If' LIPEMIA IN DIABETES The left-hand tube shows the appearance of the hlood of Case 786 (see pages 1OO, 2O6, 472) after' standing one-half hour 1 ; the right-hand tube represents the blood as it flowed into the syringe. Professor' W. R. Bloor reports the following analysis of the blood plasma. The blood was taken June 13, 1916, four-teen hour's after the last meal and no fat had been eaten for 1 the pre- ceding thirty-six hours: Fat (g 1 ycer-icles), 7.2 percent.; "lecithin," O.3 per- cent.; cholesterol (total), 1.7 per cent. Other analyses; on the same clay Allowed: blood svujar, O.27 per cent.; blood CO_>. 24 mm.. Hg. tension; alveolar air CO- ( Fricleri cia) , 21 mm. H) Cancer 07 (c) Tuberculosis .... 67 (il) Carcliorcnal and Vascular 07 (c) Infections OS (/) Inanition OS 2. Deaths with Coma 09 (a) Kther Anesthesia 70 Impaired Kidneys 70 Infections 71 () Source' of the Acid Bodies 155 (c) Percentage Relation of the Acid Bodies to One Another . 155 ((/) Mode of Elimination of Acid Bodies 15(1 2. Acidosis in Diabetes lot) () The Carbohydrate Balance 275 4. Fat 270 (a) The Increased Assimilability of Carbohydrate in Absence of Fat 277 (b) The Value- of Fat to the Diabetic 277 (c) The Danger of Fat to the Diabetic 281 5. Alcohol 282 0. Liquids .... 283 7. Sodium Chloride 283 8. Lime Deficiency in Diabetes 285 CON TENT 8 xiii SECTION V. TREATMENT. A. Prophylaxis 286 1. Early Diagnosis 287 2. Heredity 287 3. Obesity 288 4. Infectious Diseases 288 5. Pregnancy 289 6. Hygiene 289 (a) Physical Hygiene 289 (b) Mental Hygiene 290 7. Syphilis 292 8. Pancreatic Preparations 294 9. Surgery 295 B. Classification 297 1. Classification for Treatment 297 2. Classification for Prognosis 299 C. Dietetic Treatment 299 1. General Discussion 299 2. Author's Plan of Campaign against Diabetes 304 3. Summary of Dietetic Treatment 305 4. Details of Treatment 307 (a) Preparation for Fasting 307 (b) Fasting 311 (c) Intermittent Fasting 313 ((/) Determination of Tolerance for Carbohydrate .... 317 (c) Determination of Tolerance for Protein 31g (/) Determination of Tolerance for Fat 320 (g) The Caloric Needs of the Patient 321 (1) Lost Calories 322 (//) Reappearance of Sugar 323 (i) Weekly Fast Days 323 5. The Management of Mild Cases of Diabetes 324 6. The Management of Severe Casos of Diabetes 325 D. Follow-up Methods 327 E. Summaries of the Treatment Employed in Three Successive Groups of Cases 328 1. Loss of Body Weight 329 2. Loss of Body Protein 335 3. Total Calories 335 F. The Fatal Cases of 1916 336 1. Deaths without Coma 338 (a) Miscellaneous 338 (1) Inanition 338 (b) Cancer 341 (c) Tuberculosis 341 ('/) Cardiorenal and Vascular 341 (e) Infections 341 xiv CONTENTS F. The Total Cases of 1916: 2. Deaths from Coma 342 (a) and (b) Diabetes Untreated and Neglect of Preparatory Treatment 313 (c) Obesity. High Fat and Low Carbohydrate .... 344 (d) Treatment Abandoned 3-17 (r) Imperfect Supervision .... 349 (/) Increasing Severity of Diabetes 349 (&) Ether . . . 350 (//..) Carbuncle ... 350 G. Cases Unsuccessfully Treated by Fasting 350 1. Deatli Caused by the Inauguration of Fat -protein Diet in a Child 350 2. Intercurrent Infections . 352 3. Diabetes of Long Duration 358 4. Cardiorenal ('(implications 362 5. Syphilis 304 II. Special Dietetic Methods 304 1. Fasting Days 304 2. Vegetable Days 307 3. The Oatmeal Treatment 307 (a) The Place of Oatmeal in Present Diabetic Therapy . . 375 4. Other Carbohydrate Cures 370 (a) Wheat 370 (h) Potatoes 370 (c) Bananas 377 () Slow Healing of Wounds 439 (c) Exhaustion 439 ((/) Lack of Exercise 439 2. Elements Favoring Surgical Success 439 (a) Good Medical Care 440 (/>) Anesthesia 441 (c) The Introduction of Aseptic Methods 445 (d) The Avoidance of Trauma 445 3. Results 445 N. Pregnancy and Diabetes 448 1. Cases of Pregnancy Showing Large Quantities of Sugar . . . 449 2. Conclusions upon Pregnancy and Diabetes 454 3. Notable Increase in Tolerance for Carbohydrate during Pregnancy 455 O. Diabetes in Children 458 P. Diabetes in Old Age 46(> Q. Cases of Diabetes of Fifteen or More Years' Duration 408 SECTION VI. AIDS IN THE PRACTICE AND MANAGEMENT OF DIABETIC CASES. A. What Every Diabetic Should Know 470 1. The Nature of Diabetes 470 2. Faithful Treatment Accomplishes Much 470 3. Success Depends upon the Patient 471 .4. Diabetes: The Best of the Chronic Diseases 471 5. A Sugar-free Urine Spells Improvement 471 6. Treatment Depends upon Diet and not Drugs 471 (o) Carbohydrate 471 (b) Protein and Fat .471 (c) Weights and Measures Employed in Computing the Diet 472 ((/) The Quantity of Food Required 473 (c) Calories 473 (/) The Normal and Diabetic Diets Compared .... 473 (g) Observe the Method of Making the Urine Sugar-free . . 473 (/O Distribution and Exchange of Carbohydrates .... 474 (z) Special Dietetic Rules and Hints 474 7. Care of Teeth 474 8. Care of Skin 474 9. Treatment of Constipation 475 10. Exercise 475 11. Rest 475 12. Sleep 475 13. Examination of Urine 475 14. Test for Sugar 475 15. Visit to the Doctor 476 XVI CONTENTS B. Directions for Nurses in Charge of Diabetic Patients 476 C. Diabetic History Chart for the Use of Institutions and of Physicians Especially Interested in Diabetes 480 D. Chart for Dietary and Urinary Records 481 E. Actual Diets Employed to Render Patients Sugar- and Acid-free. Dietary and Urinary Charts 484 1. Diet of a Child Twelve Years Old. Case No. 923. Acidosis . 484 2. Diet of a Child Two Years Old. Case No. 938. No Acidosis . 48.1 3. Diet of a Woman Fifty-five Years Old. Case No. 759. Acidosis. Duration of Diabetes Fifteen Years 489 4. Diet of a Man Fifty-five Years Old. Case No. 181. Marked Acidosis. Duration of Diabetes Eleven Years. Intermittent Fasting- Urine Free from Sugar and Acid in Thirty-six Days 492 F. Hospital Treatment 493 G. Ambulatory Treatment 500 II. Specimen Leaf of a Book Constructed to Facilitate the Recording of Diabetic Statistics 501 SECTION VII. FOODS AND THEIR COMPOSITION. A. Diabetic Foods 503 1. Substitutes for Bread 505 () Bran Bread .... 505 (b) Gluten Bread 50(5 (c) Light Breads 50(5 (d) Various Other Substances 506 2. Substitutes for Milk 507 B. Composition of American Food Materials 508 C. Foods and Their Carbohydrate Content 511 D. Composition of So-called Diabetic Foods 519 E. Dietetic Suggestions, Recipes and Menus 530 Seven Menus for a Severe Diabetic 535 Inexpensive Menus 539 F. Tables of Equivalents 542 SECTION I. STATISTICAL STUDIES UPON THE COURSE AND TREATMENT OF DIABETES MELLITUS. A. DIABETES MELLITUS AND GLYCOSURIA. 1. Diabetes Mellitus. Definition. Definitions of diabetes are unsatisfactory, but it is safe to say that diabetes is a disease in which the normal utilization of carbohydrate is impaired, in con- sequence of which glucose is excreted in the urine. My rule in the treatment of diabetes is to consider any patient to have diabetes mellitus and treat him as such, until the contrary is proven, who has sugar in the urine demonstrable by any of the common tests. This method of procedure is safer for the patient than to make use of the term glycosuria, which begets indifference (see p. 2S(>). 2. Physiological Glycosuria. Sugar, meaning thereby glucose (dextrose, (" 6 IIi-.>O 6 ), is a normal constituent of the urine, occurring in varying quantities. Folin 1 was able by means of his technic to demonstrate the presence of sugar in nearly one hundred tests made on the urines of normal persons, and he adds : " The amount of sugar present in normal human urine is therefore probably much greater than is indicated by the negative findings recorded on the basis of the clinical qualitative tests for sugar in current use." Myers, 2 in a recent publication concludes that normal urine appears to contain between 0.08 and 0.2 per cent, sugar. He used a method of analysis similar to that employed by Benedict and Lewis for the estimation of sugar in the blood. On the other hand, the Benedict test is seldom found too delicate, although its author maintains that according to his later technic (see p. 17")) he can detect glucose in as low concentration as 0.01 to 0.02 per cent., providing the urine is of low dilution. Therefore it would seem quite impossible to demarcate sharply between normal and pathological urines with reference to the sugar output. 3. Alimentary Glycosuria. The sugar may appear in the urine after a meal rich in carbohydrates. If the carbohydrates are in ' Fnlin: Jour. Biol. f'lieni., 191"), xxii, p. 327. 2 Myers: Proc. Soc. Exp. Biol. and Mod., 1911), xiii, p. 17-V 2 (17; IS DIABETES MELIJTTS the form of starch (alimentary glycosuria e ainylo) it would .signify diabetes mellitus, but if the carbohydrates are in the form of sugar (alimentary glycosuria e saccharo) it might or might not signify diabetes. To determine this point Xaunyn 1 suggests the admin- istration of 100 grams of dextrose two hours after a breakfast of a large cup of coffee and milk and 80 to 100 grams of bread. Jf sugar is then demonstrable in the urine in a quantitatively estimable amount, diabetes mellitus exists. 4. Other Glycosurias. The drug glycosurias and those of trau- matic and emotional origin are almost invariably of a temporary nature, so that doubt concerning the diagnosis of diabetes mellitus vanishes when sugar is constantly found in the urine. 5. The Distinction between Glycosuria and Diabetes. The great variety of conditions under which sugar appears in the urine has been discussed in detail by Xaunyn, and more recently by Allen.' 2 In this latter volume diabetes and the allied glycosurias are approached from a refreshingly new stand-point. Allen proposes two laws by which glycosuria can be distinguished from diabetes. The first follows: Allen's Paradoxical Laic "Whereas in normal individuals the more sugar is given the more is utilized, the reverse is true in diabetes." Allen shows that the limits of tolerance in non-diabetic animals are all apparent, not real. Xo real limit of the power of utilization of sugar exists except death. This law applies to all species of animals which have been tested, to all methods of administering sugar, and to all sugars and carbohydrates, provided they are utilized at all. This paradoxical law of dextrose sharply distinguishes diabetic animals from every type of non- diabetic animals. In diabetic animals the limits of tolerance are real and not apparent. In totally diabetic animals the injection of dextrose causes an excretion of glucose not only equal to but frequently greater than the doses injected. In milder diabetics not only is the proportion of excretion to injection generally high, but the assimilation may be made worse instead of better by an overdose just the opposite of the paradoxical law. This law enables distinctions to be drawn experimentally between diabetes and nervous glycosuria, whether of organic or functional type. Kvery form of non-diabetic glycosuria retains the power of utilizing dextrose in amounts increasing with the dose. This law proves that the lowering of tolerance in diabetes is totally different from the lowering of tolerance in other conditions; that diabetes is not a simple overproduction of sugar, and herein lies the chief value of Allen's conception because it has been experi- mentally proven that where there is a simple overproduction of sugar the paradoxical la\v holds. It shows that diabetes is not 1 Xaunyn: Der Diabetes Melilus, Wien. 1!)()G. p. .'17. '-Allen: ( llyrosuria and Diabetes, 1'JKi, Harvard University Press. STATISTICS RELATING TO DIABETES 19 explained by an excess of renal permeability, by overaction of certain glands, or by states of the nervous system, either functional or organic, which may be accompanied by glycosnria. Earlier investigators as well as Allen were able to show that all common sugars are antidinretics, and diminish the output of urine. In other words, as respects diuresis, these sugars when given intra- venously obey the laws of crystalloids, and when given otherwise obey the laws of colloids. Salts and other diuretics have a rather similar diuretic action by all modes of administering, when allow- ance is made for the rate of absorption. In diabetes this difference is abolished as respects dextrose. In diabetes, dextrose acts as a diuretic when it is given intravenously, orally, subcutaneously, or otherwise. In other words, it obeys the diuretic laws of a salt. This is Allen's second differential test between diabetes and glyco- suria but whether it will hold, Woodyatt's 1 recent work on the rate of intravenous assimilation of glucose makes it doubtful. The further development of his researches along this line is awaited with interest. It appeared to Allen that dextrose exists in the normal body in a state of colloid combination; as a colloid it is utilized by the tissues, and, like other colloids, diminishes diuresis. It assumes the colloid form in passing through any living membrane, never circulating in crystalloid form in the normal organism except after direct intravenous injection. On the other hand, in the diabetic organism dextrose circulates as a free or very poorly com- bined crystalloid, and in this state is not available to the tissues and is a diuretic like every crystalloid. Allen also concludes that the sugar of the normal body is com- bined with some substance in colloid form which makes it available to the tissues. This combining substance he refers to as an ambo- ceptor. When the pancreas is absent or insufficient, dextrose occurs free, and he therefore assumes that this combining substance is furnished by the pancreas, and accordingly used the term "pan- creatic amboceptor" synonymously with the "internal secretion of the pancreas." Allen defines diabetes as deficiency of pancreatic amboceptor. B. STATISTICS RELATING TO DIABETES. 1. Increase in the Incidence of Diabetes Mellitus. If diabetes should continue to increase in the next thirty years at the same rate statistics show it has increased in the last thirty years it would rival tuberculosis as a cause of death, and if this rate progressed for another generation diabetes would be responsible for almost the entire mortality of the world. Such a rapid rate of increase is evi- 1 Woodyatt, Sansum and Wilder: Jour. Am. Mecl. Assn., 1915, Ixv, p. 21)07; see also p. 255, below. 20 DIABETES MELLITUS deuce of itself that a fallacy exists somewhere in the statistics. Convincing proof that this is the case is furnished by u comparison of the statistics for the registration area of the United States (luring the years 1010, 1911, 1012, and 1913, because during these years the death-rate for diabetes was nearly stationary. Such a sudden halt in the progressive frequency of the disease could not have been brought about without some obviously remarkable improve- incMit in treatment, preventive or otherwise. This we know did not take place in 1910. This striking interruption in the advancing incidence of the disease is also indicated by the statistics of New York and Boston. Thus, in Boston during 1912 and 1913 the death-rate from diabetes was less than during 1910 and 1911. lint since 1913 a remarkable advance is registered throughout the country and in New York and Boston as well. This rapid rise in mortality in 1914, 191") and 191(> 1 believe attributable to the increased attention devoted to this disease by the medical pro- fession and the newspapers following the announcement of a new method of treatment. The following tables illustrate the rapid statistical increase in the frequency of diabetes and the acceleration of the increase which this has recently undergone: TAHLE 1. THE REGISTRATION AREA OF THE TNITED STATES. Deaths from all causes. 1 Deaths from Diabetes. Kate per Hate per 1000 100,000 Census year. 2 Xumbcr. population. Number. population. 1880 169,453 19.8 237 2.8 1X90 380,212 19.0 1089 5.5 1900 512,009 17.8 2093 9.3 TABLE 2. THE REGISTRATION; AUKA OF THE UNITED STATES. Deaths from diabetes. Kate per 100.000 Number. population. 2,990 9 7 3,212 10 2 3,312 10 3 3,081 11.3 4,259 12 8 4,397 12.9 5,3;^ 12.7 5,801 13 5 0,274 13 1 7.024 13 8 8,040 14.9 8.S05 11.9 9.045 15.0 9,000 15.3 10.000 10.2 11,775 17.5 Calendar year. Number. population. 1900 . . . . . 539, . 139 17.0 1901 . . . . . 518,207 10.5 1902 . . . . . 508,040 15.9 1903 . . . . . 524,415 100 1904 . . . . . 551,354 10.5 1905 545 53:', 10.0 1900 058 105 15.7 1907 . . . . 087,034 10.0 1908 091 574 14 8 1909 . . . . . 732,538 14.4 1910 . . . . . 805,112 15.0 191 1 839 2X4 14.2 1912 . . . . . 83X.251 13.9 191:5' . . . . 890.848 14.1 1914 . . . . 89X.059 1:5.0 1915 . . 909.155 1 3 . 5 "The rouiMnition urea for I'.M ... , population of the L'nited State's, for l'.)14, 0(i.,s per cent., and for I'.U,}, (17.1 pei STATISTICS RELATING TO DIABETES 21 In Fig. 1 are displayed the falling rate of deaths from all causes per 1000 and the rising rate of deaths from diabetes per 100,000 during the last thirty-five years in the registration area of the United States. Disregarding the figures for the death-rate from diabetes prior to 1900 as quite inaccurate, it will be seen that the mortality from the disease in the registration area of the United States has increased statistically between 1900 and 1915 by 80 per cent., whereas there has been a decrease of 24 per cent, in deaths from all causes. 1880 ISO!) 1910 1915 Fui. 1. The falling rate of deaths from all causes per 1000 and the rising rate of deaths from diabetes per 100,000 between the years 1880 and 1915 in the regis- tration area of the United States. 1 believe that this increase will continue to occur for a good many years to come, and in fact until examinations of the urine of each individual in the community are made not only with each illness, but also once or twice yearly. Certainly until that time is reached, the statistical increase in diabetes must be explained simply on the ground of more accurate diagnosis. Save for the diseases of the arteries, diabetes has increased more rapidly as a cause of death in the registration area of the United States in the first decade of this century than cancer, cerebral hemorrhage and apoplexy, organic diseases of the heart, cirrhosis of the liver or B right's disease. 1 During the last thirty-five years the general death-rate of New York City has been cut in half, but the death-rate from diabetes has increased statistically sixfold. (See Table 3.) 1 Dublin: Popular Science Monthly, April, 1915, Ixxxvi, p. 313. 22 DIABETES MELLITUS TABLE 3. 1SSO 1X00 1000 1010 1011 1012 1013 1014 1015 1010 Total Old City.' 31,037 40,103 43,227 45,02.8 45,324 43,102 43,1X0 44,171 43,704 43,407 70,872 70,742 75,423 73.00S 73,002 74, so:; 70,103 77,soi 3. XEW YORK CITY. Death-rate per 1000 Old Creator City. City. 20 .1 Deaths from diabetes. Number. Kate per 100,000. Old Creator Old Greater City. City. City. City. 44 . . 3 . 21 .0 130 X .0 21 .0 20 . 233 357 11 3 10 ,4 10 .4 16 .0 440 70S 10. 10 15 . S 15 1 421 730 14 7 14 8 It .0 14 .1 521 870 17 ,-> 16 8 14 4 14 2 500 8S4 16 17 14 4 11. 540 070 17 8 is 3 14 ,0 13. 000 1100 19 3 20 2 13, G 13. 050 1118 20. 2 20 ,0 Boston shows a higher mortality from diabetes than any city with which I am acquainted save Berlin. It will be of the greatest interest statistically to note if the mortality in other sections of the community continues to approach the level reached here. To a certain extent the city with the highest mortality serves as an index to which the mortality in other communities may be expected to rise. TABLE 4. BOSTON. Deaths from diabetes. Deaths, from Death-rate diabetes. 31 77 135 138 136 138 144 195 193 Dentlis from all causes. Deaths, Death-rate all causes. per 1000. 1880 . . . 8,531 23.51 1890 . . . . 10,181 22 . 66 1900 . . . . 11,678 20 . 80 1910 . . . . 11,574 17.20 1911 . . . . 11,767 17.22 1912 . . . . 11,643 16.34 1913 1 1 839 16 . 34 1914 . . . . 11.831 16.06 1915 12 01 S 1 6 . 06 1916 . 12,776 16.80 6.90 13.71 20.07 20 . 20 19.09 19.05 19.55 26 . 06 25 . 38 TABLE 5. BERLIN. 3 DEATHS FROM DIABETES PKH 100,000. Males. Females. Total. 1871-1875 .... . . . . 2.2 1.2 1.7 1876-1880 .... . . . . 4.0 2.0 3.0 1881-1885 .... . . . 4.7 2.6 3 . 6 1886-1890 .... . . . . 6.3 3.9 5.1 1891-1895 .... . . . . 9.4 5 . 5 7.4 1896-1900 .... . . . . 11.7 7.1 9.3 1901-1905 .... . . . . 20.4 12.4 16.2 1906-1910 . . 23 4 14.8 IS 9 1911 . . . . 25.2 17.6 21.3 1912 24.8 19.5 22.0 The effeet of the limited rations incidental to the European War upon the frequency of diabetes in Europe will surely be striking. ' 1 Old C'ity of Xew York includes present boroughs of Manhattan and the Bronx. 1 ( I renter ( 'ity of Xew York includes all boroughs. '' The data for the two periods 1S71 1S75 and 1x70- 1SSO were obtained from von Xoorden, Die Zuekerkrankheit, Berlin, 1012. p. 55; the remaining figures were kindly furni.-hod me l>y Mr. Samuel 1.. Ko<_'er>, I tirector of the Bureau of the C'ensus, Washington. See "Die ( leslorhenen nach Todesursaehen " in the Statist isclies Jahrhueli der Stadt Berlin, vols. ix to xxxii and vol. xxxii, p. 1, for "Mittlere Bevol- keruiiK nach dem Ceschleeht von 1x10 l>is 1011;" also "Taliellen iiber diu Bevolker- unfrsvorfiaiifie Berlins" for 1010 and 1011. ' GottsAein and Unibev: Jour. Am. Med. Assn., 1917, Ixvii, p. 70. STATISTICS RELATING TO DIABETES 23 In Paris, too, the increase of deaths caused by diabetes is evident and T am indebted to Air. Samuel L. Rogers, Director of the Bureau of the Census for the statistics in Table (>. The table shows the deaths from diabetes per 100,000 in Paris in each of the years between 1880 and 1912, both inclusive. TABLE G. PARIS. DEATHS FROM DIABETES PER 100,000. 1880 ... 6.3 1891 . . . 12.9 1902 . . . 14.6 1881 . . . 6.8 1892 . . . 12.4 1903 . . . 15.1 1882 ... 7.4 1893 . . . 14.4 1904 . . . 16.0 1883 . . . 6.1 1894 . . . 12.5 1905 . . . 16.7 1884 ... 9.2 1895 . . . 15.3 1906 . . . 17.6 1885 . . . 11.7 1896 . . . 15.0 1907 . . . 17.9 1886 . . .11.5 1897 . . .16.0 1908 . . .17.0 1887 . . .12.9 1898 . . 15.8 1909 . . .19.3 1888 . . .13.1 1899 . . . 15.3 1910 . . . 18.4 1889 . . . 13.8 1900 . . . 17.0 1911 . . . 16.9 1890 . . . 13.5 1901 . . . 14.8 1912 . . . 17.7 The statistical increase in diabetes in various countries and cities has been instructively tabulated by Magnus-Levy. The table is so arranged that the number of deaths for the first date is set at 100 and the relative number at the later date is shown by the figures in the right-hand column. TABLE 7. MAGNUS-LEVY. 1 1. United States .... 1850 100 1880 266 2. England 1850 100 1895 310 3. Italy 1888 100 1901 176 4. Prussia 1870 100 1890 pa. 450 5. Berlin 1873 100 1903 940 6. Paris ca. 1870 100 1900 520 A similar progression in the frequency of diabetes is also seen in the statistics of England and Wales: TABLE 8. ENGLAND AND WALES. - Deaths from Diabetes per 100,000. 1851 . . . 2.; 1852 . . .2.: 1853 . . . 2.: 1854 . . . 2.- 1855 . . . 2.- 1856 . . . 2.: 1857 . . . 2.; 1858 . . . 2.; 1859 . . . 2.; 1860 . . . 2.1 1861 . . . 2.1 1862 . . . 2.< 1863 . . . 2.; 1864 . . . 3.1 1865 . . . 3.1 1866 . . . 3.5 1867 . . . 3.1 1868 . . . 3.1 1869 . . . 3.- 1870 . . . 3.) 1871 . . . 3.. r 1 Magnus-Levy: Spez. Path. u. Therap. inn. Krank., Kraus u. Brugsch, Berlin 1913, Bd. i, S. 39. 2 Registrar -General's Report of England and Wales. 1872 . , . 3.3 1893 . . . 7.0 1873 . . . 3 . 5 1894 . . . 6.8 1874 . . . 3.7 1895 . . . 7.5 1875 . , . 3.9 1896 . . . 7.4 1876 . . . 3.7 1897 . . . 7 . 8 1877 . . . 4.1 1898 . . . 8.2 1878 . . . 4.2 1899 . . . 8.5 1879 . . . 4.1 1900 . . . 8.6 1880 . . . 4.1 1901 . . . 9.1 1881 . , . 4.7 1902 . . . 8.4 1882 . . . 4.7 1903 . . . 8.5 1883 5 1 1904 9 .3 1884 . . . 5.4 1905 . '. '. 9.3 1885 o o 1906 9 7 1886 . , . 5 . 9 1907 . . ' 9^7 1887 . . . 6.2 1908 . . . 10.3 1888 . . . 6.2 1909 . . . 10.4 1889 . . . 6.0 1910 . . . 11.0 1890 . . . 6 . 5 1911 . . . 10.7 1891 . . . 6 . 6 1912 . . . 11.1 1892 . . . 6.8 24 DIABETES MELLITU8 The same upward tendency in the prevalence of diabetes is noted in Japan. This is shown by Iwai's 1 reports of admissions to the lied Cross Hospital and the Hospital of the Imperial Utiversity in Tokyo, as exhibited in Table 9. TABLE 9. INCIDENCE OF DIABETES IN TOKYO. RED CROSS HOSPITAL. Per font, of Oases of admissions Year. Admissions. diabetes. for diabetes. 1891-1S93 1,134 7 0.61 1S97-1900 1,367 13 0.95 1901-1903 1,449 17 1.17 1906-1908 1,006 16 1.59 HOSPITAL OF IMPERIAL UNIVERSITY, TOKYO. 1900-1902 24,424 59 0.24 1903-1906 39,016 135 0.34 1907-1910 47,289 207 0.43 No accurate figures are available for the total number of diabetic patients in the United States. Since approximately 1 death out of each 75 in the United States is due to diabetes, one might be tempted to assume that 1.3 per cent, of the living population will have the disease. Support for this estimate is furnished by the statistics of Barringer, 2 based upon the frequency with which sugar was found in the urines of 72,000 s adults examined for life insurance in New York City. His method of investigation showed that nearly 2 per cent. 4 of this large group of the adult population in a city had diabetes. It would be unjustifiable to apply this conclusion in this form to the entire population, because diabetes is comparatively rare in the early part of life but increases in frequency as age advances, and Barringer's data relate to adults. The incidence of diabetes might be far below 1 in 75 in the first and second decades of life and then increase to 1 in 50 or even 1 in 40 of the population from the sixth decade onward. Today our information rests upon hypothetical grounds, but within a few years far more satisfactory information will be available. For the present it is probably underestimating the number of diabetics in the country to state that approximately 1 in 75 of all individuals cither have the disease or will develop it. By this method of reckoning there arc over 1,000,000 potential diabetics in the country. Should an estimate be based upon the total deaths from diabetes 1 Iwai: Lc DiulxMc Sucre clu>z Irs Japouais. Tokyo. Translated by L Paris; Masson et, ( 'ompanie, IvlitiMii - Harrinucr: Arch. Int. Mcd.. 1'JD!) :l Actually 71,712!). * l.S'Jo per rent. STATISTICS RELATING TO DIABETES 25 annually, a different result would be reached. Thus, in the regis- tration area of the United States for J915 the deaths from diabetes were, in round numbers, 12,000, and on this basis for the whole country 18,000. Allowing a duration of five and half years for each- case, the total number of diabetics would be 100,000. Personally I believe the number of cases of diabetes in the United States is much nearer to 1,000,000 than to 100,000. 2. The Explanation of the Statistical Increase in Diabetes. (a) Greater Accuracy of Vital Statistics. The increased accuracy of vital statistics in recent years undoubtedly explains, to a great extent, the apparent increase in diabetes. It is evident from the study of the incidence of diabetes in different localities that the rate is highest where one would expect the statistics to be most trust- worthy. (6) More Frequent Urinary Examinations. The chief cause for the improvement in the vital statistics upon diabetes lies in the increased frequency of routine urinary examinations. Urinary examinations are comparatively recent. This is shown by the records of a case of diabetes at the Massachusetts General Hospital for the year 1866, where the words ''urine tasted sweet" may be seen in the hand- writing of a house pupil, later a professor in the Harvard Medical School, and now still vigorous. The frequency of diabetes in a community may be the index of the intelligence of its physicians. (c) General Increase in Duration of Life. Another cause for the apparent increase in the frequency of diabetes is connected with the general increase in the duration of life throughout the world. This is of more importance in studying the incidence of diabetes than it would be with the infectious diseases. The infectious diseases are common to the early years of life, and consequently prolongation of life would affect their incidence comparatively little; but with diabetes quite the opposite is the case. I am under the impression that just as in cancer so it is with diabetes, the longer a person lives the more liable he is to the disease. This increase in the duration of life is really considerable. The approx- imate age at death from all causes in the registration area of the Unites States in 18(50 and for following decades is shown in the subjoined table : TABLE 10. THK AVERAGE AGE AT DEATH ix THE REGISTRATION AREA OF THE UNITED STATES. 1800 22 . 7 years 1870 25.2' 1880 2t>.9 1890 31.1 1900 35.2 1910 38.7 1915 42.2 20 DIABETES MELLITUS r rhis increase in the average duration of life is also evident from German mortality tables: TAHLK 11. 'I'm; AVEKACH; KXPKCTATION' OF LIFK AT BIRTH ix (!KKMA\Y. ' Men. Woinon. 1S71-1SSO 3. r >.t> 38.4 1SS1-1SW) 37.2 40.2 1891-1000 40.1) 44.0 1901 -19 10 . 44.8 48.3 In Sweden the average length of life is said to be about ten years longer than in Germany. In quite a different manner this question has been approached by the Department of Health of New York. According to a table recently published, the life table based upon mortality during the years 1909 to 1911 inclusive is compared with a similar table for the years 1879 to 1881 inclusive. This shows the expectancy of life at stated ages for males and females. It appears that children under five years of age in the former period had an expectancy of forty-one years and that today a child of that age may look forward to an expectancy of fifty-two years, or an increase of eleven years of life. The life of a child of ten years has been prolonged more than five years. An individual from twenty-five to thirty years of age had an expectancy thirty years ago of 32.0 years, while now his expectancy is 34.3 years, an increase of 1 .7 years. On the other hand, individuals between forty and forty-five years old had an expec- tancy of 23.9 years in the earlier period, but at the present time 23.4 years, a decrease in the expectancy of six months. Dublin's 2 explanation of this phenomena is reassuring to native Americans, for he ascribes it to the influx of approximately 22, ()()(),( )()() immi- grants in the last fifteen years. Almost invariably the expectation of life in the countries from which these came was below our own. lie has also demonstrated for the State of New York that the mortality rates for native born of native parentage are far more favorable than for the native born of foreign or mixed parentage or for the foreign born. (d) Life Insurance Statistics. The statistics of insurance companies are direct evidence against any such rapid increase in the frequency of diabetes. This is shown by the records of two of the large insurance' companies in the United States in the number of rejections for sugar for successive years. (See Table 12.) 'See Stalislik '2'1. STATISTICS RELATING TO DIABETES 27 TABLE 12. REJECTIONS ON ACCOUNT OF CLYCOSURIA BY Two LARCIE INSURANCE COMPANIES. COMPANY A. Applications declined on Applications account of Hate per Year. received. suj;ar. 1000. i9i;? 501 1914 492 1915 574 COMPANY B. 1910 10,596 50 5.29 1911 11,709 77 6.58 1912 13,851 80 0.21 1913 13,805 80 5.79 1914 13,748 73 5.31 1915 15,311 90 7.27 Unfortunately I have been unable' to secure data from other of the larger insurance companies which would show whether the percentage of rejection on account of sugar being found in the urine had increased during the last thirty years. So far as I can learn from correspondence and conversation with those in insurance companies who are in a position to know, the percentage of rejec- tions has remained constant for many years. (c) Frequency by Decades at which Onset Occurs Now and Formerly. An increase in the incidence of diabetes should also be manifested in each decade of life if such an increase was actual. Von Noorden has summarized such statistics, and at first sight they favor the view that the increase in diabetes is actual and not apparent. (See Table 13.) It will be seen that successive writers at successive periods report a similar distribution of their cases among the different decades. This conclusion is not justified, because most of the writers quoted, although living at somewhat different periods, nevertheless lived before the time when urinary examinations were really part of the general routine. Von Xoorden's statistics represent a later era, and it is noticeable that lie had at least 50 per cent, more cases in the first decade of life than earlier writers. My own statistics being still more recent bring out this point even more strongly, for they show three times as many as von Xoorden's 1 1.43 per cent. namely, 5S cases or 4.9 per cent, in contrast to 1 per cent, or less which writers prior to von Xoorden report. Of GOX cases of dia- betes recently reported by Iwai, 2 in Japan, not a single case was encountered in the first decade of life! Writing even as late as in 1015, De Lange and Schippers 3 mention only 500 to GOO cases of diabetes in children being upon record. The point may be raised that on account of my interest in diabetes and the feeling of 1 See von Xoorden: Die Zuckerkrankheit, Berlin, 1912, p. 59. - Iwai: Lor. cit., p. i>4. 3 De L:inge and Schippers: Jour. Am. Mod. Assn., 1915, Ixv, p. 1150. 28 DIAKETES MELLITUS the hopelessness of the disease in children a proportionately larger number of vomit; than of older diabetics may have come to me. This is undoubtedly true, but this argument applies to the other writers cited with even more force. OF ONSF.T OF DIABF/TFS BY DKCADKS. 4. . 5. 0, 7 g. Per cent. 18.0 25 . 20.0 1 1.0 1.0 " 10.0 24.0 30.0 10.0 . 5 " 11.2 23 . 1 39.5 IS 1 3.4 " 17.3 22 . 3 32.0 10.0 3.3 " 1(1.4 24 . 9 30.7 13 .4 2 5(1 " 17.2 3H.O 20.8 9.2 1 " 10.0 21.0 17.7 4 04 0.43 mild. !t . 57 12.57 11.0 2 14 severe and mo< severe. 13.8 24.2 24 . 2 11.9 2.S Joslin . The dates of publications of the authors cited are given in the foot-notes on this page. This increase in the percentage of cases in the first decade as compared with a generation ago therefore speaks in favor of the better diagnostic methods of today rather than of actual increase in the frequency of the disease. (/) The Importance of Recognition of the Extent of Diabetes in the Community and its Influence upon Treatment. The recognition of the large number of diabetic individuals living in the United States is important. It determines the character of the treatment. The number of cases is so great that it at once becomes evident that their care must rest in the hands of the general practitioner. It is ridiculous to expect that the treatment of diabetics should be under the supervision of a specialist, neither can they all undergo hospital treatment. It is plain that a program of education not only of the diabetic patients themselves but the population as a whole must he instituted along similar lines to that which has been adopted in the treatment of tuberculosis. Hut first of all this program must begin with the doctor. The campaign against tuberculosis has given the community fresh air. A campaign for the prevention and treatment of diabetes should give the com- munity a knowledge of diet and the importance of physical exercise. The gain to the community from the dissemination of such knowl- edge will almost oil'set the harm caused by diabetes. In this campaign I would especially emphasize the importance of the education of the general practitioner in dietetics. To treat the 1 Freriehs: T'eber don Diabofes, Berlin, 1SS4. "Soe-icn: Der Diabetes Melims, Merlin, I Aufll., 1S70, III Aufl., IS'.):}. 3 Cirube: Diiitetisehe BchandlunK tier Zurkerkruukheit, Morm. is'.ls. 4 Schinit/: Merl. klin. \Vorh., 1S7:{, xviii, S. KS; also Merlin, 1S91, S. :57:5. : 'Pavy: On Diabetes, London, IMi'.l; Different iat ion in Diabetes, London, 1900. fi Ki'ilz: Klinisrhe Krfahnmnon i'lber Diabetes Melitus, Herausg. von Kunipf, Jena, 1899. 7 Von Xoorden: Lor. cit., Berlin, 1912. TREATMENT OF DIABETES MELLITUS 29 100,000 to 1,000,000 diabetic patients in the United States there were in 191() 145,241 physicians, so that a physician will rarely see over 5 to 10 diabetics a year. Under normal circumstances he cannot give much time to the study of so few patients. In other words, suggestions for treatment must be of the simplest character. The greatest mistake that can happen in the treatment of diabetes today would be to advocate dietetic methods which would involve complicated analyses of urine and blood. The greatest advantage of the fasting treatment introduced by Dr. Allen lies in its simplicity and in the removal of the need for quantitative urinary examinations. C. THE IMPROVEMENT IN THE TREATMENT OF DIABETES MELLITUS. For the successful treatment of a chronic disease it is essential that the physician be convinced that his methods of treatment are imp r oving. This I thoroughly believe to be the case in diabetes mellitus. It is freely granted that occasionally one sees diabetic patients in middle life who claim that they have disregarded treat- ment and yet have maintained health and strength, but investiga- tion usually shows (1) that the type has been mild; (2) that the individual is unusually intelligent, and (3) although he disclaims it, that he has really adopted a mode of life and diet essentially con- sistent with sound treatment. Case Xo. 354 is one in point. While at the head of a large insurance business at the age of fifty, sugar was found in his urine. Although he claimed to have disregarded routine treatment, he really lived on a diet containing little sugar and much fat. Sixteen years later sugar disappeared permanently from the urine and he was able to eat sweets without harm. Al- bumin and casts then appeared, the blood-pressure rose to ISO mm., and the patient died at the age of seventy-one years. 1. Indirect Improvement. Various factors have contributed indirectly to an improvement in treatment such as an early diagnosis, a better understanding of the diet and the disease, and the better treatment of complications. (a) Early Diagnosis. Success in the treatment of diabetes, as well as of consumption, depends upon an early diagnosis, and all will agree we are in a much more favorable position to make early diagnoses now than were the physicians of a former generation. 1. liuntine Urinary Examinations. The favorable outcome in a considerable percentage of the cases can be attributed to the early detection of the disease. Xaimyn especially urges the importance and good results of early treatment. A notable percentage of the total number of my diabetic patients were apprised of their disease, not so much because of their symptoms, but because of routine urinary examinations. It appeared, therefore, of interest to examine ;]() DIABETES MELUTUS into the duration of life of the living and fatal cases of this group, because of the possibility that the duration of the disease among these patients might be longer or shorter than the average of all the patients, With this end in view I constructed Table Y.V in my first edition, but a more careful attention to this point during this last year has led me to discard it, because of the difficulty of deciding whom 1 should include in this class. The issue, however, is ade- quately met in a statistical way in the scries of cases which com- prises those diabetics who became cognizant of their disease through examinations for life insurance (see p. IOC)), When a routine urinary examination is neglected, the damage done to the patient is considerable, and to the physician's reputation is extreme. The physician is never forgiven! The time is probably not far distant, if not already here, when the question will be asked of the diabetic: " How long did the disease exist before it was discovered?" and "Who was your physician?'' We are familiar with the same questions in tuberculosis. 2. Examinations for .Life Insurance. An early recognition of the disease is promoted by examinations for life insurance. The percentage of the population in the I'nited States which is undergoing such examination is steadily increasing. This is shown in Table 14, which is based upon the number of policies, not includ- ing industrial insurance, in force at the end of the years indicated as shown by the Spectator Company's Year-book for li)10. TABLE 14. LIFK I \sru.\\rK POI.ICIK.-- ISSTHD i\ THE UNITED STATES. IVr cent, of populn- Popuhition. tion buying policies. f>o,if>r>,7x:5 " i (12,947,714 2 7"),!)'.M..")7r) 4 ) 7 100. :;<><>.:; ix (Vst.) S The above figures include policies in force in all companies in the country. The number of individuals carrying one or more extra policies reduces the total number insured, but this is more than offset by those' examined for insurance, and then refused. Seventy-six cases of my series discovered the disease as a result of an examination for life insurance. All of these cases were men. In other words, life insurance examinations were responsible for the detection of diabetes in 1 1 per cent, of my male cases above the age of ten years. It is acknowledged that in many the disease was undoubtedly mild, but its detection afforded an unusually early opportunity for treatment. The favorable outcome of these cases is encouraging. (See Table MO, p. 107.) Year. 1XSO . . . IX'.H) . . . Total. . . f.OX.IiXl . . l,27l>, 107 1 '.)()() . . . niio . . . i'.iir> . . . :*.()7l.2f>:i . . 0,040,017 S.284.281 TREATMENT OF 1)1 ABET E8 MELLITUH .'51 (/;) Knowledge of Diet More General. An accurate knowledge of the diet is responsible for a large share of the improvement. For this we are indebted first to the publication by the United States Government of Bulletin No. 28, l and especially to the vigorous campaign conducted under the auspicies of Professor Street at the Connecticut Agricultural Experiment Station in bringing to com- mon knowledge the carbohydrate content of so-called diabetic foods. Twenty-five years ago a knowledge of the carbohydrate content of the food of an individual was almost negligible. One is free in making this statement because even today relatively few physicians can reckon up the carbohydrate in the diet of their patients. For this lack of instruction in diet medical teaching is largely responsible. I know that as hospital assistants our estima- tion of the carbohydrate which our patients took began and ended with the statement that most gluten breads were fraudulent. It is not at all strange that physicians in general practice are unacquainted with carbohydrate food values, because it is only within a decade 2 that such information has begun to appear in the literature and a much longer period is required for it to become widely circulated. (c) The Disease is Better Understood. It is no longer considered almost miraculous for a patient to void a urine with 11 per cent, of sugar upon one day and a few days later to become sugar-free. It is recognized that a blood sugar of 0.4 per cent, falls slowly and that with renal complications may not reach normal, and that under these circumstances sugar may persist for days or even weeks in the urine. Successes with bizarre methods of treatment, such as an exclusive buttermilk diet, are now explained by the accompanying undernutrition which they entailed. The merits of a new method of treatment are more quickly determined today than a generation ago, and this contributes to better treatment. The number of new remedies exploited upon the profession and upon the laity is gratifyingly growing less. Physicians realize better than formerly that the value of any remedy in the disease must depend upon its power to increase tolerance for carbohydrates, and as no remedy so far has been shown to be capable of bringing this about, confidence in drugs is less frequently secured. This means much, because formerly valu- able time was lost and the strength of many patients exhausted in useless therapeutic procedures. (<7) The Better Treatment of Complications. Complications are less numerous at the present time, and when they do occur are more intelligently treated. The older authorities believed tuberculosis to 1 Chemical Composition of American Food Materials, Bulletin Xo. 2s, U. S. Department, of Agriculture. This can he obtained by sending 10 cents in coin to Superintendent of Documents, Washington, D. C. - Bulletin No. 28 first appeared in 1906. 32 DIABETES MELLITUS be about the worst foe of the diabetic patient, and they were prob- ably right in their conviction. Today preventive medicine removes tliis danger to a large extent. Similarly, septic processes are avoided. Frerichs 1 found in 55 autopsies upon diabetic patients tubercu- losis to be present twenty-five times. In his private practice at Frankfort, von Noorden- noted tuberculosis in 5.5 per cent., but in his hospital practice 15.1 per cent. In Vienna the percentage rose to 27. Among my cases only 2.9 per cent, were known to have open tuberculosis, and among 241 new cases seen during the year ending December 1, 1910, tuberculosis was demonstrated to be active in 3, or 1.2 per cent. 2. Direct Improvement Shown by Statistics. Positive proof of improvement in treatment would be afforded if it could be shown that the average duration of life of diabetic patients was greater today than twenty years ago. This I unquestionably believe to be true. In order to secure information upon this point 9X per cent, of 945 cases of diabetes seen by me in private practice from 1X93 to November IS, 1915, were traced. Of these 922 cases it was learned that 420 had died. One year later, December 1, 1910, another compilation was made and the condition of 97.4 per cent, of the 11X7 cases seen since 1X94 was ascertained. The fatal cases now number 5 Hi. The second series of data, therefore, will serve for comparison with the former, and if the treatment of diabetes is improving, some indication of it should become apparent. Both series of cases have been grouped in decades according to the onset of the disease. The actual onset of the diabetes in each case was sought with care and for a large number of the cases the records not only show the probable onset, but the date on which the diag- nosis of the disease was proven by urinary examination. Nearly all the histories have been personally taken by me, and my records there- fore are uniform. Before recording my own statistics reference should be made to the data of Naunyn and von Noordcn and the statistics of the City of Boston and of the Massachusetts General Hospital. (a) Massachusetts General Hospital, 1824-1898, 1898-1913, 1913- 1916. These data are of especial value because they show the duration of life of diabetic patients for three 1 epochs. The first period includes all eases from the beginning of the records of the hospital in 1>>24 1o 1X9X and the second from 1X9X to 19K5- a period which is sufficiently remote to exclude the results of modern treatment. The third period (191:5 1910) represents one year of the older methods of treatment, the first year of modern treat- ment (a transitional year) and one complete year under the con ditions of the treatment today. For the first period I personally examined each history upon the medic;!! side of the hospital, for in those early years there was no index, and collected the diabetic cases and recorded whether the information as to duration was trust- 1 Freriohs : ( 'iiivl by Xaunyii, lor. cit .., p. 2 I'.). - Von Xoorden : Lor. fit., p. 196. TREATMENT OF DIABETES MELLITUS 33 worthy. These data with others were published conjointly with R. H. Fitz. 1 The statisties for the years LSUS to 1913 and 1913-1916 were collected, under my direction, by my secretary, Miss Helen Leonard, and the duration of each case verified by my former assistant, Dr. F. Gorham Brigham. For the privilege of inserting these data, which are of value as a measure of the improvement in treatment, I am deeply indebted to the Superintendent and Staff of the Massachusetts General Hospital. TABLE 15. THE DURATION OF LIFE OF CASES OF DIABETES WHO DIED AT THE MASSACHUSETTS GENERAL HOSPITAL BETWEEN THE YEARS 1824-1898. Duration, years. Number of cases. Percent. 0-1 27 67.5 1-2 7 17.5 2-3 3 7.5 3-4 1 2.5 4-5 5-6 1 2.-5 6-7 1 2.5 40 Of the total number of 172 diabetic patients treated during these seventy-four years, 47 or 27 per cent, died within the hospital walls. The duration of 7 of the fatal cases could not be determined. TABLE If). THE DURATION OF LIFE OF CASES OF DIABETES WHO DIED AT THE MASSACHUSETTS GENERAL HOSPITAL BETWEEN THE YEARS 1898 TO NOVEMBER, 1913. Duration, years. Number of cases. Per cent. 40.6 18.8 8.7 8.7 2.9 7.2 1.4 2.8 \A 1.4 1.4 2.9 1.4 Total 79 Of the total number of 280 diabetic patients treated during these years 79 or 28 per cent. died. Whereas the percentage of patients who died in the hospital during the two periods was the same, it is evident that the duration of life of the patients was distinctly greater in the second period. The most striking feature, however, is the lessened fall in mortality during the first year of 07.5 per cent, to 40.0 per cent. The statistics lend weight to the inference that the 1 Fitz and Joslin: Diabetes Mellitus at the Massachusetts General Hospital, 1824-1898, Jour. Am. Mcd. Assn., 189S, cxxxix, p. 70. 3 0-1 . 28 1-2 . 13 2-3 6 3-4 6 4-5 2 5- 6 6- 7 . 1 7-8 2 9-10 1 10-11 1 15-16 1 16-17 2 20-21 1 G9 Unknown because of lack of past history . . 10 34 DIAHKTKS MKLLiri'S treatment of the disease as 51 whole in the community had improved, but that the intensive treatment of such cases as came to the- hospital remained stationary. TAHLE 17. THE DURATION OF LIKE OK TWENTY-SEVEN CASES OF DIABETES WHO DIED AT THE MASSACHUSETTS (!ENERAL HOSPITAL BETWEEN NOVEMBER, 1913, AND JANUARY, 1917. Duration, years. Number of cases. Per cent. 0- 1 10 47.0 1-2 1 4.8 2-3 2 9.5 4- 5 3 14.3 5- ... 2 9.5 0-7 1 4.8 10-11 1 4.8 15-10 1 4.8 Unknown' because of lack of past history . . Total 27 During this period 243 cases of diabetes were treated at the hospital and the number of deaths was 27, or 11 }>cr rent. Such a reduction in mortality is gratifying to all, speaks, well for the care the patients received in the hospital, and is reliable evidence of the improvement in treatment of the disease. (/>) City of Boston. The duration of life of the fatal cases of dia betes occurring in Boston for the period 1895-1913 and for 1915 are given in the subjoined table. TABLE 18. DURATION OF LIFE OF 1057 FATAL CASES OF DIABETES IN BOSTON DURING 1895-1913 AND IN 1915. 1S05- 1013. i 1015. Number of Number of Per cent. eases. Per cent. 32 . 7 32 27 . 3 21.9 10 13.0 10.8 18 15.3 5.9 5 4.2 8 0.8 14 11.9 2 1.7 3.2 3 2.5 . 8 3 2 . 5 .3 1 0.8 .4 4 3.4 0.8 1 0.8 0.4 3 2.5 1.8 2 1.7 .4 2 1.7 0.3 1 O.s 0.8 2 l.C 20 51 4 195 ton Medical ami Surgical .Journal, I'.llii, clxxv, p 54. Duration, years. cases. 0- 1 309 1- 2 . . . . . 207 2- 3 102 3- 4 50 4- 5 71 5- C) . . . . . 20 0- 7 . . . . . 24 7- 8 30 8- 9 . . . . . 8 9-10 . . . . . 59 10-11 . . . . . 4 11 12 . . 8 12 13 . . . . . 4 14 15 . . . . . 17 15-10 . . . . . 4 17-lsO . . . . . 3 20-3 . . 8 940 Duration loi i(i, but uncertain . . . 207 I'liknown . . 507 Diabetic jianjz :n-ne . 50 Total . . 1704 1 Morrison: HoMoii Medica TREATMENT OF DIABETES MELL1TUS 35 Here again a decrease in the mortality during the first year of the disease is manifest in the later year. (r) Naunyn. In a separate group should be placed the statistics giving the duration of life of 00 of Xannyn's severe cases. These represent a special type and should be kept distinct from the tables which precede and follow. TABLE 19. THE DURATION- OF LIFE IN '"PURE" AND '"ORGANIC" DIABETES (XAUNYN, 1906). 1 Duration, Number of cases. Number of cases. years. "Pun 1 ." Per cent. "Organic." Per cent. 0-1 22 33.3 20 28.2 1-2 19 28.8 15 21.1 2-3 16 24.2 7 9.8 3-4 6 9.1 7 9.8 4-5 . . 5 7.0 5-6 1 1.5 5 7.0 6-8 2 3.0 4 5.6 8-10 .. 1 1.4 over 10 7 9.8 In Xaunyn's statistics, however, the duration is based upon the period at which the patient left his observation, but he states that none of the cases could have lived many months after this time. (r/) Von Noorden. Von Noorden, 2 writing in 1912, says that dia- betes in children under ten years of age seldom lasts more than for one and a half or two years; in the second decade it seldom exceeds two to four years; in the third decade four to six years, occasionally even ten years, and sometimes longer. People who acquire the disease after the thirtieth year have a considerably greater length of life, and with favorable conditions the disease may last from ten to fifteen years, or even longer. The first half of the fifth decade, however, is a more critical period, but when this corner is passed the tendency of the disease is to remain mild provided doctor and patient have not themselves to blame. Such patients may live fifteen, twenty or even thirty years. Still, a length of life of more than twenty years belongs to the exception. ( Living Cases of Diabetes with Actual and Percentage Mortality, The duration of life for the living cases both up to December 1, 1915 and to December 1, 1916 is shown in Tables 23 and 24. The large accession of new cases raised the per- centage with a duration under one year, but the general tendency of the table is to show an increasing number of cases who have had the disease over a longer period than was indicated in the statistics of one year ago. Thus, one year ago the percentage over ten years was 17.9, and the later series shows 20.5 per cent. (4) Comparison of Duration of Life by Decade of Onset of 500 Fatal and 640 Living Cases of Diabetes. Tables 22 and 23 show the actual duration of life in years of both fatal and living cases arranged for comparison by decades, and as before comprise the two series of cases ending December 1, 1915 and December 1, 1910. 38 DIABETES MELLITUS - i | CO ,_-' TREATMENT OF DIABETES MELLITUS 39 M 'C t- t ?l >n O* C: 40 DIABETES MELLITUfl TABLE 24. THE DURATION OF LIFE OF Six HUNDRED AND FORTY LIVING CASES OF DIABETES. Duration, years. 0- 1 Number of . . 29 1- 2 . . . 2- 3 . . 66 61 3- 4 . . . 4- 5 . . 47 46 .">- 6 . . . 6- 7 ... 7- S . . . 8- 9 . . . 9-10 . . . 10-11 . . . 11-12 . . . 12-13 . . 43 . . 43 . . 27 . . 24 . . 15 . . 26 . . 10 12 13-14 . . . 14-15 . . 8 4 15-16 . . . 16-17 . . . 17-18 . . . is-in . . . 19-20 . . . 20-21 and over 7 . . 2 . . 4 7 . . 4 . . 5 1S9.3 101.' 1893-1910. 5.9 13.5 12.4 9.6 9.4 8.8 8.8 5 . 5 4.9 3.1 5 . 3 2.0 2.4 1.6 0.8 1.4 0.4 0.8 1.4 0.8 1.0 cr of cases. Per cent 58 9.1 56 8.8 61 9.5 66 10.3 72 11.3 60 9.4 42 6.6 34 5 . 3 36 5 . (') 27 4.2 31 4.8 23 3.6 17 2.7 10 1.6 8 1.3 8 1.3 4 0.6 5 0.8 8 1 .3 5 0.8 9 1.6 408 640 TABLE 25. COMPARISON OF DURATION OF LIFE OF FATAL AND LIVING CASES OF DIABETES BY DECADE OF ONSET. DECEMBER 1, 1916. Age at onset, years. 1893-10 Hi. Number of cases. Average duration, years. Fatal. Living. 10 11 20 21 :',() 31 10 41 50 51 no ni-7o 71 so si (to 33 S.I 1.x 37 7 4 ''2 3.4 1.22' 1 . 33 2 4.44 3 2 . 59 4s 11 .X 27 5 . 5 55 1 1 () 40 ( 1 . 2 2.70 2 . 75 2.70 2.02 40 O.S 50 10.2 51 10 2 01 0.5 3 .30 3. O.S 4.00 5 . 32 53 13.0 71 14.5 (it 12 s 94 14.7 4.43 4.42 6.12 7.1(1 71 17.4 140 21). Xl 95 1!) 1X4 28.7 . OS (1.42 7.04 7.28 07 23. S 120 24.5 1 1 5 23 1(13 25 . 5 n.n3 (1.02 0.20 n . 20 ~ > -) 12.7 55 11.2 (17 13 4 (13 O.S 0.00 5 . Of) 5 . 3X . 00 \ 1 3.4 1 1 > ) in 2 12 1 .0 3.71 3.5S 4.45 5 .25 1 0.2 1 0.2 0.33 1.33 Inspection of Tahle 2o slum's that the duration of life of the living cases has already exceeded that of the fatal cases in five 1 If C;iso \o. ss7. whose duration was twenty-nine years, is ineluded, the average duration of the ca>es would lie 2.0(1 years. Sec also Table 22. - Xot including ('ase \'o. ss.7. n See Talile 2-'.. Note there were 2 ca.-es in the fir.-t decade under 1 year of age, one year ago, but now there are > - 1 ::::::: ^H in **. H ra tn . ~- / < S ^ cc h w W '' ai ^ 2 < " >. P ^ >> W /. '"' [2 S t " " >. O P o * - >. (z; I.-5 M w >. S Z i H . ~' X Q S! 2 1C Q - i g jj w >> '*. t m o Q H , . T " H . . . . "* ' H <; x - - - 02 o EH ci >, . cs X CM b -- 1 >, ^ Ti * E co ' ' |25 rt >> K >1 71 H K C f* tH ^ >. . . . .01 . T-H . CO 5 o '!' TI r ::::--:: fl ^ U 3d o rt >> : : : :-H : : : r- 1C o C-. >. ::::-..: -> ^ Q 05 .- .0 2 ,_, ,_, >o Rd . -< 3 a x "S M 2 * 3 "" ~> < s K ~ s "^ x ' '"!-* ' M t- -^ " K a ff, ", K cc o 3 - >, 01 2 - 1 >, : : : : SN : * ; 01 ? DIABETES IN MALES AR DECEM ^ >> ^ ti V. 2 ; \ '. '. ' n '. . : : : : : : :- :--oi : : : : i-i-i-- 1 : : : :-H(M"s : : : : :^- r '" : : : : -t-.'o- : . ~i - * ^ 7i h ^ o -f rt CO X > IM O '^ J " CO s OF DIABETES IN FEMALI Dt ' >, ^^ x o| I s - r^ o ^ ^1 10 ;- : : : ^ -H -M ^H : '. ! ! ! r-< IN "1* -H ! ~- ' : TO -f -M ; ; ; ; M -H ci i^ ; J 71 t- X X a ^ Ui"^ ; I-H I-H IM ! '.<-*'-< rl <-> TO "H _ -f . . -H t~ -r cr. c^ i.o oi ic /: -rfi -? t~ -t 1 : Cl 71 ~ 1- -^ f Cl T-H CO s 01 IP q oi 01 iO o t~ X M~~ 6 O-l O) ii u - = '--'-'---< t^ ._ J P) T3 ^^ W J a '- h "--'--*--' ^! ^ i -H Ol CO T" 10 O 1^ CXD 1 I ~o H CJ O a ~ = - 01 ro Tt" >c o t^- HEREDITY 43 The apparent clinical preponderance of diabetes in men may be ascribed first to the greater consideration which they receive, second to the large percentage discovered by life insurance examina- tions, amounting to 11 per cent, of my male cases, third to the hesitancy of women to speak about their urinary troubles and their neglect to regard such seriously, and fourth, to the type of practice of those who have reported large series of cases. It is suggestive that in Japan 80 per cent, of the cases of diabetes reported by Iwai 1 were males, that 70 per cent, of the German cases reported are males, but in England and the United States only (50 per cent. (See Table 2(.) Morrison's 2 figures for the city of Boston between the years 1895 and 1913, like the national statistics, show of) per cent, of 1775 deaths from diabetes to have been females. The fact that more men than women receive special medical treatment for diabetes and that more women than men die of the disease is evidence in favor of the usefulness of our present thera- peutic procedures. In Tables 28 and 29 the duration of the disease is given for males and females separately. These tables show clearly the longer duration of diabetes in males. E. HEREDITY. The influence of heredity in diabetes has always interested me, and I have taken great pains to secure data from my patients upon this point. Xaunyn said that the more carefully he inquired into the family history, the more commonly he found heredity to be present, and I am quite sure that this rule holds. It is easy to exaggerate the importance of heredity, for diabetic patients naturally would be more liable to know of the presence of diabetes in the members of their families than would patients who did not have diabetes. This introduces a considerable error into the statistics of the heredity of diabetes. Among 100 non-diabetic patients, Heiburg found 7 who had relatives with the disease, and upon inquiry from 100 diabetic patients he learned that 18 had relatives similarly affected. Among 500 consecutive histories of my own non-diabetic patients 25 had relatives with the disease. Having seen the increase in diabetes in the community to be at least in great part apparent rather than real, I hesitate to attach too much importance to heredity. Undoubtedly, heredity will appear more prominently in the case reports of diabetes as time goes on, but this may only reflect the greater accuracy in vital statistics. 1 Iwai: Loo. cit., p. 24. 2 Morrison: Loe. pit., p. 34. 44 1) I A HE T JUS MELLITUf! Statistics ii])on heredity in diabetes are susceptible to criticism because of what is included under the term "hereditary." I sus- pect that often by this term familial as well as hereditary cases are included. So far as my own statistics are concerned "heredity" signifies the presence of the disease in a parent, and occasionally information about a grandparent was obtained, as well as uncle, aunt or child. The term "familial," on the other hand, embraces brothers, sisters, and cousins and it is quite evident from the litera- ture that brothers, sisters and cousins are usually included in com- piling statistics on heredity. Conjugal diabetes, of course, does not belong to either group. In securing statistics upon heredity, a record should always be made upon the history blank as to whether the information appears reliable. In 4 cases of my series, Cases Xos. 105, 473, 50)] and 954, the diabetes developed in the child or children before it occurred in the parent. The age of onset and the duration of the disease in these four cases to death or the present time is shown in Table 30. TABLE ;>(). DIAHKTKS i\ THE CHILD AND PARENT COMPARED. Child. Parents. Case AKC at Duration, Case Afie at Duration, Xo. onset, years. Condition. Xo. onset. years. Condition IT) 2.S Dead :'.) 47 5.7 Dead 14 1.4 Dead 47:5 4S 0.2 Dead Ml I 5.1 Alivo 711 54 8 . X Alive 2 1.0 Dead 12 0.2 Dead 9 ' 54 49 1 ' Allvo A history of diabetes in grandparent, parent, uncle, aunt or child was obtained in 151 cases of my series in other words, 13 per cent. Of this number, S4 were males and (57 females. On the other hand, the group of cases showing a familial tendency included 100 cases, or ( S per cent, of my series. Of this group, 5!) were males, and 41 females. If we combine the two groups, we have 251 cases showing an hereditary or familial tendency, or 21 per cent, of all the cases. Older writers record data which are similar. The fact that the series of cases of several of these writers showed a higher percentage of hereditary cases is interesting. Personally, I believe that the statistics on the subject should be gathered with increasing care, and that as time goes on the percentage of hereditary cases will constantly grow. For comparison, Table 33, taken in part from von Xoorden, is inserted: HEREDITY 45 IO t~- O CO X r- 1C Ol H - CO -H M< X O X I "f -, c o o o 9 c c 46 DIABETES MELLITUS 1. Earlier Statistics. TABU: 33. Heredity Author. No. cases. per rent. Crube 8.0 FrericJis ...... 10.0 Seegcn 14.0 Srhmitz 20.0 Kill/ C92 21. (> Williamson 100 13.0 Bouchard 25.0 Naunyn 39X 17.0 A . ^ f Hereditary. 18.51 - , VonXoorclen i Familial, <.9J 2;) ' 4 ,,,-. j Hereditary, 13.01 1( ' slm 11S ' (Familial, X.O/ AVhen I personally examined the records of the Massachusetts General Hospital from its formation until 1SOS, and in connection with Dr. Fit/ published the same, I found that in 42 cases the question of an inherited tendency to the disease was raised, and this factor was .stated to have been present in ID cases and absent '2. Notable Diabetic Families Showing Heredity. The most interesting diabetic family history that I have encountered was furnished me by a nurse, who stated that her mother and her mother's eleven brothers and sisters all died of diabetes save one, who, like the rest, has the disease but is still alive. Two of her aunts each had a child with diabetes. To my repeated inquiries about the accuracy of the information, she writes: "1 am positive that diabetes was proven by examination of the urine in the case of five of the first generation; the others are on record as having died of diabetes. All the patients were stout." (See Table 34.) TABLE 34. Preceding gone-ration. Date of birth. DateofdoatI Duration year.- of life, Ouii-e of death. Andrew . I lannali his wile . Oct. IX, 1X03 Aug. 9, 1X05 1X53 1848 50 43 Lockjaw Childbirth First generation: Kmilv Feb., 1X27 1XX1 54 Diabetes Daniel . . . . Aug., 1X2X 1X13 15 " Hannah . . Mar., 1X30 1905 / 5 " (leorge . Elizabeth . . . Aug., 1X31 . Jan, 1X33 1890 1XX3 59 50 " Andrew . . Mav. 1X34 1X74 40 i .lane .... . Feb., 1X3(1 189(5 (i() ' Addie . . . . Dec., 1X37 1904 l>7 ' Karhel . . . . Oct., 1X41 1X73 32 ' William . . . Dec., 1X13, 1X95 52 ' John Julv, 1X15 1X7X 3,3 n Catherine Mar., 1X48 Living 52 I o date. December, 1915. Second general ion : .Jennie, daughter < f Fli/.abeth . . . IX Jane, daughter of Family 42 HEREDITY 47 The family showing the most marked diabetic heredity with which I have personally come in contact was that of a Jewish patient, Case No. 759, who developed diabetes at the age of fifty- five years. A detailed history of this case is reported on page 489. A brother died of diabetes and tuberculosis at forty-eight years; a sister of diabetic coma at fifty-five years; a sister of acute indi- gestion at fifty-eight years, having had diabetes for many years; a sister of Bright's disease at fifty-six years, having had diabetes for ten years; a sister, who is living, has had acute indigestion and also has diabetes; finally a brother died of influenza at 62 years, having had diabetes eight years. The father succumbed to paralysis at sixty-two years, and the mother to heart disease at sixty-five years. A niece, Case Xo. 430, also has had diabetes for nine years, during six of which she has been under my observation. She became pregnant in 1915, and I am glad to record that the carbohydrate balance, which amounted to 15 grams for a year or more before pregnancy began, rose to about 65 grams. She had a normal delivery in April, 1910, nursed her baby for several months, and both mother and child are in excellent condition (May, 1917). The tolerance for carbohydrate fell after delivery and again when nurs- ing was given up. Xo sudden change in tolerance, however, was observed. 3. The Favorable Influence of Heredity. The influence of heredity upon the disease is not uniform. It may be serious, or it may be favorable. Xaunyn has pointed out that the onset of the disease in successive generations is apt to take place at a successively earlier age, and consequently the diabetes is more apt to be severe. Von Xoorden has emphasized the importance of the character of the diabetes in the progenitor as indicative of the character of the disease in the offspring. My cases of hereditary diabetes have frequently been mild, and this has so commonly been the case that I always look upon heredity as a favorable omen. Several cases in young people have been unusually favorable, and these are summarized in the following Table 35, which should be compared with Table 25, p. 40. TABLE 35. THE FAVORABLE INFLUENCE OF HEREDITY UPON THE COURSE OF DIABETES. Case Age at Duration. XD. onset. Hereditary history. 203 ... 7 Mother Alive 10 1 295 ... 14 Father, also brother Coma 12 304 . . . .13 Grandfather, also 2 Coma 9 2 brothers 310 ... 17 Mother, also brother Coma 21 and sister 48 DIABETES MELLITUS All cases which have come to my attention of youthful patients with diabetes living for such long periods of time have been heredi- tary. The case of Sehmitz quoted by Xaunyn 1 was hereditary, and deserves regard because of its encouraging features. This was a four-year-old child, whose mother and older sister were diabetics. The urine was frequently examined for sugar, and always found sugar-free, the last examination being November 22, 1ecember 3, 3.5 per cent.; December 7. PEKCE\TA<;K OF JEWISH CASES FOUND i\ SUCCEEDING; CROUPS OF 100 PATIEXTS. 4th f>th fith 7th 8th 9th Kith 100. 100. 100. 10(1. 100. 100. 100. ,5 3 446 r> 17 11 13 16 8 10 It shows little more than the increase in the Jewish population in the community during the period of observation. What is, however, of much more interest than the percentage of Jewish patients to the total number is the course which the disease has taken among these individuals. This is shown in Table 38. TABLE 3S. DURATION" OF LIFE OF THIRTY FATAL CASKS OF DIABETES IN JEWISH RACE. Df- I'mlcr 1 radcs. 1 yr. I yr. 0- 10 11-20 31-40 41- .")() 61-70 71-KJ Total 3 1 2 1 1 1 1 3 2 . . 1 1 . . . . 1 1 . . 4 1 1 1 8 1 1 1 Wallach: Deutsch. mod. Wrhnsrhr., 1*93, xix, p. 779. 2 Morrison: LOC. cit., p. 34. THE CURABILITY OF DIABETES 51 For the sake of comparison, the statistics for the Jewish patients have also been arranged in a table showing onset by decades, and as in former instances, the duration of life of the fatal cases up to December 1 , 191(1, is given. It is surprising to me that the course of the disease among my Jewish patients was less favorable than the general average. The reason for this is not clear, but I believe it to be due to the fact that they understood the diet less well, and so have paid the penalty of poor cooperation. The causes of death of the 30 fatal Jewish cases were as follows: Coma 1<> Tuberculosis 2 Cardiorenal and vascular 8 Tonsillitis 1 Cancer of kidney 1 Suicide 1 Unknown 1 30 F. THE CURABILITY OF DIABETES. The course of diabetes occasionally ceases during the lifetime of an individual, but this happens very seldom. Both Xaunyn and von Xoorden believe that a diabetic patient may recover. I am unwilling to state that any of my patients have been cured, though I am watching with interest several cases of the 40 now alive who have had the disease for fifteen or more years. (See p. 408.) Other facts regarding my own cases are set forth in Tables 23 and 24. The term "arrested" instead of "cured" has found general acceptance in the literature of tuberculosis, and it is equally appro- priate in diabetes. It is better for the present to take the con- servative stand-point and be very slow to report cured or even arrested cases. This has not been done in the past, and I am con- stantly being told of patients who have been "cured," of which proof does not exist. It should be remembered that a sugar-free urine and a normal blood sugar do not constitute a cure; both should be normal for years. Allen has admirably classified the subject, and pointed out that recovery is possible in three types of cases: 1. Acute Diabetes Arising from Curable Causes. It is quite possible that diabetes which has followed trauma might subside, either with or without operation. Recognizing the hyperglycemia which occurs in the course of infectious diseases, it is easy to under- stand that in a few cases glycosuria may develop, last several days, and after recovery from the infectious disease may disappear. Schmitz's remarkable case, already described on page 49.) 2. Exceptional Cases in Childhood. A considerable number of cases of mild or temporary diabetes in children is beginning to appear in the literature. (See Kiesman, already cited, p. 4N.) Case No. 203 of my series is instructive. This little boy first showed sugar at the age of seven, April o, 190N, although symptoms were present the preceding year. His mother has diabetes and is Case No. 15"). He was promptly put upon a careful diet by his mother and physician, Dr. Warren \Vhite, of Roxbury, Mass., and has remained sugar-free 1 during the subsequent nine years. lie was very faithful to his diet and remains in good condition. By advice he is purposely keeping a trifle underweight and eats a normal diet except for a little less sugar. The faithfulness of his mother to her diet, maketh a doctor's heart glad. Though she has had diabetes sixteen years she "would not touch sugar any more than she would poison." Yon Xoordeu reports a striking case of a seven-year-old boy who, on a strict diet, constantly excreted 20 to HO grams of sugar and considerable quantities of acetone bodies. lie became sugar- free only with the help of oatmeal and vegetable days, lie remained on this restricted diet for some years and at the age of twelve years was again seen by von NOorden, when he was in perfect health, eating an ordinary diet without a trace of glycosuria. H. Diabetes Associated with Organic Diseases. This group is subdivided bv Allen into (1) those cases with nervous disorders, ETIOLOGY OF DIABETES 53 such as the cases described by Naunyn in association with tabes. (2) Cancer involving the pancreas which may be followed by the disappearance of sugar. Xo instance of this kind has occurred in my practice. (3) Xeither have I observed cirrhosis of the liver replace the diabetes, such as Claude Bernard and Lepine describe. (4) On the other hand, with the development of severe nephritis, diabetes has ceased in a few cases, notably Case Xo. 354 (see page 29) . In Cases Xos. 457 and 872 (p. 418) it developed. (5) The fifth group mentioned by Allen is tuberculosis. Xaunyn and many other writers have observed the subsidence and disappearance of the disease following tuberculous infection. I would refer to Case X'o. 344 of my series (see page 410). In this patient the severest degree of diabetes, as shown by the urinary analyses of December 25-20, 1911, changed its character entirely, as proven by the urinary reports of March 17-18, 1912. During the interval pulmonary tuberculosis broke out and advanced rapidly. AYeight decreased from 147 pounds to the neighborhood of 80 pounds and the patient died without acidosis. Many diabetics are cachectic at the time of death, but this condition by no means has modified the disease. On the other hand, when cachexia, due to other causes than diabetes, occurs in a diabetic, it may be responsible for improvement in the diabetes. G. ETIOLOGY OF DIABETES. In any discussions of the etiology of diabetes it is recognized that some fundamental disturbance in the metabolism is at fault, presumably associated with the pancreas. This is the cause of diabetes. In the discussion which follows, various factors pre- disposing to the development of diabetes are considered. They are: (1) obesity, (2) dietary excesses, (3) multiple etiology, (4) heredity, (5) strenuous life, (()) nervous element, (7) infections, (8) arterio- sclerosis, (9) syphilis, (10) trauma, (11) pancreas, (12) hypophysis, (13) liver, (14) renal glycosuria, (15) gout. 1. Obesity. In over 40 per cent, of 1063 of my own cases avail- able for statistics in this respect, marked obesity has preceded the outbreak of diabetes, and if my early data were more exact I believe that the percentage would be fully twice as great. Contrariwise, in the presence of wasting disease diabetes is practically unknown. One of the most striking instances is that of a little child, Case Xo. 1139, a healthy baby until his second year, when he entered upon a series of rhinopharyngeal and bronchial attacks, including a mild bronchopneumonia. During eleven of these feverish ill- nesses, exercise was curtailed and he was kept abed to safeguard his heart. A cow was bought that he might have fresh, rich milk, and to the 50 ounces of this milk which he daily consumed were added 54 DIABETES MELLITUS three tablespoonfuls of a malt preparation, amounting in*nutritive value to 120 or more calories. At the end of his first year lie was of normal weight, but at eighteen months his weight was above that of a, child of two years. When two years old his weight had increased to that of ii child of three years and a half, and when two years and a half old, his weight was above that of a child of four years. Is it surprising that when mumps developed, glycosuria appeared? It vanished with a change in diet, to return threemonths later, during the presence of another mild febrile disorder, and again in six months after the removal of adenoids, but si nee the urine has remained sugar-free. Kisch 1 directed attention to the frequency of the development of diabetes in the later lives of fat children, and 1 never see a fat child without thinking of this possibility, for a fat child, like a fat man, is prone to diabetes. Particularly serious must it be for a child or adult of slight stature to put on weight which would be excessive even for a man of large frame. Case Xo. 1142 first consulted me at the age of forty-one years, and 1 can just remember his little spindling legs and delicate frame as a boy before he entered the primary school; many years later lie entered a grocery store, "worked like the dickens and ate the .same way," until, at the age of forty lie was 00 pounds above the average weight, and sugar was found at a life insurance examination. Of oOS applications for life insurance rejected for obesity Romanelli 2 reports 21 per cent, as having glycosuria. A poor musculature usually accompanies obesity. One-half of the carbohydrate in the body is stored in the muscles and a large part of the sugar of the body is burned in them. Therefore it does not appear strange that fat people develop diabetes. It would almost seem as if the muscles became so loaded with fat that they were unable to burn carbohydrate. ^ It is a serious matter to get fat. One of my thin relatives whose religion was not of the eut-and-dried type used to laughingly say to her fat but pious sisters-in-law, "To get fat shows a lack of moral character." And my learned friend who supplies me with epigrams says, "Laugh and grow fat grow thin and laugh longer," and also, "What is added to the waist line is cut from the life line." i\Iy own cases have been analyzed in two ways to show the incidence of obesity prior to the development of diabetes. Of the lOni) cases available for the purpose above the age of twenty years, I considered from the appearance and history that obesity was or had been present as an etiological factor in 4 1 per cent. It is possible, however, to arrive at a more definite conclusion from the examina- tion of a table constructed to show the average highest weights at given heights of a large number of male and female diabetics. Such 1 Ki-ch:.Inur. Am. Mcd. .\ssii., March. I'.H.V Ixiv, p. Kl.'is. 2 Romaiu'lli: .lour. Am. Mcd. Assn., I',)!,"}, Ixv, p. HI.". TABLE 39.- -A COMPARISON OP THE AVERAGE HEIGHT AND WEIGHT OF NORMAL AND DIABETIC INDIVIDUALS. Weight. IIei 8 ht - Normal.' Diabetics (K. P. .1.). (Shepherd). Male. Female Ft. In. Cm. Lbs. Kg. No. of cases. Lbs. Kg. No. of eases. Lbs. Kg. Ages 15 to 24 4 10 147.3 4 11 149.9 5 152.4 120 54.5 1 150 70 . 8 2 105 47 7 5 1 154 9 122 55 4 5 2 157.5 124 55 . 8 3 110 49.9 5 3 100.0 127 57 . 2 117 53 . 1 o 4 102.0 131 59 . 5 o 105 47.7 5 5 105.1 134 00 . 8 1 157 71.3 5 107.7 138 02.0 3 139 03 . 1 1 119 54.0 5 7 170.2 142 04 . 4 2 141 04.0 2 147 ; 00.7 5 8 172.7 140 06.3 8 148 67.2 5 9 175.3 150 68. 1 4 182 82.6 5 10 177.8 154 09 . 9 i 1 101 73.1 5 11 180.3 159 72.2 1 136 01.7 (i 182.9 105 74.9 3 191 80 '. 7 (i 1 185.4 170 77.1 1 148 07.2 2 188.0 170 79.9 1 200 90 . 7 (i 3 190.5 181 82.1 Averag e number of poum s overwei glit of 38 diabetics 3 pou lids. Ages 25 to 29 4 11 149.9 1 120 54 . 5 5 152.4 125 56.7 5 1 154.9 120 57.2 5 2 157.5 128 58 . 1 58 . 1 5 3 100.0 131 59 . 5 1 128 5 4 102.0 135 61.3 1 248 74.9 5 5 105.1 138 02.0 , 1 210 1 105 75.3 5 6 107.7 142 04 . 4 4 101 73.1 2 166 66 . 3 5 7 170.2 147 06.7 3 158 71.7 1 146 5 8 172.7 151 08.5 3 170 79.9 5 9 175.3 155 70.3 2 173 78.5 o 10 177.8 159 72.2 2 141 04.0 5 i 11 180.3 104 74.4 4 100 72.0 (i 182.9 170 77 . 1 1 208 6 1 185.4 177 80 . 3 1 200 90 '. 7 2 188. 184 83.5 3 190.5 190 86 . 2 Average number of pounds overweight of 27 diabetics 51 pounds. 1 Ages 30 to 39 4 11 149.9 1 125 50.7 5 152.4 129 58 . 5 5 ! 1 154.9 130 59 . 3 140 03.5 o 2 157.;' 132 59 . 9 2 123 55.8 5 3 100.0 13-5 61.3 4 103 74.0 5 4 162. i', i 139 63.1 2 140 (Hi . 3 1 250 5 5 105. 1 142 64.4 3 171 77.0 4 109 70.7 5 (i 1(17.7 140 66.3 3 177 80.3 3 191 , 80.7 5 7 170.2 151 68.5 7 179 81.2 1 178 80.8 5 8 172.7 155 70.3 7 175 79.4 1 145 05 . 8 5 9 175.3 160 72.8 6 172 78.1 5 10 177 . 8 105 74.9 7 131 82.1 1 225 5 11 180.3 171 77.6 10 193 87 . 4 212 182.9 177 80.3 2 181 82.1 6 1 185.4 183 83.0 1 135 88.5 2 188.0 190 86.2 6 3 190 . 5 197 89 . 4 Averag number of pounds overweight of 72 1 diabetie s^23 po uncls. Ages 40 and over 4 10 147.3 4 179 81.2 4 11 149.9 5 181 82.1 5 152.4 133 00 . 4 2 1S8 85 . 3 16 172 78.1 5 1 152.9 135 61.3 1 174 79 . 11 175 79.4 5 2 157.5 13S 62.6 ..' 14 109 70.7 5 3 100.0 141 (i t . 14 180 81 .7 5 4 102.0 144 65.4 4 187 84 '9 14 171 77.0 5 5 105.1 148 07.2 12 178 80.8 17 209 94 . 8 5 6 107.7 152 69.0 20 183 83.1 1 1 190 80.2 5 7 170.2 150 70.8 24 204 92 . 6 3 199 90.3 5 8 172.7 101 73.1 39 200 90.7 4 207 93.9 5 9 175.3 166 75.3 42 211 1 222 5 10 177.8 171 77.0 18 218 5 11 180.3 177 80.3 . 20 212 3 230 6 182.9 183 83.0 17 235 6 i 185.4 190 86 . 2 2 337 6 2 188.0 190 88.9 4 225 6 3 190.5 201 91.2 1 270 Average number of pound ,s overweight of 320 diabetics 37 pounds. 1 Shepherd: Med. Examiner, 1899, ix, p. 209. Dl A BE TES MELLIT US a table has been prepared, and for eomparison Shepherd's Table of Height and Weight of Healthy Adults is placed alongside (Table 30). I have not a very satisfactory collection of heights and weights of diabetic children, but the question is so important that I append the standard height and weight table of healthy children selected by Dr. John L. Morse for his ( 1 aw Histories in Pediatrics. (Pub- lished by \Y. M. Leonard, Boston, 1913.) TAHLK 40. HKICHTS AND WEIGHTS OF CHILDKRX. Height. Weight. Age. Boys. Girls. Boys Girls. Inches. Cm. Inrhes. Cm. Pounds. Kg. Pounds. Kg. Birth 1 20.0 52.") 20.5 52.2 7 . 55 3 . 43 7.10 3 . 20 1 year 29 . 73 . X 2S.7 73 . 2 21.0 9 . 54 20.5 9.31 2 years 32 . r> ,s2 . s 32 . 5 S2 . s 27.0 12.27 20.0 11.81 3 vcars 35 . o 89 . i 35 . 89 . 1 32 . 14.55 31.0 14.09 4 yoars 38 .0 90 . 7 3S . 96.7 30.0 10.30 35 . 15.90 The heights and weights in this table are net; i. f., without shoes or clothes. A>_'<> at last birthday. 5 years 41.7 105.9 41.3 104.9 41.0 18.6 39 . 18.0 years 43 .9 111.5 43.3 110.1 45 . 2 20 . 5 43.4 19.7 7 years 40.0 HO.s 45 . 7 116.0 49.5 22.5 47.7 21.7 8 years 48.8 123.9 47.7 121.1 54 . 5 24.7 52 . 5 23.8 9 years 50.0 127.0 49 . 7 120.2 59 . 27.0 57.4 20 . 10 years 51.9 131. S 51.7 131 .3 05 . 4 29 . 5 62.9 2S . 5 1 1 years 53.0 130.1 53 . 8 130.0 70.7 32 . 1 09 . 5 31.5 12 years 55.4 140.7 50 . 1 142.4 70.9 34 .9 7S . 7 35.7 13 years 57.5 140.0 5s . 5 14S.5 84.8 3S . 5 8S . 7 10.3 14 years 00.0 152.4 00 . 4 153.4 95.2 43.2 9. Those make up for boys approximately S per cent., and fo girls 7 per cent., of the gross weight. The term, ''ag at last birthday," is liable to give a wrong impression, because the figures given are vally average figures taken from all the children from that birthday to the next. A lore accurate term is the succeeding half-year; age approximately for succeeding half-year, ;'. c.. five and a half years instead of live years, the age at the last birthday. A study of the tables plainly shows that the average diabetic, whether nude or female, has been greatly overweight. In only four of the forty-three groups is there an exception to this rule. With these facts before him, what an opportunity for the general prac- titioner to practise preventive medicine. The question of overweight is of importance in a far broader field than diabetes, and for that reason I insert from the Traux- actionx <>/ flu 1 Actuarial Socicfi/ of America the mortality figures of a large series of individuals according to their weight and height. This table shows convincingly the danger of overweight. 1 The data for the first lour years, taken by Morse from Holt's "Diseases of Infancy and Childhood," are here given as published by Holt in his 7th edition. ETIOLOGY OF DIABETES C 5 s i -= g! 1 < X o W c I S j i S S < -S - x x o ooo] o] o-) co i .0 LOO CCO) OOt^ LOOO O) O) O O O O) O C C OOO XXX H) C) o O5 o Ci XXX CMOO 0) M T X X X O XXX -O C LO a CO t CM CO coxo I CO LO o co o a -O O) CO IO CO- 3S2-S ris o o t- 01 01 o) -* co c LO LO LO IO Tf -f Tf CO CO CO CO CO 01 OO ', 0>) ' ' O CO CO CO TO Ol O) CO O) CM x x oxx en 0) t c. >.i CO CO -f COO1O) CM CM CM CM 00 -t LO X b- XOO CO -f X O X t~ O i OO CO CO COCOTf CO * * COCO- CM CM CM CM CM OC CM O)O)CO COCOCO COCOCO O) CM CO O)O]CM CM' ~ 2 - 22- CO O O X O) O OOO OO OO CC X X O C O ooo o o c o o c c o c o o c o o LO -f-p-i r^xx oxx r^ t- -*^i COO 000 CCO COO 00 Cc ooc C-)CMO 1 0)0)0) -f M O) X O) O) o o ooc c o : OOO OOO CO OO >o to c ~ ~ x xxx cox ox mo 01 o o oc ooo ooo OOO OOO OOO XOO O O X OXX C. XX OXX C C. O O XOO XXX XXX OXX OXX OXX < CM O c; c tccox cox c >o o -r co O) . OOO OOO OOO XXX XXX XXX OXX OXX CO! CO O) X t^ Ol CO o o xxx xxx b x x b x x : 01 o c ooo C. OO XXX XXX OXX OXX co o o") c ooc o o b iO rf O CM O * M OC O LO t^ LO O O 01 co 01 01 o o c c ooc -tox cooio xoco -tcoi t-co^t OOO XXX XXX OXX OXX OOC OOO O O X OXX OXX OXX 5S DIAHKTEfi MELLITUS Table 41 is at first confusing. Tt shows the percentage of mor- tality in the community for different ages and heights according to whether the individual is below or above the average standard weight for the given height. 1 Tims, the liability of death of an indi- vidual forty years of age, under five feet seven indies tall, who is of standard weight (see Shepherd's Table, p. .">.">) is only !Ki per cent, on the basis of the mortality for all of this age. If he is fortunate enough to be 10 or even 1") pounds under weight, his liability for death is reduced to SO per cent., but if, on the contrary, he weighs 15 pounds above the standard, his misfortune raises his liability to death to 104 per cent., and if ">() pounds over weight, to 1.">1 per cent. The patient in my office today, as I write this, saw that with his weight of oOO pounds his chance for death had more than doubled, and it was unnecessary for me to urge him to cat less. It is obvious that one's chance for life is increased after the age of thirty-five by being lo to 20 pounds underweight, and that obesity carries with it an invitation to death. TABLK 42. KXPF.CTATIOX OF LIKK AMF.KICAX TABI.K. Exp. 48.72 48. OS 12 47. If) i:{ Hi. so 11 Hi. Hi I.'. 45.50 1C. 44.s. r ) 17 14.1!) is 1.T53 19 12. S7 20 12.20 21 41.53 22 10. So 2:5 10.17 21 159.19 2:, 3s.si 2C. :5S.12 27 :57.K5 2s :5C..7:5 29 :5(i.o:5 :5t) :55.33 ::i :5i.c.:5 53 is. 79 7r> ti.27 The remark is often heard that from middle-life onward the advent of diabetes does not tend to shorten the life of the individual. To this general statement I cannot agree. It is true that the restricted diet often exerts a most beneficial influence upon not only the diabetes but also upon obesity, heart disease and an accompany- ing Hright's disease, yet on the whole as yet we cannot advise 1 patients to endeavor to contract diabetes in order to live long. 'Flic expectation of life of patients is not sufficiently considered by 1 This table is to he u-ed in conjunction with Shepherd's t:il>lr. (Sec Table .'59.) It may lie as-limed that normal women between the ;im>s of fifteen and torty- fonr \veiiili about MX p >und- le-< than men, and after that at:e about three pound- less. For a careful .-tu 49 50 16, -> 7s 5 .11 57 16 05 79 4 .74 58 15, 39 SO 4 .39 59 14 74 8] 4 .05 ()( ) 14 10 S2 3 .71 til 13 .47 S3 3 .39 C.2 12 .86 si 3 IIS c,:5 12 ,26 S5 > * i til 11 .67 80 > .47 05 11 10 s7 > .18 00 10 54 ss 1 .91 07 10 ,00 S9 1 .60 OS 9 .47 90 1 .42 69 s .97 91 1 .19 70 s IS 92 .98 71 s ,00 9.H ,80 72 7 55 94 64 7: 1 , t ,11 95 50 74 () c.s ETIOLOGY OF DIABETES 59 physicians, and for my own benefit as well as others I insert Table 42, the American Fxperience Mortality Table. A study of this table will show how many years we often fall short of the normal expec- tation of life in our present treatment of diabetes, and it sets a goal to which we must aspire with each case. 2. Dietary Excesses. Dietary excesses figure quite prominently in my records as precursors of diabetes. Allen's dogs artificially predisposed to diabetes by removal of a considerable portion of the pancreas became diabetic when overfed. It is, however, the excess of food rather than of carbohydrate which does the harm. Indeed, a high percentage of carbohydrate in the diet does not appear to predispose to diabetes. Thus, the Japanese live upon a diet con- sisting largely of rice and barley, yet so far as statistics show, the disease is not only rare but mild in that country. Four years ago Germany consumed half as much sugar as was eaten in the United States, and yet diabetes was quite as prevalent there as here. The increase in the quantity of sugar consumed per capita in the United States during the last century is very great, as shown by Table 43, and when one compares this table with the rising incidence of diabetes shown in Tables 1 -4, it would seem as if the two must stand in relation. Fortunately, the dietary habits and the statistics upon diabetes of Japan would seem to save us from this error, but I am not quite convinced, and shall be glad to study the more accurate figures for the incidence of diabetes in countries where carbohydrate food is predominant, which will come with time. It will also be of interest to learn the incidence of diabetes of workers in candy factories. The marked increase in the consumption of sugar shown in Table 43 might appear of great significance in seeking for a cause of the greater frequency of diabetes today were it not for the fact that, whereas the consumption of sugar per capita between 1900- 1917 has increased 17 per cent., the mortality from diabetes has nearly doubled. Nevertheless, such a marked alteration in the diet of a nation is noteworthy and deserves attention. TABLE 43. THE CONSUMPTION OF SUGAR IN THE UNITED STATES, 1800-1917. Population Pounds per average for Tons, capita, Years. decade. yearly average. yearly average. 1800-1810 .... 6,146,343 33,952 ' 11 1810-1820 .... 8,280,041 35,152 8 1820-1830 .... 11,038.448 47,930 9 1830-1840 .... 14,720,126 85,055 12 1840-1850 .... 19,824,542 158,995 2 1850-1860 .... 26,905,025 350,653 26 1860-1870 .... 34,645,094 391,013 23 1870-1880 .... 43,777,206 732,758 33 1880-1890 .... 55,912,152 1,233,408 44 1890-1900 .... 68,818,801 1.920,886 56 1900-1910 .... 83,275,548 2,722,048 65 1910-1917 .... 97,837,494 3,568,014 73 GO DIABETES MELLITUS 3. Multiple Etiology. The majority of diabetic patients preseirt multiple causes for their diabetes. This is well exemplified by the history of a gentleman, aged forty-nine years, Case No. 054, who consulted me on December 12, 1915. One of his children died in 1001 at the age of two years, and another in 1013 at the age of twelve years, both of diabetes. As a child he had measles, scarlet fever, and whooping-cough, und at twenty-four years was ill for eighteen months with inflammatory rheumatism, and the pericardium was tapped twice. At the age of thirty-three years his weight was 200 pounds, and for his height 5 feet 11 inches was 17 per cent, above normal. Prior to this time he indulged in considerable alcohol three evenings a week, and his use of tobacco has been, more than moderate. lie was fond of sweets and occasionally ate half a pound of candy in an evening. During the last two years he took little exercise, and recently led a strenuous life on account of his active business. An attack of gall-stones, which was accompanied by an infection of the biliary tract, led to an operation on November 15, 101"). Prior to the operation the urine was examined and found normal. The anesthetic was ether. Convalescence from the operation was satisfactory, but while at the hospital his friends, knowing his fondness for sweets, sent him much candy, which lie ate. On December 11, 1015, he observed polyuria, and later he recalled that when nervous and working hard this symptom had occurred on" and on for a day's duration during several years. On December 11, 1015, sugar was demonstrated in the urine, and upon the fol- lowing day, when he came for treatment, the specific gravity was 1045 and the percentage of sugar was 7.2. Albumin and diacetic acid were absent, and the sediment was negative. The weight of the patient was approximately 1S5 dressed, shortly after the operation, and on December 13 was 1(17 naked. Physical exami- nation was negative. The patient began fasting by going without his supper on December 12, and the twenty-four-hour quantity of urine ending December 14 contained only a trace of sugar, and even this was absent the following day. Improvement was uninterrupted. The important ctiological factors of diabetes were all present in this case. There was a tendency to diabetes, as illustrated by the death of his two children with the disease; obesity and dietary excesses existed; a strenuous life; nervous excitement connected with the operation; the presence of an infection; the trauma of the operation, and presumable interference with the normal action of the pancreas and liver. The other factors which it is customary to consider in the etiology of diabetes, namely arteriosclerosis, syphilis (negative \Vasserinann), disease of the thyroid gland, acromegaly, gout and Bright's disease were absent. ETIOLOGY OF DIABETES 61 It was interesting that the disease did not break out earlier in the case of this patient, when one realizes that obesity had existed several years before. It plainly suggested that obesity alone was not a sufficiently strong factor to lead to the disease. Even when to this were added dietary excesses and the strenuous life, the disease still remained latent and only a multiplicity of influences, such as were brought about by the operation, were sufficient to make it declare itself. 4. Heredity. The influence of heredity has been elsewhere dis- cussed. (See p. 33.) 5. Strenuous Life.- A strenuous life has been considered by most writers as of importance in the etiology of diabetes, and it has so impressed me. Yon Xoorden records that S per cent, of all his mild diabetic patients were physicians, and this can easily be connected with the strenuous life which medical work entails. The frequency diabetes among physicians, however, is undoubtedly in part due of to the opportunity they have for detecting the disease. In my own series of male patients, above the age of 20, 10 per cent, have been physicians. Of these 15 have died, and the average duration of the diabetes was seven years ; 47 are alive and the average duration of the diabetes up to December 1, 1916, was likewise seven years. 6. Nervous Element. The nervous element in diabetes has received additional importance since the publication of the work of Cannon, Shohl and Wright 1 and the investigations of Folin, Denis and Smillie 2 upon the appearance of sugar in the urines of 192 insane patients. Of these, 22 showed sugar with the standard tests. The latter investigators also examined the urines of students after important examinations. "Of 34 second-year medical .students examined before and after an examination, 1 had sugar both before and after the examination. Of the remaining 33, 6, or IS per cent., had small but unmistakable traces of sugar in the urine passed immediately after the examination. A similar study was made on second-year women students at Simmons College. Since these students were younger and presumably much more ex- citable than our medical students it was thought that even more striking results might be obtained. This expectation did not prove to be well founded. Out of 36 taking the examination and who had no sugar in the urine on the day before, 6, or 17 per cent., eliminated sugar with the urine passed immediately after the examination." Case Xo. 10 developed severe diabetes when on an unusual business trip involving much responsibility. The appear- ance of polyuria and polydipsia was acute, but he lived twelve years. In comparatively few instances in this series was the presence of an 1 Cannon, Shohl and Wright : Amer. Jour. Phyis., 1911, xxix, p. 2.SO 2 Folin, Denis and Smillie: Jour. Biol. Chsm., 1914, xvii, p. 519. 02 DIABETES MKLLITl'X organic nervous disease demonstrated. Allen's conception of the functional nature of diabetes shows how the disease may be brought about through the nervous system by its action upon the pancreas. Thus the Claude Bernard puncture produced diabetes in one of Allen's dogs, predisposed to the same by the removal of a portion of the pancreas, and yet proved to be non-diabetic before the puncture. Jn this dog characteristic changes were found in the islands of Langerhans at autopsy. 7. Infections. The influence of infections has received consider- able attention of late. In only a few instances have I been able to associate infectious diseases with diabetes; in fact, in only 2S cases of my series, \\hen one considers the frequency of infectious diseases in a community and the rarity with which diabetes develops after the same, one is not inclined to assign great importance to infections. On the other hand, the marked lowering in tolerance for carbohydrates in the diabetic patients when an infection appears is an emphatic demonstration that this subject should be thoroughly investigated, and I certainly remain open-minded upon it. My data do not show how important a part the teeth and gums play in the etiology of diabetes. S. Arteriosclerosis. The frequency of arteriosclerosis with diabetes is perhaps best explained by the increasing incidence of the disease as age advances. 9. Syphilis. The influence of syphilis in causing diabetes is possible, either directly by disease of the pancreas, or indirectly by its effect upon the bloodvessels, which in turn interfere with the blood supply of the gland. Reliable statistics concerning the relation of syphilis to diabetes are still lacking. Nineteen cases in my series have given a history of syphilis, but such data are of comparatively little value. A Wassermann test has been made upon 107 patients, and has been found positive in (> instances. The last .")() Wassermann tests upon diabetic patients have all been negative. (See p. 2 ( ,)2.) 10. Trauma.- A definite history of trauma immediately preced- ing the disease was present in Ca*es Xos. 7 and !)Sl2 of my series. In Case No. 7 the patient observed the first symptoms of diabetes directly after being injured by a cow. He was seen by me once on October _!.">, 1S9S, four weeks later, and died in coma five months after the onset. No record of a previous urinary examination exists. Case No. DSL.' was seriously hurt in his back in a foot-ball game, in the fall of 1 !>!:>, and became unconscious. A broken neck was suspected. In the following January he again had a serious fall in the woods and again injured his back, so that he was incapacitated for three weeks. At the expiration of this time he observed that his mouth was dry, and in February sugar was found in the urine. Case No. 954 developed the diabetes after the trauma of an operation ETIOLOGY OF DIABETES 63 but there were many other etiological factors present, and further- more such an event must be extremely rare. Case No. 11SS probably represents best of all the indirect harm which may result from trauma, though his case resembles most cases of the supposed influence of trauma upon diabetes by lack of evidence showing its absence before the accident. Case Xo. 11S8 was an ice-cream manufacturer in 1915, at the age of thirty-one, weight 242 pounds. He was in the habit of eating, in addition to his regular meals, two quarts of his own ice-cream. During the subsequent year, his weight fell to about 212 pounds, when on October 27, he was accidentally shot with a rifle, and his leg badly shattered, requiring frequent painful dressings. The urine was first examined on November 27, 19 1C), and sugar found. The quantity amounted to 37SO c.c., and the percentage of sugar was large. Under treatment he gradually became sugar-free and acquired a tolerance for about SO grains carbohydrate. The fracture of an extremity is often accompanied by glycosuria. Among (>1 cases at the Surgical Clinic at Kiel, 24 showed alimentary glycosuria, and in 3 unsuspected diabetes was discovered. Hyper- glycemia was present in 31 out of 3(> of the cases examined. 1 It would not seem strange if many of the cases of diabetes following trauma might be due to the enforced idleness and rest combined with forced feeding at the hands of friends. 11. Pancreas. Disturbance in the physiological action of the pancreas or disease of the gland is a basal etiological factor which all recognize to exist in diabetes. I await with interest the results of the experimental observations of Allen upon the gland and his summary of the literature, and believe it far better to refer the reader to his report, which will surely be included in his forthcoming publication. 12. Hypophysis. The intimate relation of the hypophysis to glycosuria and diabetes has been emphasized by dishing, and is described in detail by him. 2 I.']. Liver. The connection between diseases of the liver and diabetes has been frequently pointed out. Such an association is not easily demonstrable in my cases, for out of 51(5 cases there have been only 3 who have died in which cirrhosis of the liver appeared to figure as a cause of death, and but 7 of cancer of the liver. The interesting series of 20 cases (Table 140, p. 295) in which gall-stones have been a factor probably should be credited etiologically to the pancreas. 14. Renal Glycosuria. Glycosuria with normal glycemia rela- tively independent of diet. In other words, the excretion of sugar must be due to abnormal permeability of the kidney, while the 1 Konjetzuy and Wetland: Jour. Am. Med. Assn., 1915, Ixv, p. 2264. 2 Gushing: The Pituitary Body and its Disorders, Philadelphia, 1912. 04 DIABETES MELLITUS tissues still retain the normal power of utilizing dextrose. 1 A diagnosis of renal diabetes should be very cautiously made. Most of the cases in the literature have been observed for short periods. No eases have been proven congenital and few eases have had blood- sugar examinations extending over a period of years. Bonninger's case 2 was observed for six years. The glycosuria is seldom absolutely independent of the diet. Theoretically, the renal diabetes might be of a severe type, as described by Gallambos, 3 or of a mild type, according to the depression of the sugar threshold. Lewis and Mosenthal's case 4 was carefully studied and exhibited some evidences of depression of renal function, as disclosed by a slightly diminished phenolsulphonephthalein excretion, and a slightly elevated Ambard's reaction. Murlin and Niles r ' report another instance in which throughout a period of nineteen days the blood sugar varied between O.OcS and 0.105 per cent., the sugar in the urine between 19 and 34 grams and the carbohydrate in the diet between 15 and 100 grams. Strauss describes the case of a boy, aged thirteen years, whose urine contained a non-quantitatable percentage of sugar for three weeks on a non-carbohydrate diet, but one hour after a meal rich in carbohydrate showed 0.04 per cent, sugar. Clinically, it is dangerous to make a diagnosis of renal diabetes until the patient has been under observation for several years. Among my cases of diabetes of a duration of fifteen years some may belong to this group (see p. 4(iS). Unfortunately all the recent cases of renal glycosuria, so carefully worked up, have been observed for very short periods. A duty rests upon all physicians who report such cases to describe their future course. (See page 29() for Miller's and my own experience in reporting the disappearance of sugar after removal of a fibroid tumor of the uterus.) 15. Gout. Gout is frequently mentioned as a precursor or companion of diabetes. So seldom has this occurred in my experi- ence; that I cannot attach much importance to it. Dr. J. II. Pratt, of Boston, who has seen an unusually large number of cases of gout, and has a series of '.>() cases of chemically proven gout, assures me that in no instance was sugar found in the urines of his patients. H. CAUSES OF DEATH IN DIABETIC PATIENTS. The end-results of all the 11X7 diabetic patients whom I have seen in private practice from 1X94 to December 1, 191G, have been 1 Allen: Lor. cit., p. ."> 1 1. 2 Bonniimcr: Verb. d. Cunt:, f. Inn. Mod., lOl.'i, xxx, p. 17. 330), because these are the cases from whom lessons for the improvement in the treatment of diabetes can be learned. In fact, it is on account of these 38 deaths that I justify myself in taking the time and going to the expense of maintaining my statistics. Forty-six of the fatal cases were among males. This constitutes 0.0 per cent, of the total males traced. Thirty of the deaths were among females, or 0.5 per cent, of the females traced. Coma was the cause of death of 04 per cent, of all the cases when the statistics were collected up to November 18, 1915; but now (December 1, 1910) the percentage has fallen to 00 per cent. In no other way will improvement in treatment show more strongly than in the gradual decrease of this percentage. A death by coma in diabetes is like a death by perforation in duodenal ulcer a death to be prevented. Consequently a decrease in the percentage of deaths in diabetes by coma speaks well for the change in treatment. A new cause of death appears this year in the mortality tables, and it, may increase as a factor, but I hope for only a few years to come. This is inanition, and to it three deaths must be credited. Otherwise, save for numerical changes in the individual causes, the table- remains the same. 1 . Deaths without Coma. The causes of death of these patients will be discussed under six headings, , into the first of which I will place the miscellaneous and uncommon cases. i DIABETES MELLITUS TABLE 44. CAUSES OF DEATH OF FIVE HUNDRED AND SIXTEEN DIABETIC PATIENTS SEEN IN PRIVATE PRACTICE <)\E OR MOKE TIMES FROM 1X94 TO DECEMBER 18, 1915, COMPARED WITH SIMILAR DATA EXTENDING TO DECEMBER 11, 1910. DEATHS WITHOUT COMA: To Doc is, To Doc 1. Miscellaneous: icir,. line,. Pernicious anemia 1 1 Cirrhosis 2 3 Oldajie 2 2 Suicide 2 2 Drowned 1 1 Diabetes 8 11 Cause unknown (i 10 Total miscellaneous 22 30 II. Cancer: Breast 1 Face 1 1 Esophagus 1 1 Stomach 4 6 Intestine 1 Kectum 1 1 Liver 6 fl Kidney 1 Bladder 2 3 Bones (sarcoma) 1 1 Location unknown 1 1 Total cancer 17 23 111. Tuberculosis: Pulmonary 1(1 17 Peritoneal . . 1 Meningeal 1 Total tuberculosis Ill 19 IV. Carcliorenal and vascular: Cardiac .28 34 Chronic nephritis . 14 lo Cerebral hemorrhage ...'... .14 20 Arteriosclerosis 6 ('> Pulmonary edema . . 1 Knlariied prostate .... 2 Total cardiorenal and vascular, renal and pros- tat ic (12 7S V. Infections: Pneumonia . . !"> 22 Influenza . .3 3 Tonsillitis . ... .... 1 1 Krysipelas . .1 1 (iu'll-stones .... ... .1 1 Acute appendicit is . ... 2 ~> ( 'urbuncle .... 1 ."> Acute abdominal 2 Abscess, iun>r 1 Sepsis or fiangrene of lower extremities . . 9 14 Total infections VI . Inanition o DEATII> WITH COMA ........ . 27:! 30S or or 64 per cent. 60 per cent. CAUSES OF DEATH IX DIABETIC PAT I EXT 8 07 (a) Miscellaneous. In this group fall 1 death of pernicious anemia, 3 of cirrhosis of the liver, 2 of old age (ages seventy-eight and eighty-six years), 11 eases of whom 1 can secure no other details than a death report signed diabetes, as well as 10 who died of causes unknown to me. The small number of deaths from cirrhosis always appears remarkable. One patient was drowned and 2 committed suicide. Of the latter, one was mentally unbalanced, and the latter took her life after becoming pregnant a second time soon after an abortion. Today I should not advise an abortion unless vigorous and prolonged treatment failed to render the urine sugar- and acid- free. Xo patient has died as a result of pregnancy since this patient took her life in October, 1014. (6) Cancer. Twenty-three of the patients died of cancer. The primary growths were in the breast, face, esophagus, stomach, intestines, rectum, liver, kidney, bladder, and I include 1 case of probable sarcoma of the pelvic bones. The average age at death of these patients was sixty-five years and four months. Cancer apparently developed after the diabetes in 15 cases, and may have developed before or coincident with the diabetes in 8 others. Incidentally, I would say that 2 cases- of cancer which have been operated upon are alive. I of the uterus. Case Xo. 799, and 1 of the breast, Case Xo. 1088. Of two patients with cancer of the bladder one died suddenly of pulmonary embolism, nine days after its removal and the other with recurrence in the abdominal scar. (('} Tuberculosis. Pulmonary tuberculosis was responsible for the death of 17 cases, thus adding but one death to those of a year ago. This is significant, for so far the predictions which followed the introduction of the fasting treatment have not been fulfilled. Those gentlemen who saw tuberculosis in every fasting diabetic, forgot that fasting simply ushered in the new diabetic campaign for life. The medical profession has altogether too pessi- mistic a view about this complication. \Yhen the treatment of the diabetes is faithfully carried out. these patients do quite well. The trouble in the past has been that consumption was usually advanced when diagnosed. In a diabetic, temperature, pulse and respiration, may give no clue to the diagnosis, and the loss of weight is attributed to the diabetes. One patient died of meningeal tuberculosis, the pulmonary symptoms having subsided in large measure and another of tuberculosis of the peritoneum. (d) Cardiorenal and Vascular. Cardiorenal and vascular changes, uncomplicated by coma, caused the death of 78 cases. Of these, 34 died of heart disease, in four instances suddenly with angina pec- toris; 15 died of chronic nephritis; 20 from a cerebral hemorrhage; G of general arteriosclerosis. The average age at death of these patients was sixty-four years. If the methods we employ in the OX DIABETES MELLITVS treatment of these same arteriosclerotic conditions in non-diabetic patients are correct, it is difficult to see how modern treatment with restricted instead of forced feeding can fail to prolong the life of these individuals. Two of the patients succumbed to an enlarged prostate. (c) Infections. 1 he advent of an infection lowers the tolerance of a diabetic for carbohydrate and thus increases the severity of the disease. This is an old and reliable clinical fact. It has lately been emphasized by the Chicago School. 1 Case No. cS13, aged eight years, sugar-free for eighteen days, developed tonsillitis and with- out change of diet a positive carbohydrate balance of 50 grams dropped to 14 grams, with reappearance of acidosis. Recent experiences with fasting treatment show that in the presence of an infection a diabetic becomes sugar-free very slowly, and if sugar- free, sugar may return. An infection is an additional load for the diabetic to carry and to it he often succumbs. If those cases are excluded in which coma was an element, the number of deaths from infections is comparatively small, for there were but of). This is a distinct increase over my figures of a year ago. It is quite possible, indeed probable, that acidosis existed in company with the infection, but the infection appeared so much more important that it was disregarded or overlooked as a cause of death. Fortunately, most patients, no matter what the disease, are unconscious when they die and if coma did occur in this group it was not attributed to diabetes., The point is a hard one to decide, and too much weight must not be put upon statistics. Of general infections, pneumonia heads the list with 22 deaths, influenza claims )>, and tonsillitis and erysipelas each 1. A con- siderable number of patients have passed through pneumonia suc- cessfully, e. , 4(>, .V>S, 4ii5, Si)") and 1274. Of local infections, septic and gangrenous legs account for 14, car- buncles for f>, and acute fulminating appendicitis for f> deaths. Coma was a terminal event in some of the cases of appendicitis, though the abdominal condition appeared to be sufficiently severe without it to cause a fatal issue. One patient died of gall-stones without coma. Two died of acute abdominal infections and of these the physician thought pancreatitis probable in one. One patient died of an abscess of the lung. (/) Inanition. The three deaths of inanition were the first of this character in diabetes which I remember to have seen in twenty-three years. The later symptoms of these patients were painless, the weak- ness was by no means overpowering, and the same tendency to hopefulness was exhibited by the one adult, Case Xo. 1085, which is so frequently seen in the consumptive. These three cases all showed 1 Wondyatt : Abstract Proc. Seventh Annual Meeting Am. Hoc. Advancement Clin. Invest., 1'Jl.J, p. 25. CAUSES OF DEATH IN DIABETIC PATIENTS 69 acid poisoning at one time or another. One, Case No. 785, died eleven months after leaving the hospital; one, Case No. 1120, died in the hospital, and his case is described on page 338, and the third case, Case Xo. 1085, died six days after discharge. In each case an attempt was made to increase the protein and fat in the diet, and in each case it was evident to the relatives as well as to us that if this increase was maintained acidosis would continue to increase and death would follow. Two of these patients died sugar- and acid-free, but 1 do not know the details of the remaining case. The one change which 1 should make in the treatment under similar circumstances would be to fast more rigorously the first few days, hoping thereby to be able to increase more rapidly the calories in the diet in the subsequent days. Thus 2CS deaths are explained. Of this number a good many could be saved today. One death from suicide almost certainly could have been prevented; tuberculosis could be earlier diagnosed, or better, by prophylaxis its development prevented, and even if it did occur, treatment would certainly today accomplish much more than formerly; likewise, present-day methods, I am convinced, will accomplish much in prolonging the lives of the cardiorenal group. Our better knowledge of the disease cannot but help in the presence of an infection, especially when caused by local sepsis. 2. Deaths with Coma. Coma was the cause of 308, or 60 per cent. , of the deaths. 1 Hiring the last fifteen years there have been treated under my personal supervision in hospitals 406 cases of diabetes, of whom 21 have died of coma, making the total deaths in hospitals for this period 26. For many years my attitude toward coma was to consider it the culmination of the diabetes, and after the death of the comatose patient I could honestly unite with the family, the physician and the nurse in the feeling that no more could have been done. But for me this comfortable creed has passed, and in its place has come the opinion that coma by no means represents the culmination of the disease, that it is not a justifiable accident, and, though not in all, yet in most cases, it is avoidable. My conception of coma in diabetes cannot better be expressed than by quoting from the valuable essay of Moynihan 1 on "Inaugural Symptoms" in abdominal emergencies with especial reference to duodenal ulcer, only changing the word "ulcer" and the phrase "in the abdomen" to "diabetes," and the word "perforation" to "coma." With the wording thus changed the text would read: "It is in dealing with the acute catastrophes occurring in diabetes that we shall probably derive the most instant and striking advantage from 1 Moynihan: The Pathology of the Living and Other Essays, Philadelphia, 1910. 70 DIABETES ME1JJTUS an attentive study of inaugural symptoms. . . . To take a specific example, the onset of cnina in ditilu'lcx. But let me first say that a catastrophe of this kind is almost always capable of being forestalled. Though the onset of count in dinl/ctcx is acute, the didlx'trx itself is chronic. It is a disease that has existed for months or years, and it has given in almost every instance, not only sustained evidence of its existence, but a recent warning that the pathological processes engaged in it were becoming more acute. The warning, however, is commonly ignored, because the significance and impor- tance 1 of it are not understood, and accordingly a disaster is precipi- tated. There are few catastrophes occurring in diabetic that are veritably acute." bearing this point of view in mind, the deaths from coma which have occurred in my practice in hospitals will be analyzed and at the same time reference will be made to those patients dying else- where, whose history was of especial significance. (See also Table 154, p. :mj (d) Ether Anesthesia. Case Xo. 720, a severe diabetic, three months after her last visit to me, without my knowledge, was taken to a dentist's office, given ether by her physician, and all her teeth extracted. This was on a Monday. She was taken home, became unconscious Wednesday, and died on Friday. It is hard to believe that her death represented the culmination of diabetes, or was even accidental. Case Xo. 34X, before an operation for removal of a prostate, was free from acid and sugar and tolerated 20 grams of carbohydrate. After light etherization followed by three days' fasting, ;>:> and 41 grants of sugar appeared in the urine on the second and third days respectively, and the ammonia was )).)> grams. It is true that he recovered, just as all but 7 of H4 patients undergoing major operations have also done, but the ether lowered his tolerance and made 1 his diabetes temporarily worse. This does not mean that diabetics should not be operated upon, but it demonstrates that ether anesthesia is a burden which a light case 1 of diabetes may easily be-ar, which may change a mode-rate 1 to a severe case 1 , and to a severe case 1 may be fatal. (b) Impaired Kidneys. IHabetic patients with vulnerable kidneys are 1 peculiarly suse-eptible te> coma because the power of elimination of acid bodies is impaired. Many will re-call cases ot Bright's disease 1 in diabetes in which an apparently mild acidosis preceded coma. Years ago (Joodall 1 and I pointed out that acidosis was miie-h more ivadily tolerated by the young than the old diabetics, and this e-tn we'll be 1 attributed to the 1 deficient kidne'vs of the 1 latter. Few could void on the 1 verge of coma the ."> or more 1 liters of urine which with the 1 old alkaline 1 treatment was ncerssary tor ive-overy, 1 (loud;, 11 and .lo.-lin: The Clinical Value of the Diabetes, Boston Med. and Sun:. Jour., l!H).x, clviii, CAUSES OF DEATH IN DIABETIC PATIENTS 71 and is recorded to have taken place in most patients who did recover. Coma by no means is exceptional in the old; thus the percentage of coma in my fatal cases under fifty-one years was 07 per cent., but above that age 33 per cent. A good many instances of this type could be specified, and in this group might well be classed a fe\v of the cases of death in pregnant women. (c) Infections. Already attention has been called to the fact that an infectious process renders the diabetes more severe. A con- siderable percentage of the cases of coma occurred in connection with either general or local infectious processes. It not infrequently happens that the infection is not recognized. Better statistics upon this point and in general about the circumstances attending coma should be accumulated. Thus, Case Xo. 830, seen in con- sultation one evening, was found to be in partial coma, but 1 was able to demonstrate to the physician a membrane in the throat, and three hours after the patient's death the following morning, the Board of Health reported a positive culture for diphtheria. But what I consider of far more importance is the number of procrastinating cases of mild infections in mild diabetics, chiefly in their lower extremities, which frequently prove fatal. The youngest case save one of sepsis or gangrene of the legs in a diabetic in my per- sonal experience has been fifty years of age. In other words, these conditions develop at a time of life when diabetes is mild, and why should they so frequently be fatal? Please consider with what these mild cases of diabetes have to contend. Handicapped by a lingering infection, which only too often is allowed to continue for months, with kidneys less efficient for throwing off the acidosis attack, deprived of exercise that recently proven stimulus to sugar consumption for whoever heard of a poor old gangrenous diabetic taking exercise these pitiful patients frequently meet a fourth enemy in ether anesthesia. Is it any wonder that a for- n.erly innocent disease becomes virulent and the victim dies of coma? There is no doubt in my mind but that if such cases had been treated vigorously, even with the dietetic methods of a few years ago, a large percentage of the legs amputated might have been saved. In fact, Dr. Stetten 1 has most ably demonstrated this. If one will read his paper it will be seen that his success in these 2 cases was due to two factors: (1) that his patients were given the very best medical treatment of the time, and (2) that they had the advantage of expert surgical care. All are very well aware that if a diabetic patient has gall-stones to be removed he instantly commands the services of the leading surgeon on the senior staff, but if a diabetic patient has a sore toe there is no house officer too young to dress it, until a few weeks later, if the patient 1 Stetteu:.Tour. Am. Mecl. Assn., 1913, Ix, p. 112G. 72 DIABETES MKLLITUS survives that long, the surgeon in tlie amphitheatre amputates the thigh. Any success that I have had with surgical patients in diabetes has been due to the fact that J know no surgery, and learned that never-to-be-forgotten lesson twenty-one years ago; and further, that no matter how trivial the ailment 1 have secured the very best surgical skill for my patients. (i7) Mental Excitement. An occasional case of coma is precipi- tated in a severe diabetic by mental excitement. A violent n't of anger in one of the hospital patients, already in a precarious con- dition, was accompanied by vomiting and inability to retain liquids. Coma soon appeared, and this case 1 know to have been duplicated by another outside the hospital. A patient with exophthalmic goitre went into coma with far more ease than is the rule. (/') The Influence of a Fat-protein Diet. The explanations of coma thus far given account for many deaths, but by no means for the majority of deaths from coma in diabetes. In these fourteen years at the hospital, 9 cases of coma have occurred under my care which could not thus be explained. It is easy to say that the patients die soon after admission, but I cannot get around the thought that if a patient reaches the hospital alive 1 am responsible for his departure from the hospital alive, too. I will partly excuse myself for the death of 1 patient five hours after entrance and for a child, aged two years, who was in coma within ten hours, but I will not plead any excuse for the other 7 deaths. A study of these patients shows none of the factors hitherto mentioned as predisposing to coma, but there is one factor which is common to all, namely, the diet consisted largely of fat and protein with little carbohydrate, or just prior to entrance or after entrance an excessive quantity of fat had been given. Diabetic patients will live untreated for many years without the appearance of coma. They suffer from complication after complication. They are tormented with sepsis, neuritic pains and pruritus; yet they still live. Their diet is atrocious. Along conies an enthusiastic young doctor, and imme- diately fat is increased, carbohydrate diminished, and the patient goes into coma. Out of carbohydrates it is impossible to form the acid bodies. When, therefore, carbohydrate is suddenly replaced with fat, we deliberately furnish our diabetic patients with material which though it acts partly as a food, acts far more as a poison. At a recent meeting in New York upon acidosis in children, Dr. .lacobi struck the keynote. lie said that prevention is the treat- ment of acidosis in children, and that those susceptible to acidosis ought not to have fat. \Vhat he said for children holds for diabetics. Diabetic patients need fat; it forms the chief constituent of their diet; but they must not be poisoned with it, they must be gradually accustomed to it. CAUSES OF DEATH IN DIABETIC PATIENTS 73 The treatment of acidosis in the past has been unsatisfactory. Like typhoid fever, it should be prevented not treated, and in uncomplicated cases this can be done. Although most of our diabetic patients will quickly and safely become sugar-free by simple fasting, for the cases already showing acid poisoning, cases of long duration, elderly patients and diabetics with infections, I believe that prior to giving any fasting treatment a possible severe acidosis may be anticipated by taking away the causes namely, by the absolute exclusion of fat from the diet, without otherwise changing the diet or the habits of the patient. After two days, or longer if desired, omit protein another, though lesser contributing factor to acidosis and thereafter daily halve the carbohydrate, hitherto unchanged in the diet, until 10 grams are reached, and then proceed with routine fasting treatment if, as seldom happens, the patient continues to show sugar. (/) Rapid Loss of Body Fluid. Vomiting at the onset of coma usually presages death because the patient is deprived of fluid with which to eliminate acids. Body liquids are so needed that the body in the course of coma becomes obviously dry. This is really only another sign of defective elimination. For many years my interest has been aroused to the necessity of salt for a diabetic. Recently, too, my attention was drawn by a patient to the fact that he lost 13 pounds during the first four days of fasting treatment. Inquiry developed that he was only given, during this period, water and alcohol. Had he been given broths or some mineral water, suffice it to say that during fasting, even for a week, he would probably have lost little or no weight. This is important because patients prior to coma frequently lose weight rapidly. Thus, in the presence of multiple carbuncles and septicemia one of my patients lost 35 pounds in eleven days preceding his death. (See Case No. 513.) Years ago in seeking for the cause of edema in a severe diabetic patient who was taking sodium bicarbonate I withdrew sodium chloride from his diet. Prompt loss of weight followed and symptoms of coma appeared. (See Case Xo. 135, page 127.) Since then I have been careful not to restrict salt and I can well understand the opinion of Hodgson 1 that water, rich in salts, is really helpful to diabetic patients. 'Hudson:. Jour. Am. Med. Assn., 1011, Ivii, p. 1187. SECTION IF. IMPORTANT FACTORS IX THE TREATMENT OF DIABETES MELU1TS. Xo disease demands a knowledge of its pathological physiology for successful treatment more than diabetes mellitus; with such knowledge treatment is almost self-evident. For this reason I will discuss various aspects of its pathological physiology as well as cer- tain conceptions of the disease, namely, those of Xaunyn and Allen, which I have found peculiarly helpful in the care of patients. A. NAUNYN'S CONCEPTION OF DIABETES MELLITUS. Xaunyn 1 thoroughly believes in the unity of diabetes, notwith- standing the manifold, and even doubtful, causes which appear to lead up to it. He sees in heredity the common bond which unites the different forms, or as he says, " to speak more exactly, the hered- ity of the diabetic tendency." Variety in the etiology of diabetes becomes understandable if one sees in the disease the development of an individual tendency. Almost any illness or injury, no matter how slight it may be, may serve as a cause. Furthermore, the experience that an individual, who at one time has been diabetic or even has had a suspicions glycosurin which has passed for years as cured and has even dropped out of memory, perhaps for decades, may again become diabetic under favoring influences, is in con- formity with this view. This diabetic tendency is generally congenital indeed, in many cases hereditary, and this heredity is demonstrable in I'O per cent, of the cases. These cases in which heredity is demonstrated differ in nowise from those in which heredity is not demonstrated. This hereditary tendency is associated also with neuropathic and gouty tendencies. The essential point in the diabetic tendency is a weakness of the metabolism which is manifested in the utilization of carbohydrate. Although we know that lesions and diseases of certain organs, above NAUNYN'S CONCEPTION OF DIABETES MELLITUS 75 all those of the pancreas, may be causes of this disturbance of metabolism, it may well be that the diabetic tendency is not so much to be sought in the lessened activity of one of these organs as in the disturbance in their harmony of action. At this stage the manifestation of the diabetic tendency is still shrouded in darkness, and it is too early to try to assign it to the pancreas or some other of the diabetic organs. By experiment with these organs, we know that diabetes can be produced even without any individual tendency to it. It might be that disturbances in the function of a diabetic organ might take place even in the absence of causes of diabetes. But this does not appear to be the rule, since we find among the cases of diabetes of the nervous and hepatic- type the same large proportion of hereditary cases as elsewhere. Besides the diseases of diabetic organs, we have as causes of diabetes most manifold general diseases and most manifold injuries, surgical operations, fits of anger, overexertion, indigestion, excesses of all kinds, and it is hard to say how these act. Yet this is certain, that so far as it concerns these, we are dealing only with an existing tendency which has been let loose by a special cause. Whether this is connected with the disease of one of the diabetic organs is hard to say. According to the manner in which the development of the diabetes stands to the diabetic disposition, Xaunyn distinguishes three forms of the disease: 1. The diabetes of young people, chiefly between thirty and forty years of age: Xaunyn's "true 1 diabetes." In this group the congeni- tal weakness of the sugar metabolism, of itself or often in conjunction with some accompanying circumstance (illness, accident, exertion, excesses), may lead to an insufficiency of the metabolism of sugar even without the addition of the disease of a diabetic organ. In this type one must conceive of an especially severe tendency to the disease, and this accounts for the especial severity of its form, but only in part, for this severity depends in high degree upon the great demand which youth makes on metabolism. 2. The diabetes, usually mild, of elderly people. In this type, the tendency is less severe. The disease comes late to development, and for this it requires a lowering of the vitality which comes with age. The age of a man depends on the condition of his arteries. So here, too, arteriosclerosis comes into play, and with the arterio- sclerosis come all those conditions to the front which favor the development of diabetes overnutrition, luxurious living, and especially excesses in alcohol. Syphilis, which is so important according to many authors in the cases of arteriosclerosis, strange 1 "Pure." 76 FACTORS L\ TREATMENT OF DIABETES MELLITUS enough to say, appears to play no great role here. Just as old age can come on with a lowering of vitality without arteriosclerosis, so, too, a growing old of the diabetic organs, quite independent of arteriosclerotic processes, may he important. )>. The organic diabetes. The role which the tendency plays in the different cases of this group varies. In general, the tendency is present, although it need not be. To the organic type of diabetes belong those eases in which the disease of the diabetic organs appears as the cause of the diabetes. Thus, diseases of the liver, of the nervous system, whether organic, functional, or traumatic, diseases of the thyroid gland and of the pancreas are here found, and the pancreas, according to experiment and autopsy, holds the first place. In these organic diseases arteriosclerosis is important, and in this way arteriosclerosis may be the cause of diabetes. B. ALLEN'S HYPOTHESIS. DIABETES- A WEAKNESS OF THE PANCREATIC FUNCTION. Allen 1 '- sees in diabetes a weakness of the pancreatic function, and he was led to this view through his own experimental work, as well as that of others. 1 . The Production of Diabetes. Removal of portions of a dog's pancreas produces a lowering of the sugar tolerance. Removal of nine-tenths of the gland results in severe diabetes. When the remnant is larger (for instance, one-eighth) milder types of dia- betes result, and the course is chronic, extending over months, and ends fatally. Such animals furnish a closer imitation of human diabetes than is produced by total pancreatectomy. 2. Pathological Anatomy of Diabetes. In the foregoing proced- ure duct communication between the pancreatic remnant, and the duodenum is preserved, and in consequence the acinar tissue does not atrophy, and in some cases it may hypertrophy. The islands of Langerhans show typical progressive alterations, which may be summarized as vacnolation of cytoplasm and loss of granulation, pyknosis of nuclei, loss of cells, and finally disappearance of islands. By suitable controls, Allen showed the changes to be specific in diabetes. IIomans :! has made a careful histological study of such changes in the islands of diabetic cats by the aid of the admirable methods of Bensley. )>. Pathological Physiology of Diabetes. An idea has long existed that the disturbance of carbohydrate metabolism in diabetes might 1 Allen: Studies Coni-emim,' Diabetes. Jour. Am. Med. Assn., 1014. Ixiii, p. fKW. -Allen: Tin- Treatment of Diabetes. Huston Med. and Sum. Jour., I'.H."), clxxii, p. 7i:i: also see Jour. Am. Med. Assn., I'.IUi. xlvi. p. 1.VJ.1. Allen and DuHnis: Arch. Int. Med., I'.HO, xvii, p. 1010. Allen: Am. Jour. .Med. Sc., 1917, eliii, p. JiKi. 3 Ilonituis: Jour. Med. Research, l'Jl-1, xxx, p. -10. ALLEN'S HYPOTHESIS 77 perhaps be explained by the assumption that sugar exists in a combined form in the normal body, but that this combination is lacking in diabetes. This combining substance has been designated by Allen with the figurative title of "amboceptor," to indicate its function as a bond between tissue and sugar. The substance which thus combines with dextrose, or some link in the combination, is supposed to be furnished by the islands of Langerhans. It was pointed out by Thiroloix and Jacob that some dogs after partial pancreatectomy are free from glycosuria on a meat diet, but show glycosuria when carbohydrate is fed, and that continuance of carbohydrate diet finally brings the animals into a state of severe diabetes in which they excrete sugar even on meat diet. Independently of these writers, Allen made similar observations, and thinking them of much importance, repeated and extended the work, about which he was able to make the following statements: After removal of sufficiently large fractions of the pancreas, as above described, dogs develop a severe diabetes, in which they show heavy glycosuria on meat diet and also during considerable periods of fasting. The condition progresses steadily downward to a fatal end. \Vhen the remnant of the pancreas left in situ is slightly larger, a condition may be produced in which the fate depends on the diet. On meat feeding such a dog is free from glycosuria, and remains so for months, eating his fill every day and maintaining full health and nutrition, with no sign of downward progress, but subcutaneous tests show that the dextrose tolerance is very low, and bread feed- ing readily produces glycosuria. A return to meat diet stops the glycosuria; but if the bread diet and accompanying glycosuria are maintained for too long a time, the glycosuria then continues even on meat feeding. The diabetes thus produced is not inferior in severity to that resulting from simple removal of larger fractions of pancreatic tissue, and the downward course and fatal termination are similar. When the pancreatic remnant is still larger, glycosuria is absent on meat diet, and on bread diet may be absent or transitory. Such animals may remain in excellent condition indefinitely on bread diet, free from glycosuria or any downward tendency, but if suffi- cient sugar is added to the diet, glycosuria can be produced and maintained. After a period of such glycosuria, the animal reaches a condition in which it has glycosuria on bread diet. By prolonging the glycosuria on bread diet, the dog finally reaches the condition of severe diabetes, with glycosuria on meat diet, and continues downward progress. When the pancreatic remnant is still larger, sugar feeding may produce transitory glycosuria, but it cannot be made to continue. 7S FACTOR* /.V TREATMENT OF DIABETES MELLITUS The sugar tolerance is lower than in normal dogs, hut nevertheless the doses of sugar necessary to produce glyeosuria are higher than can be tolerated as a daily routine l>y the gastro-intestinal canal. The changes previously described as occurring in the islands of Langerhaiis occurred whether the diabetes followed directly on the operation alone, or was induced in the prepared animal by means of modification of diet. The best of the control animals were still alive when Allen published his paper. lie found the control tissues, examined up to that time, indicated that in animals predisposed by operation, when diabetes is prevented by the carbohydrate-free diet, the signs of exhaustion and degeneration in the islands of Langerhans were absent. lie pointed out that if further study shows that this is the case, it will a f lord additional evidence that these changes have a direct relation to the diabetes and that they are a result of functional overstrain. In a Harvey Lecture 1 delivered in the autumn of 11)10, Allen reported his MICCCSS in producing acidosis in partially depancreatized dogs. These observations fill in the gap which has existed between the artificially produced diabetes in dogs and the spontaneous diabetes of man. The significance of this discovery is fully as great as his earlier contribution to the treatment of diabetes. Heretofore attempts to combat acidosis have been limited to observations upon human beings, but now experiments upon animals can be tried and the knowledge derived thereby will be applicable to mankind. Taking a suitable dog, free from cachexia, Allen sketched what happens "when lie is forced either to hold or to gain weight in the presence of severe diabetes. lie cannot long hold weight on carbo- hydrate or protein; the one food for the purpose is fat. In some dogs, and in any dog if the diet is not carefully adjusted, vomiting, diarrhea and loss of weight prevent a perfect result. The tendency to digestive disturbances is like that of human patients on similar treatment. If digestion and absorption remain adequate, the breaking do\vn of metabolism is manifested by increasing acidosis. There is repugnance to fat and hunger for carbohydrate as in human patients, but the animal's \vi>hes must be disregarded as has been done in human case>. and the fat given forcibly if necessary. The highest fat diet is the most quickly toxic, but excessive quantities of fat are not required, and both protein and carbohydrate aid digestion and do not interfere with the result, so long as fat is continued. For the sixe of dogs used, the acidosis diet has some- times been I .">() to 200 gin. suet and L'OO to 100 gin. beef-lung, or 100 to 1.10 gni. suet, '200 gin. lung and ">0 to 1.10 gm. bread. Lipemia 1 Allen: Am. Jour. M^l. Sc., 1917. cliii ALLEX'S IIYl'orilKMK 79 is present; there is malaise and depression of spirits as in patients with aeidosis, and digestive upsets increase. If the animal is well suited for the purpose, if the diet is properly adjusted, and if there is enough day and night watching of all details, it can be shown that dogs thus go into fatal diabetic coma on full mixed diet. "Second, we may take the customary treatment of moderate diabetes and illustrate it in dogs. Suppose that suitable operation and overfeeding have produced a condition where there is marked glycosuria on a kilogram of lung, but sugar-freedom on M)() gm. lung, along with a fair state of nutrition and entire absence of ketonuria. Now place the dog on (>00 to SOO gm. lung and 100 to 200 gm. suet, according to the classical method. There is no glycosuria, weight is gained, and the condition is splendid for weeks and possibly months. The treatment is highly successful. Closer examination shows the presence of hyperglyeemia and slight ketonuria, which are usual in the patients of corresponding type. (Hycosuria follows, illustrating the spontaneous downward progress which the authorities describe. This is cleared up by a few fast days on the Xaunyn plan, and the diet is again adjusted; it may now be 400 gm. lung and 200 gm. suet. The gain in weight continues as before, with hyperglyeemia, ketonuria, and subsequent glycosuria. Againlhe fast days are used and the protein diminished so that the diet is perhaps 200 gm. lung and 200 gm. suet. The same cycle is repeated. Now the dog is in splendid condition and spirits, the coat sleek, the appearance such that he might create a good impres- sion out walking in the park, only he has difficulty in remaining sugar-free on even the protein minimum, and the fat may be pushed higher to maintain nutrition against the repeated fast days. If the dog has actually been kept fat, a fasting period about this time may diminish tin' glycosuria or it may remain high. The previously lively and hungry animal begins to show a curious little mournful- ness, and complete repugnance to food. A day or two later, vomiting of clear mucus begins, and the dog drinks and vomits water. The acetone reaction is heavy; the ferric chloride may be heavy or slight. The alkali reserve of the blood falls low, and the complete picture of patients who go into fatal aeidosis on fasting is reproduced. Dogs of the type first described are also subject to this result of fasting if they have been kept fat enough, but fattening is easiest in absence of glycosuria. "The third type of aeidosis in dogs is exemplified by diabetic animals kept free from glycosuria by regulated diet, or by those in which the amount of pancreatic tissue removed is not quite sufficient to give rise to diabetes. They are free from aeidosis on protein diet or on fasting; but on a carbohydrate-free diet high in fat they sooner or later develop marked ketonuria. The protein ration may SO FACTORS IX TREATMENT OF DIABETES MELLITUS be governed by the capaeity of the stomach. Probably high protein tends to increase susceptibility to diabetic glycosuria and diminish the tendency to ketonnria. 1 hose experiments also may extend over weeks or months, but we have proved upon many dogs that, with enough fat in the diet the result is invariable. The qualitative acetone test is heavy but the quantitative output relatively small, generally below 1 gram. Partially depancreati/ed non-diabetic dogs on a diet of 150 to MOO gm. snet and perhaps an equal amount of lung, may thrive in spite of ketonnria for a longer or shorter time. Ketonuria is apt to be slight. But the final outcome appears in one or two forms. One may be digestive failure and consequent loss of weight and strength, with cessation of ketonuria. In the other form, the routine measures against vomiting and diarrhea may succeed, but at the end there is foul breath, fatty feces, pros- tration, muscle-quiverings or actual convulsions, da/ed mentality and sometimes total unconsciousness. Against forced feeding the organism protects itself by vomiting, diarrhea, and remarkable cessation of absorption. The small proportion of protein contained in cream or snet gives little protection. The same result follows more slowly whenever the proportion of fat to protein or carbo- hydrate in the diet is too high. The craving of diabetic patients for carbohydrate is often illustrated in such dogs. It should be worth while to determine a law of balance for normal animals. Not only has the diabetic animal a specific sensitiveness to fat, but on low protein ration it must be unable to bear as much fat as an animal on high protein. If the danger of glycosuria prevents increasing the protein, intoxication can be avoided by diminishing the fat. The animal is thinner but safer, hungry instead of nauseated. The large quantity of fat in the diet led patients to acquire a repugnance to the prescribed diet and refuse to endure it. By will-power they sometimes endure it for a time. They live in fair comfort on moderate protein and little or no carbohydrate' as long as the fat is kept suitably low. They behave much more rationally toward simple hunger for all classes of foods than they did toward the former excessive craving for carbohydrate. Lack of self-control still claims many victims, but the proportion of patients willing to follow diet faithfully has been increased by reason of the more natural balance of foods in the diet." While various factors are undoubtedly concerned in the produc- tion of human diabetes, clinical observations indicate' that diet is an important one. It is generally recognized that there is a higher incidence of diabetes among those' races or classes of people who use an excess of carbohydrate. Luxurious living and seden- tary life are thought to predispose to this disorder. The experi- mental observations on animals suggest an explanation of this ALLEN'S HYPOTHESIS 81 relationship between diet and diabetes in the human subject. If individuals differ in the strength of the pancreatic function as in other functions, in some this may be so weak that diabetes comes on in early life irrespective of the diet. In others this function may be only a little stronger, so that diabetes may be delayed until later in life, or even to the period of senility, when there occurs an impairment of various functions. Others may be more or less below the average in pancreatic functional power, but under ordinary circumstances this is sufficient to prevent the occurrence of diabetes. In these persons, however, who might otherwise go through life with no sign of diabetes, an excess of starch in the diet may serve as an exciting cause of diabetes, and the degree of such excess may help to determine the earlier or later onset and the milder or severer type of the disorder. In other persons of this sort, as in the corre- sponding type of dogs, the pancreatic function is able to deal safely with as much starch as can be digested, but a sufficient excess of sugar is an effectual cause of diabetes. It is impossible to produce diabetes in the normal dog by an excess of carbohydrate feeding, since if too much sweet or starchy food is taken, indigestion results and automatically stops the inges- tion. But in dogs after operation and in predisposed human patients, pancreatic weakness may reverse the normal relation, so that the organism can digest and absorb more carbohydrate than it can combine and assimilate, and in this condition the production of diabetes by improper diet is possible. The best established and most generally accepted theory is that diabetes results from deficiency of the internal secretion of the pan- creas. In a relatively small number of cases there is gross destruc- tion of pancreatic tissue by infection or otherwise, and here early surgical intervention, in the form of drainage of the gall-bladder or other suitable measures, is presumably the most important treat- ment. In the great majority of cases the gross appearance of the pancreas is normal. Changes in the islands of Langerhans occur in animals with experimental diabetes, and changes in the islands of human patients are being found more frequently as they are sought more carefully. Even after the longest and severest clinical diabetes, changes in the islands may be difficult to discover, and complete destruction of all the islands in the human pancreas must be a great rarity, if it ever occurs. In animals the islet changes are secondary to the diabetes. In human cases it is still unknown to what extent the changes are primary, producing the diabetes, or to what extent they may be secondary, produced by the diabetes. It seems that a considerable functional factor is generally present; that we are never dealing with destruction of tissue which cannot be replaced, but always with a certain element of disturbed function, 6 82 FACTORS IX TREATMEXT OF DIABETEH MELLITUS which can be broken down by overstrain or strengthened by rest. On the other hand, a full return to normal function is practically never obtainable. Allen cites Macleod's 1 simple objective description of diabetes. Maeleod states how it appears at first as a weakened function of carbohydrate metabolism, next there is a weakened function of protein metabolism, and then in the severe cases an imperfect metabolism of fat. Allen says that if we follow this plain, simple idea, it guides us to a rational therapy. Diabetes is commonly looked upon as a progressive, fatal disease. Of course in one sense it is a disease. But in another sense it may be beneficial to implant the idea in both physicians and patients that diabetes is not a disease, lie considers that there is no evidence that it is an infection, or an auto-intoxication, or anything else of that order. He is not aware that an inherent downward tendency has ever yet been demonstrated in typical cases. For practical purposes he believes in keeping to the simple idea mentioned above, that diabetes is merely the weak- ness of a bodily function, namely, the function of assimilating certain foods. If diabetes is a weakness of the pancreatic function, one can understand why the break-down is most frequent in elderly persons, but generally most serious in young persons, as emphasized by Xaunyn. If a person overtaxes a weak stomach, the resulting distress punishes the error and forces him to desist. If he overtaxes a weak pancreas, nothing but intelligence can show him what is wrong. But if the conception is correct of diabetes as the simple weakness of a bodily function without inherent downward tendency, then if the patient is obedient he may be kept from going down hill simply by preventing him from overtaxing his weakened func- tion. The weak pancreas may never become a strong pancreas. The patient may never be entirely normal again. But Allen believes that if this idea is fully correct, this precaution may save life. C. IS THE TENDENCY OF THE DIABETIC GLYCOSURIA TO INCREASE? Hitherto most writers have been unanimous in the belief that the tendency of the diabetic glycosuria, particularly if untreated, is to increase. There are numerous exceptions to the rule, if the diabetes has been treated, but I know of none where the disease has been allowed to take its course without medical intervention. Since my assistant, Dr. Ilornor, and I have paid especial attention to the group of diabetics of fifteen or more years' duration, a good many cases have come to light in which the disease has shown little 1 Muck'od: Jour. Am. Mvd. A^n., 1'Jl 1. l.xii. p. 1222. IS THE TENDENCY OF DIABETES TO INCREASE 83 if any progress. All of these patients, however, have been treated to a certain extent and though the treatment may have been very slight, strictly speaking, it has been enough to take them out of the class of untreated cases. If treatment has been thorough the question changes and an answer to it will be found in the following section (I)). Nevertheless, even with some cases treated most conscientiously in the past the diabetes has become more severe. In the future I believe it will continue to do so in a much smaller number of instances. This forecast is based upon the favorable course in the past of many mild cases and the ease with which cases rather more severe, in character are now controlled. Case Xo. 8 shows this tendency of the glycosuria to increase. The case dates back many years and of course strictly modern methods were not followed, yet the patient was carefully treated. The description of the case is as follows: A woman showed the first symptom of diabetes in the spring of 1899, at the age of sixty years, and 5 per cent, of sugar was found in June. She had gradually lost during the preceding fifteen years 20 pounds, and weighed 105 pounds when the diagnosis was made. Under rigid diet the urine promptly became sugar-free; the tolerance rose to 130 grams and save for very transitory intervals remained so for nine years, until 1908. During 1908 and until the autumn of 1909 it returned, but except at one analysis was less than 1 per cent. In October, 1909, the sugar amounted to 4.0 per cent., and a carbuncle appeared. With prompt surgical care, vaccines, the restriction of carbohydrates, and the temporary utilization of the oatmeal diet, the sugar disappeared, and the carbuncle healed promptly. But the urine did not remain per- manently sugar-free, although only about 30 grams of sugar were excreted daily. In the spring of 1911, the sugar again rose at the time of an attack of lobar pneumonia, but as recovery took place and a restricted diet was instituted, the sugar disappeared. Kvi- dently the patient could be freed from sugar, but upon a diet containing only about 30 grams of carbohydrate. This seemed too narrow (compare treatment in 1912 and now) for the patient after thirteen years of dieting, so that it was practically impossible to keep the urine free from sugar continually. Residence in a hospital for a few days in September, 1912, in order to have several teeth removed, lowered the sugar to O.S per cent. Except for the brief periods of illness due to the carbuncle and pneumonia, the patient remained well during all these years, and was unusually strong and vigorous for a woman of seventy-three years, until she finally succumbed to a lingering illness subsequent to a hemiplegia, and death finally occurred due to a terminal pneumonia in 1913. 84 FACTOR* IX TREATMENT OF DIABETES MKLLITVK Yet I confess that doubt arises as to the progressive character of the disease when one sees a ease of diabetes such as Case Xo. l ( i(), who was untreated for years, and yet was able to become sugar-free with restricted diet, and apparently was able to acquire a tolerance for 20 grams carbohydrate. If the disease was actually progressive, then such a patient should ultimately, barring death from accidental causes, lose all tolerance for carbohydrate, but a case of this type has not yet come to my observation. 1 confess that these long-standing, neglected cases are apparently easily amenable to modern treatment, but the question arises as to whether they can be kept in a sugar-free condition and maintain Aveight consistent with life. Case Xo. !!)() has had diabetes for twelve years, and still has a carbohydrate tolerance for IS grains. Case Xo. .'>.") 2 has had diabetes for ten years, and has a carbohydrate tolerance for 2(> grams. The attractive illustration cited by von Xoorden to show the decrease in tolerance with neglect of treatment will not help with our present knowledge. The steady loss of tolerance in his case during one month would appear to prove the point, but today we know that the possibility of storage of carbohydrate, or rather the lack of storage of an excess of carbohydrate, in the body, must be considered in any such experiment, and in the case as reported by von Xoorden, determinations of sugar in the blood, determinations of the respiratory quotient and changes in the weight, of the patient were not recorded. Further evidence of value would be afforded by a study of very mild cases of diabetes discovered by routine or insurance. If such easels, often very little treated, remained stationary for years, it would be evidence against a progressive tendency. I'pon this point (>2 cases, which have extended over a period of fifteen or more years, offer some information. X'early all of the )>7 cases now liv- ing appear today in about the same condition as for several years. A low percentage of blood sugar in cases of long duration is also evidence against a progressive tendency. Thus in the (>2 cases of fifteen or more years' duration, of the !>(') analyses of blood sugar which have been made, 20 were under 0.20 per cent., 10 were be- tween 0.21 and 0.2."> per cent., and (> between 0.2(1 and 0.4M percent. D. GAIN IN TOLERANCE FOR CARBOHYDRATES WHEN URINE IS SUGAR FREE. The tendency of the diabetic patient to gain in tolerance for car- bohydrates when the urine becomes sugar-free is the fundamental principle upon which all treatment has been and is rightly based, and that by which the value of all therapeutic measures is deter- THE TOLERANCE FOR CARBOHYDRATES 85 mined. This proposition is a reverse of the preceding, and it would appear to hold so far as evidence is available, both experimentally and clinically, yet here again the evidence should be far more carefully weighed than heretofore. I have particularly observed this question of a gain in tolerance among my cases of long duration who have become sugar-free in the last year, and it is my general impression that even these cases increase their tolerance. Case Xo. 194 is an excellent illustration of this gain in carbohydrate tolerance in a child and its later loss when the treatment was interrupted. TABLE 45. CHART OF CASE Xo. 194. VARIATIONS i.\ CARBOHYDRATE TOLERANCE. o Q Volume, c.c. Diacetic acid. M "O" X 4. Nitrogen, gin. Sugar in urine. 255 si 3 : | . , -| ^ *j rj be I " bfi "^ H rH 2 S: o Carbohydrate balance, gin. NaIICO 3 , gin. Naked weight of patient, kilos 1908 April ls-19 1800 + 50 90 +40 8 19-20 2610 1.9 . . 7S 90 + 10 8 49.2 20-21 1710 + + 44 04 +20 8 49.2 21-22 1890 + + 00 60 8 49.2 22-23 1000 + 20 . 1 1.9 r .. 42 GO +20 8 49.4 23-24 1650 + 20 40 +20 8 49.5 24-25 1650 + 17 20 +5 8 49.0 25-26 1590 + 10 15 8 50.2 26-27 2970 + 1.5 12 10 8 50.4 27-28 1050 12 + 10 8 50.0 May 1- 2 1830 13.0 10 + 15 50 . 2 Nov. 11-12 1500 50 +50 53 . 1909 Fob. 1- 2 1520 90 +90 May 10-11 90 +90 i 55.9 July 27-28 !0 +90 55.9 Sept. 18-19 2820 13.0 0.5 17 11 15 20 53 . 22-23 3280 + + + 13.0 11.2 0.6 42 24 15 -25 20 52 . 9 23-24 2577 + + + 14.3 8.0 0.5 50 43 105 4115 20 52 . 3 24-25 2410 + + + 14.1 10.0 .. 19 15 5 20 52 . 1910 Oct. 25-26 2300 + + + 15.9 2.4 80 04 - 53 . 6 Nov. 29-30 2000 + + + + 04 - 52.6 Dec. 10 Death i n coma. Female, born August 16, 1S93, single, no occupation, onset of diabetes at age of thirteen years in February, 190S; sugar in the urine March, 1U)S; came under observation April IS, 190S; died in coma December 10, 1910. There was no history of diabetes in the family. Father died of pneumonia, mother and brother well. The past history included scarlet fever, dysentery at three years of age, measles, mumps, whooping-cough, chicken-pox, enuresis noc- 80 FACTORS IX TREATMENT OF DIABETES MELLITUS turna which ceased at the age of four years. (I have repeatedly observed this symptom in the early history of diabetic children.) Always a voracious appetite, sometimes eating six potatoes at a meal; ate much candy. 1 hiring the year preceding the onset of the disease the patient developed rapidly both in height and weight. In February, 190S, she showed weariness. Karly in March polydipsia, polyuria and polyphagia were present, and sugar was demonstrated in the urine. In the previous year the urine was normal. During the whole period of illness the patient remained in good condition and attended school with comfort. The diet was rigidly adhered to and not relaxed except when the diagnosis was at one time doubted by the local physician. Catamenia was established for the first time in March, 1900. The patient died in coma on December 10, 1010. Case No. f)f>4 (see p. 27X) shows a remarkable gain in tolerance from a minus carbohydrate balance of 50 grams to a positive balance of o5 grams, lasting for years. Case No. 2(K> (see p. 52) also illus- trates a gain in tolerance, for now this boy is able to take a free diet without the appearance of sugar. Case Xo. (>5.'> showed 5.S per cent, and 174 grams sugar on a free diet in September, 191.'->, but now, three and a half years later, at the age of fifty-six years, eats between 200 and .'!00 grams carbohydrate and is free from sugar. This patient, like several others, takes comparatively little fat. The remarkable case of (ieyelin and DuBois, 1 who progressed from a minus carbohydrate balance of 75 grams to a positive balance of 1(50 grams, is another striking illustration of gain in tolerance. Case No. 30 showed sugar for IS years, and yet now, on a very liberal diet, is sugar-free. Case Xo. )>21 had a tolerance for 90 grams carbohydrate in 1910, at the age of fifty-six years, but now takes carbohydrate LSI grains, protein 57 grains and fat 102 grams, without a trace of sugar. Case Xo. (>.'>2 (see p. )>S2) had difficulty in becoming sugar-free in July, 191.'), ;it the age of thirty-one years, upon carbohydrate 15 grams, but now maintains an active life, playing golf better than ever, and has reached a position of emi- nence in his country in his military vocation, yet he keeps sugar- free 4 upon a diet of carbohydrate 29 grams, protein S5 grams, and fat 150 grams. In general it may be said that a gain in weight of a diabetic patient lasting for several years, in conjunction with a sugar-free urine, can only take place when the total diet, in other words the total food tolerance of the patient, has increased. The best writers are all most emphatic in their approval of the endeavor to promote tolerance for sugar by rendering the patient 1 Gcyclin and DuBois; Lnc. cil., p. 112. HYPERGLYCEMIA 87 sugar-free. Xaiinyn says, " From my experience I consider it highly ])rol)al)le that among the early, strictly treated cases which passed in the beginning as severe, but later took a favorable course, there is many a one for which one must thank this early strict treat- ment; moreover, on the other hand, there can be no doubt that the cases which run ultimately a severe course have undergone little, if any, energetic care." And again he urges not to be "content to maintain the patient for a time in just an endurable condition, but rather to strive to improve the diseased function, or at least hold in check further inroads on the same." So universal are the two principles that there is an increase in severity the longer the disease lasts and that the progress of the disease is checked by making the patient sugar-free, that it makes any glycosuria, no matter how inconsiderable, worthy of energetic treatment. E. HYPERGLYCEMIA. The percentage of sugar in the blood of normal individuals ranges from 0.0(i per cent, to 0.1 1 per cent., most of the observations being in the neighborhood of 0.10 per cent. In diabetes the per- centage of sugar rises, and in the untreated diabetic usually lies between 0.20 to 0.40 per cent. Under treatment, it may fall to normal. In a few cases values are obtained above 0.40 per cent., but, as a rule, such cases appear to be on the verge of coma, complicated with nephritis, or as in Case Xo. 1015, associated with suppression of urine. The percentage of sugar in the blood of normal individuals rises promptly after a meal, and may reach 0.17 per cent. This is due to the carbohydrate and the protein as well in the food. Jacobsen 1 observed in 15 persons receiving 100 grams dextrose on an empty stomach before breakfast a hyperglycemia which usually occurred within half an hour. He also made the interesting observation that of 14 persons receiving 107 grains of bread, all reacted with hyperglycemia and with glycosuria. The only difference in action between the dextrose and the bread was in the rapidity of the rise and fall of the sugar curve, and this could be explained by the difference in the rate of absorption. Strouse- has reached similar conclusions. His analyses show that there is no fixed value for the normal blood sugar, and the normal varies with each individual depending on the diet. In two or three hours after the administration of 100 grams glucose to normal individuals, the percentage of sugar in the blood may fall below normal, but with diabetics the maximum is usually 1 Jacobson: Biochem. Ztschr., 1913, Ivi, p. 471. 2 Strouse' Johns Hopkins Hosp. Bull., 1915, xxvi, p. 214. 88 FACTORS IX TREATMENT OF DIABETES MELLITUS reached in two hours, but the decline often occupies eight to ten hours.' Other causes than food may raise the percentage of sugar in the blood. It may be increased in apoplexy, pneumonia, typhoid, tuberculosis in the presence of fever and in some cases of cancer. There is a decided increase after ether anesthesia and in operations on all kinds of cases. If the anesthesia is less than an hour the increase is .">2 to 89 per cent., but if more than an hour the increase is greater. Tain, fear of operation or of anesthesia exert a negligible influence upon the blood sugar. 2 Epstein and Baehr 3 have drawn attention to the significance of blood volume in blood-sugar estimations. They point out that the increase in the percentage of sugar in the blood is only indicative of a relative, that is percentile, but not of absolute hyperglyeemia, and suggest the advisability of studying the blood volumes and com- puting thereby the total blood sugar in cases of diabetes mellitus, and in a later publication 4 Epstein defines "hyperglyeemia as an increase in the total amount of blood sugar over the normal, and not merely an increase in concentration or percentage. Thus it is possible to have a hyperglyeemia even when the percentage of sugar is normal or below normal." Recently Myers and Bailey 5 have recorded a series of blood- sugar determinations in diabetics with nephritis in which the per- centage of sugar varied between 0.19 and O.SO per cent. In the last instance; the patient died the day after the observation, and there was evidence of nitrogen retention, as shown by the presence of 55 mg. of urea nitrogen per lOOc.c. of blood. Ilagelburg'' found an abnormally high sugar content in the blood in 2(5 cases of nephritis or arteriosclerosis, or both combined. For my own experience with patients having Bright's disease see p. 419. Case Xo. 1015 of my series deserves detailed mention because the percentage' of sugar found in the blood, so far as I am aware, exceeds that hitherto reported. A schoolmaster, aged forty-seven years, highest weight 21(5 pounds. Family and past history negative. Indefinite onset of diabetes in February, 191(5. Sugar first discov- ered March 12, 191(5, and three days later the urine contained (5 per cent.; albumin was reported absent. The patient came under my observation for the first, time on March is, 191(5. During the preceding twenty-four hours he had been fasted except for one ounce of whisky in three ounces of black coffee, which he had taken every 1 Martin and Mason: Am. Jour. Mod. $<., 1017. eliii, p. .">(). - Kpstein and A-rhaar: Jour. Am. Mod. Assn., I'.lKi, Ixvi, p. HI2'.). 3 Kpstoin and Maohr: Jour. Mini. ('horn.. MM 1, xviii, p. 21. 4 Kpstoin: I'mo. Soc. Kxpt. Mini. Mod., 101(1, xiii, p. (17. 6 Myers and Bailey: Jour. Hiol. ('hem., 1011), xxiv, p. 117. B HuwlburK: Cited Jour. Am. Mod. Assn., 1012, lix, p. 17/32. II YPERGL YCEMIA 89 two hours! Xo other liquid was given save about a pint of water in which a tablespoonful and a half of sodium bicarbonate had been dissolved. I Hiring this period he had vomited fluid which contained blood. He was dull, but conscious, and there was no hyperpnea. Xo edema. He had no fever; pulse 104; arteries not sclerotic. The systolic blood-pressure in the right arm was 00 mm. mercury, in the left arm SO mm. mercury, and the diastolic pressure in the left arm 50 mm. mercury (Tycos apparatus). These observations were controlled by another physician. The heart was little if any enlarged, and there was a systolic murmur at the apex. The liver was 2 cm. behnv the costal margin. During the twenty-four hours, 30 c.c. of urine were obtained by catheter, showing a slight trace of albumin, no diacetic acid, and a positive reaction for sugar. The sediment contained many coarsely and finely granular hyaline casts, pus, and six to eight red blood corpuscles to a field. Death was preceded by edema of the lungs and coma, although in nowise suggesting diabetic coma. The blood sugar taken twelve hours before death, showed by the Bang method, 1.15, 1.45, and 1.49 per cent, of sugar, or an average of 1.37 per cent. The accuracy of the solutions used in the test was immediately controlled with a standard solution of glucose. Estimations of the blood sugar are undoubtedly of considerable value in the treatment of diabetes, but it is certain that such inves- tigations will be made in but a very small proportion of the cases for some years to come. To promote a better understanding of the subject, an attempt has been made in the following tables to correlate the blood-sugar analyses with the age of the patient, the duration of the disease, the urinary analyses, the carbohydrate tolerance of the patient, and to give the first and last observation of the blood- sugar in a series of cases. In a former edition, I arranged all the cases upon which blood examinations had been made in clinical groups, absolutely independent of the results of the blood-sugar analyses, but in this edition I purposely omit the long tables and instead will give the conclusions derived therefrom. The method of Ivar Bang (see p. 204) was employed in the earlier analyses, but the Lewis-Benedict method since the spring of 1910. Unless otherwise stated, the blood was taken in the post- absorptive state, fourteen hours after a meal, usually between seven and nine o'clock in the morning. The influence of age upon the sugar in the blood of diabetic patients is shown in Table 46. In general the younger the patient the lower the blood sugar, but the converse is not true, and indeed, for the sixth and seventh decade, the values were nearer those of early childhood. Cases Xos. 1 192, 1224, 1231 and 1200, all under the age of ten, showed respectively, 0.06, 0.04, 0.08 and 0.07 per cent. 90 FACTOR* IX TREATMENT OF DIABETES MELLITUR of sugar within a few days of the beginning of treatment. Their course in the hospital was by no means as favorable as such low percentages would imply. Their future behavior will be watched with much interest. TAHI.F. 4(5. THE IXFLVKNTK OF Aci: ri'ox THE BLOOD SUCAR i.\ DIAUKTF.S. Ae of patients, Cases, Analyses, lilood sufiar. per cent. years. No. No. Lowest. Average. Highest. o- 5 3 3 0. 1(5 0. 1(55 0. 17 (5-10 . . . . 8 20 . 07 0.20 0.33 11-15 . . . . 1(5 43 . 07 0.19 . 45 10-20 . . . 7 13 0.13 0.23 . 34 21-30 . . . . 27 131 0.10 0.24 . 57 31-40 28 145 0.09 23 39 41-50 . . . . IS 53 0.09 . 22 0.47 51-00 35 67 09 20 0.3S 61-70 . . . . is 44 0.12 0.21 . 45 71 and over 2 3 0.20 . 27 . 36 The duration of the disease does not necessarily boar any relation to the percentage of sugar in the blood. This welcome fact is clearly brought out by Table 47 and is in striking contrast to impressions which are current in the literature. The blood-sugar analyses made upon my patients therefore give no support to the theory that diabetes becomes more severe the longer it lasts. TABU; 47. Tin: I\FLrK\a-: OF TIIF. DURATION- OF TIIK DISKASE ri>o\ TIIE BLOOD SrciAR i\ DIABETES. Duration of disease', years. I'mlrr 1 . . 1-2 . . . 2 3 Cases, No. . 54 32 23 Analyses, No. 141 120 70 Bl. Lowest. 0.09 0.07 10 3- 4 . . . 4- 5 . . . 5-10 . . . 10-15 . . . 15-29 . 14 . 12 . 2(5 . 12 i 31 37 81 19 2(5 0.12 0. 10 0.09 0. 11 0.09 The presence of sugar in the urine Muring the preceding twenty- four hours) is almost invariably accompanied by an increase of sugar in the blood. Among 207 of my blood-sugar analyses, there were only 5 instances in which this did not occur. On the other hand, the sugar in the blood may be as high as ()..">() per cent, without glycosuria. The cases in which there was glycosuria without an accompanying hyperglycemia were as follows: Cases Xos. 50M, 70(1, 1171, 1224 and 12(>0. In Table 4S is shown the range of the percentage of blood sugar in my cases and the quantity of sugar excreted in the urine in the preceding twenty-four hours. HYPERGLYCEMIA 91 TABLE 48. THE BLOOD SUGAR IN RELATION TO THE TOTAL QUANTITY GLUCOSE EXCRETED IN THE PRECEDING T\VENTY-FOUH IIoi'IlS. Excretion of filucosp in preceding 24 lirs., Cases, Analyses, Blow 1 sugar, per ( cut. grains. No. No. Lowest. Average. Highest. o 135 321 0.07 0.19 . 50 1- 5 . . . . 41 62 0.11 . 24 . 50 <>- 10 . . . . 19 29 . 09 0.24 . 35 11- 20 . . . . 27 38 0.18 0.28 . 50 21- 30 . . . . 18 22 0.15 0.2G 0.43 31- 50 - . . . 18 25 0.19 . 24 0.45 51- 70 . . . . 11 12 0.11 0.28 0.45 71-100 . . . . 8 9 0.24 0.32 0.40 101-150 . . . . 6 6 0.19 0.31 0.38 151-200 . . . . 1 1 0.13 0.13 0.13 318 . . . . . 1 1 . 25 . 25 0.25 From the above, it is evident that one can hardly speak of a renal threshold in diabetes. Various writers have suggested different levels for normal individuals. Just what the threshold is at which sugar passes out from the blood through the kidneys into the urine is not known. Ilamman and Hirschman at the 1910 meeting of the American Society for the Advancement of Clinical Investigation reported from a study of 50 cases that if the blood sugar was not above 0.17 per cent., sugar failed to appear in the urine, but that when it reached 0.18 per cent, or more there was a development of glycosuria. Foster at the same meeting, found the renal threshold of permeability to lie between 0.149 and 0.1 64 percent., basing his observations upon studies made with patients after undergoing ether narcosis. He also pointed out that blood sugar values between 0.25 and 0.35 per cent., frequently occurred without glycosuria in patients with a long history of diabetes usually in the presence of renal disease. Woodyatt likewise discussed the question and reported that the threshold could be lowered somewhat by diuretics and raised by antidiuretics. Acids tend to raise the threshold and neutral salts in diuretic concentration to lower it. A study of a series of my cases has shown an intimate relation between the blood sugar and the carbohydrate balance. Thus a blood sugar of 0.44 per cent, was found both when the carbo- hydrate balance for the preceding day was 41 and +1, and on the other hand with a blood sugar of 0.11 the carbohydrate balance varied between 52 and +45. The variable diet of the patients accounts for these differences in part, but the quantity of stored carbohydrate must also be important. In order to arrive at some conclusion as to whether the percentage of sugar in the blood would give a clue to the character of the disease, I have classified cases upon which blood examinations have been made into clinical groups wholly independent of the results of the 02 FACTOR* IX T If E ATM EXT OF DIABETES MELLITUS blood-sugar analyses. In each instance the case number, agent onset, duration of the disease at, the time the analysis was made, the percentage of sugar in the blood, the quantity of sugar in the urine, or the length of time during which the urine has been sugar- free, the condition of the kidneys as shown by examinations of the urine and the blood-pressure, and, so far as possible, the record of the diet for the preceding day were recorded. r \ he results of a critical analysis of these data are herewith reported. It. is striking to note that among a group of seven clinically very mild diabetics, of the two who sho\ved a percentage below 0.11, each presented higher values subsequently rind not a case in the group could be classed as renal diabetes. For example, Case No. 70, with 0.24 per- cent, of sugar in the blood, has been under my observation for twelve years, during which period the percentage of sugar in the urine has never risen above 2.(>, although the diet lias invariably been liberal. The higher percentage of blood sugar in this patient is in marked contrast to Case Xo. 1007, who suggested at the first observation a moderately severe type of diabetes of sixteen years' standing, yet whose blood sugar was 0.1") per cent. Cases Xos. ',}'.}], !>!)2 and ")().'), who showed 0. 1(1, 0.1(5 and 0.00 to 0.1 percentages respectively, all were of the familial or hereditary type. Case No. T.SO showed only 0.07 per cent, of sugar in the blood on November 2!), 1 !)!.">, but on March 10, 1!)1<> this had risen to 0.1S per cent. This case was that of a Harvard student, who made' the interesting observa- tion that despite a liberal carbohydrate meal at noon no sugar appeared in the urine if he exercised strenuously immediately thereafter. The diabetes of Case No. So."), with 0.1(5 per cent, was discovered by life insurance examination, and he repeatedly showed traces of sugar. ( 'ase No. *,) IS, blood sugar 0.11, was obese, and the urine never contained more than a trace of sugar. A closer relation between the percentage of sugar in the blood and the severity of the disease' became apparent when the cases of long duration were collected and divided into those with low and those \vith a distinctly high tolerance for carbohydrate. (Table 4S, 1st Kdition.) The average percentage of sugar in the blood of the 1 ") patients with a tolerance for only 20 grams or under, as deter- mined by thirty-seven tests, was 0.2)!. The average percentage of sugar in the blood of S patients whose tolerance was distinctly high, as determined by eleven tests, was 0.1*. It is evident that, as a rule, the lower the tolerance the higher the percentage of blood sugar, and il is also evident that when the tolerance is distinctly high the blood sugar is but little above normal, although there are exceptions. Case No. S.~>7 showed 0.2 per cent, of blood sugar, and the urine contained neither albumin, casts nor sugar, but the blood- pressure rose to 2 10 at one time and the diet was only slight lycst noted. HYPERGLYCEMIA 93 It is significant that among the cases of long duration it was not the rule for the blood sugar to be particularly high, no matter whether the tolerance for carbohydrate was low or high. ]t was the exception for the blood-pressure to have risen to an abnormal degree; furthermore, it is also interesting that, even in the presence of a high blood-pressure, frequently the sugar in the blood Avas not high. The average percentage of sugar in the blood of 29 patients with severe diabetes as determined by one hundred and eleven tests was 0.21]. It is rather striking that the percentage of sugar in the blood in only one of nine children (Case Xo. 7S5) was particularly high. Among these clinically severe cases in adults, there were two whose low blood sugars 0.13 and 0.11 respectively were excep- tions to the rule. Case Xo. 5(54 represented a severe type of diabetes, although ultimately becoming sugar-free, and now has remained so for a period of two years. Case Xo. 70(5, it is true, represents a serious form of diabetes, but in spite of this fact he eventually became sugar-free and at the present writing has remained so for five and a half months. The blood sugar indicated better than the clinical impression the true nature of the case. Repeatedly severe cases of diabetes take an extremely favorable course and ultimately prove to be mild. One would expect that these cases would show a particularly low percentage of blood sugar. Such, however, is not the fact, for in many instances in my series the blood sugar was high even though the patient was sugar- free. On the other hand, cases classed clinically as severe, but becoming mild with treatment, occasionally give an indication of this in the blood sugar. The average percentage of blood sugar of 10 such patients, as determined by twenty-three observations, was 0.20. Of three moderate cases which ultimately became mild, the average percentage of blood sugar as determined by four observations was 0.1(5. The following table (Table 49) shows a series of simultaneous analyses of the blood and urine made upon different patients during the course of treatment. The table is divided into two parts: the first comprises the cases which are still alive, and the second is made up of fatal cases. It will be seen that, as a rule, treatment is accom- panied by diminution of blood sugar. It is probable that the fall of blood sugar would have appeared more marked had the blood been secured upon the first day when the patient came under observation when sugar was abundant in the urine. With the earlier patients this was seldom the case, as is evidenced by the fact that the urine was usually sugar-free at both the time of the first and last blood tests. It is of interest to compare the earlier figures in the table with those later obtained with other patients. It seems remarkable TABLE 49. -ANALYSES OF BLOOD SUGAR MADE UPON DIABETIC PATIENTS DURING THE COURSE OF TREATMENT. Case * No. ge at Dura- nsct, tion, ears. yrs. Blood per sugar, cut. Time intervening between tests. Total sugar in urine during 24 preceding the time of blood analyses, gins. First. Last. First. Last. 430 20 0.7 . 25 . 25 62 days Slight 479 35 5.0 . 33 0. 19 25 " 138 oil 43 12.0 0.20 0.10 150 " C32 30 3.0 . 20 0.21 50 " 786 38 2 . 7 0.27 0.18 28 " 36 814 12 0.3 0.20 0.29 13 months 934 14 1.2 0.18 0.11 9 days 948 30 0.1 0.11 0.10 61 " 951 30 3.0 0. 13 0.22 28 " (1.3t visit) 4.0 . 20 0.15 14 " (2J visit) 960 52 10.0 0.17 0.10 24 " 965 40 3.5 0.17 0. 10 2 " 909 42 0.0 . 20 0.21 13 " (1st visit) 7.0 0.33 0.22 9 " 40 4 (2d visit' 970 58 i 0.1 0.10 0. 12 13 " 979 33 10.5 . 20 0.24 33 " 26 981 28 , 1.7 0.20 . 22 13 " 9s2 19 , 2.0 0.10 0.10 35 " 983 55 ; 0.2 0.38 0. 18 13 " 123 991 04 0.7 0.20 0.17 29 " 994 15 ! 2.1 0. 18 0.13 9 " 995 ! 0.2 0. 19 . 20 22 " 1001 12 I 1.2 . 30 0.21 13 " 54 1008 40 14.0 . 35 0.20 41 " 40 1011 25 1.8 . 32 . 24 45 " Trace (1st visit) 2 .2 0.27 0.25 IS " (2d visit) i 2.4 . 34 0.19 13 " 24 (3:1 visit) 1012 13 1.1 0.36 0.29 39 " 20 1013 43 2 . 0.14 0.24 16 " 7 1022 53 10.0 0.15 0.20 2 " 1020 22 0.1 0. 19 0.15 24 " 18 1028 29 0.9 0.25 . 28 9 " 1029 32 3.0 0.29 0.26 10 " 14 1033 50 4 . 1 0.27 . 30 19 " 40 1034 35 0.0 0.25 . 24 40 " 20 13 106(5 3s 3 . . 20 0. 19 8 " 1070 20 0.1 0. 13 0. 13 15 " 1075 43 . 1 0. 18 0.09 26 " 1 1070 53 7.0 0.13 0. 10 9 " 1083 53 1.0 0. 10 0.21 11 " 10S4 34 0.1 0. Hi 0. 15 17 " 10.SO 00 2.9 0. 10 0. 17 3 months 6 days (1 1090 18 3 . 3 0. 10 0. 17 9 d iys 1094 51 2.0 0.30 0. 13 30 " 05 1097 42 1.0 0.27 0. 15 " 100 109S 10 7.1 0. 11 0. 14 5 1101 52 7 . 7 0.20 ( ) . 2s 28 " 11 1 1 02 21 4.7 0.20 0. 10 51 " 1103 l(i 4.9 0.47 0.13 39 " 110 1110 01 3.2 0. 15 0.24 13 " 0.2 per cent. 1125 3 1 2 . 5 0. is . 1 s 25 " 1130 21 1.1 ( . 25 0.20 19 " 60 1131 20 0.3 0. 10 0. 10 14 " 1137 14 0.2 O.I 1 .) 0. 15 S5 " 107 1145 00 7.5 0. 19 0. 10 10 " 2 1147 33 2.0 0.29 0.17 IS " 15 1151 3 7.0 ( 1 . 25 ( ) . 22 30 " 42 1157 5s (1.3 0.45 0.20 ;-;o " 08 1159 30 2.4 0.34 0.29 13 " 75 1 100 25 1 . ,s 0.40 0. 17 33 " 92 3 1 173 15 0.2 0.45 0.11 11 " 15 11S1 21 1.1 0.57 ( ) . 22 20 " 101 1182 3S 1.4 o. 1:5 0.09 i 11S5 15 0.2 0. 10 0.07 5 9 Av = 0.24 0.19 < ( 94) HYPERGLYCEMIA FATAL CASES. 95 Case No. Age at onset, yrs. Dura- tion, yrs. Blood sugar, per cent. Time in- ; tervening between tests. Total sugar in urineduring 24 preceding the time of blood analyses, gms Cause of death. First. Last. First. Last. 416 54 8.0 0.24 0.22 42 days Uright's. 500 06 4.0 0.13 0.17 2 " Cancer of liver. 7(55 21 1.0 0.31 0.19 23 " 43 Coma. 821 23 1.8 0.34 0.17 42 " 54 Coma. 904 38 10.0 0.21 0.36 7 o 23 Carbuncle, coma. 942 29 1.2 0.16 0.20 2.6 r ; (1st visit) Coma. 2.0 0.29 0.28 40 " 42 (2d visit) 900 .'59 0.5 0.21 . 23 30 ' 01 Coma. 974 12 0.1 0.17 0.17 11 (1st visit) Coma. 0.7 0.43 0.33 9 ' o~ 4 (2d visit) 990 62 1.2 0.33 . 29 25 ' 10 Coma. 1004 44 0.4 0.41 0.28 3 ' + + Coma. 1005 28 0.3 . 24 0.24 26 ' 13 39 Coma. 1010 05 7.0 . 30 0.20 2 " 09 25 Coma. 1025 20 1.0 . 28 0.15 23 " 3 Coma. 1049 27 0.7 0.20 0.22 18 " Trace Abscess of . lung. 1058 38 0.6 0.14 0.14 9 " Coma. 1085 34 0.5 0.25 0.22 104 " 3 5 Inanition. 1120 G 0.7 0.32 0.28 68 " 16 8 Inanition. AY. = 0.26 0.23 that so many patients should become sugar-free and yet the blood sugar remain so high. At first I thought this due to the short period of time intervening between the first and last blood test, but such an explanation will not suffice. It would seem to indicate that rigorous dietetic treatment should be continued for a long period after the patient becomes sugar-free. Martin and Mason 1 also observed that whereas the blood sugar usually fell to normal with the fasting treatment, "with gradual increase in the diet, how- ever, the blood sugar rose to an abnormal degree, usually in the absence of glycosuria." This is a disappointment, for one always regrets being obliged to discharge a patient from the hospital with a normal urine, but with an increased blood sugar. Rogers 2 was most fortunate with his 27 cases, for at the time of their discharge their average blood sugar was 0.12 per cent. I must confess that an increase of blood sugar during treatment causes me many mis- givings as to the policy of treatment adopted in the case. The rise in blood sugar of Case Xo. 904 was accompanied by the extensive development of a carbuncle. On the other hand, Case Xo. 1004, who entered the hospital in beginning coma, showed a decrease in blood sugar as the coma deepened. Martin and Mason: LMC. cit., p. 88. - Rogers: Boston Med. and Surg. Jour., 191(5, clxxv, p. 152. 96 FACTORS IX TREATMEXT OF DIABETES MELLITUS F. BLOOD LIPOIDS. 1. Blood Lipoids in Health. In the first edition a summary of the literature with a statement of the existing knowledge of the blood lipoids was kindly written for me by Profesor Ploor of the Harvard Medical School. In this edition 1 have drawn freely from the earlier article, and have added material from subsequent papers of Professor Bloor, in many cases quoting his own words. The tables have been prepared by my assistant, Dr. Horace Gray, who is spending half of his tiir.e for eight months in Professor Bloor's laboratory. Lipoids (fat) are normally present in blood in four forms: 1. Glycerides of the fatty acids, usually oleic, palmitic or stearic, of which an example is triolein- ,,OOr CnHss C.-.H 5 HOC (\-Iln N oor r 1T H 3 3 glyceride of oleic acid (a compound (ester) of glycerin and a fatty acid). These contain about 9o per cent, of their weight of fatty acids. 2. "Lecithin" a compound of glycerin with two molecules of fatty acids and one molecule of phosphoric acid, which is in turn combined with one molecule of cholin. The fatty acids compose about 70 per cent, of the whole compound, and the phosphoric acid about 12 per cent. In addition to various kinds of lecithin other phosphatides are present in the blood, notably eephalin, which is very nearly like it in composition and is determined along with it. In the graphic formula? the fatty acids are in black-face letters, the alcohols (glycerin, cholesterol and cholin) in capital letters and the phosphoric acid in italic capital letters. OUT ('iTlIs-j C.-,H 5 OOf (YJI.-n Oleo-palmito lecithin contains: 1 molecule of glycerin; 2 mole- cules of fatty acid (oleic and pal- N O mitic acids); 1 molecule of phos- phoric acid; 1 molecule of cholin. (CH t } -N HO .'!. Cholesterol- a secondary alcohol belonging to the terpene series of compounds and containing one double bond, OH C^H;,/ CH CH : CH, BLOOD LIPOIDS 97 About two-thirds of the cholesterol in the plasma is combined with fatty acids, thus forming esters, but in the corpuscles all of the cholesterol is free. In the cholesterol esters the fatty acids con- stitute about 44 per cent, of the whole molecule. 4. Cholesterol esters combinations of cholesterol with a fatty acid, ordinarily oleic or palmitic acid. The formula for cholesterol oleate is as follows: xOOCnH M ^CH CH CH : In a discussion of the lipoids of the blood one usually refers to (1) the total fatty acids, including in this term (a) the fatty acids in the glycerides, (/;) the fatty acids in lecithin, (<) the fatty acids in the cholesterol esters; (2) lecithin; (.'>) the cholesterol existing free or combined with fatty acid, as a cholesterol ester. The methods for determining the lipoids of the blood are com- plicated, a full analysis by an expert requiring at present three or more hours. For convenience the methods devised by Professor Bloor are given on pages 207 and 20S. The total fatty acids represent the predominant form in which fat appears in the blood, and normally vary in the whole blood between 0.29 and 0.42 per cent. The lecithin comes next and varies normally between 0.2S and O.oo per cent. Cholesterol is least abundant and varies between 0.19 and 0.2") per cent. In Table 50 are given Bloor's figures for the high normal, average normal and low normal values of total fatty acids, lecithin and cholesterol in the whole blood, the plasma and the corpuscles. It will avoid confusion if little attention is directed to the figures for corpuscles, because their content of lipoids varies but slightly in health or disease. The figures for the plasma vary most of all, and these afford the best index of changes in the lipoids of the blood. TABLK 50. LIPOIDS OF NORMAL BLOOD. l Compiled from tables of W. R. Bloor, Jour. Biol. C'hem.. 191(). xxv, p. 5So. Cor- Whole Per cent, variation of hifih above av. 14.0 20.0 '32.0 10.0 24.0 14.0 12.0 35.0 20.0 Highest normal . 0.42 0.47 0.4") 0.33 0.2(1 0.4s 0.25 . :U 0.24 Av. (19; 2 normals . 0.37 0.39 034 30 021 042 0.22 0.23 0.20 Lowest normal . 0.29 0.30 0.27 0.2s 0.17 0.35 0.19 0.19 0.17 Per cent, variation of low below av. 22.0 23.0 21.0 7.0 19.0 17.0 14.0 17.0 15.0 98 FACTORS IX TREATMENT OF DIABETES MELLITUS The reasons for the variations in the reported values of the blood lipoids in the past are due: ( 1 ) to the inadequate methods employed ; (2) the failure in many instances to take into account the relation of the time the blood was taken to the previous meal, and (3) to the probable presence in the blood of enzymes, which may alter or destroy the lipoids if the blood is allowed to stand for any consider- able length of time after being drawn. The analyses above given, as well as those which will follow, relating to my own patients, have been made by Professor Bloor himself. Samples of blood have invariably been obtained, unless otherwise specifically stated, in a postabsorptive condition, eight to sixteen hours after the last meal (with diabetics usually fourteen hours after the last meal), and the analyses have been begun not later than two hours, and generally in half an hour, from the time the samples of blood were obtained. The influence of food upon the blood fat is striking, but as yet no test fat meal has been adopted, to serve as a standard by which variations from the normal can be detected. Speculation suggests that as wide a variation might exist in the blood lipoids, after the ingestion of fat in diabetes, as takes place in the blood sugar in diabetes after the ingestion of carbohydrate. Fasting or narcosis cause an increase in the blood lipoids, but this is not constant, apparently depending upon the nutritional con- dition of the animal, because in those instances where no marked increase results this could be obtained after a period of stuffing with fat food. However, it should be borne in mind that Ivosen- feld observed that the fat mobilization ordinarily produced by phosphorus poisoning in dogs does not take place in starving animals. It is of interest that the total fatty acids of the blood have been found to be increased in nephritis, pneumonia, pregnancy and in extremely severe anemia in animals, conditions in which acid poison- ing is supposed to play a part. Lecithin appears to be an intermediate stage in the metabolism of fat, and Bloor points out that the blood corpuscles, and in this instance the red corpuscles, are the principal carriers, take up the fat from the plasma and transform it into lecithin, and that most, if not all, of the absorbed fat is so transformed. The increase of lecithin in the blood during fat absorption suggests that lecithin is a stage through which the fats must pass before they can be utilized in metabolism. Lecithin has been found high in nephritis and high values have been observed in the blood corpuscles in leukemia, but in the cachexia of carcinoma low values obtain. It has also been reported increased in experimental conditions in animals, such as in the anemia produced by continued bleeding, and in BLOOD LIPOIDS 99 depanereatized dogs. Lecithin is destroyed upon standing by the action of an esterase in the corpuscles, and this may account for the low values occurring in the literature. Cholesterol is a relatively well defined and stable chemical substance, and is the lipoid which can be most readily determined. It does not appear to increase in the blood directly after feeding fat alone, according to recent work, but is increased in narcosis, alcoholism and pregnancy, as well as in all cases of jaundice, because of the stoppage of one of its normal paths of excretion. It is in- creased in nephritis, but decreased in cachexia of various origins. Cholesterol appears to play a minor part in the phenomena of fat absorption. Cholesterol and lecithin are constituents of all living cells and probably constitute most of the "built in" or invisible fat of the tissues. Both may be synthetized in the body. Lecithin probably takes, an active part in fat metabolism and is the first stage through which the fats pass in their utilization by the organism. The function of cholesterol is less well known. Where the vitality was low McCrudden found a relation between the severity of the con- dition and the low values for cholesterol and blood sugar. When the patient improved under treatment, the values for both blood sugar and cholesterol increased. Table 59, on page 107, illustrates the relation which has been found between cholesterol and blood sugar in my diabetic cases. Fat gets into the blood in two ways about GO per cent, passes into the chyle and in this way reaches the blood stream, but the remainder is probably absorbed directly by way of the intestinal capillaries. Formerly it was supposed that the fats, after their hydrolysis in the intestine, were resynthetized during their passage from the intestinal wall and passed into the blood stream in essen- tially the form in which they were ingested that is, as glycerides, and that alimentary lipemia was due to nothing more than the addition of these glycerides. Today it appears more likely that instead of a lipemia, a lipoidemia exists, in that, along with the increase of glycerides of fatty acids, there is also an increase of lecithin and cholesterol. Thus, it has been shown that the fat of the chyle has a somewhat different composition from that of the food fat. Since fat persists in the blood in starvation, the inference appears plain that the blood is mobilizing fat from its fat depots for its nutritional needs, and it is readily conceivable that in the presence of abnormal metabolism the fat of an obese subject might be more easily drawn upon, and that the effect would be the same as that of ingested fat. For this reason the lipoids in the blood of fat and thin people present an interesting field for study. 100 I' ACTOli'S IX TltKATMKXT <)!' DIMiKTEX MKLUTl'X 2. Blood Lipoids in Diabetes. Witli an excess of fat diabetes begins, and from an excess of fat diabetics die. In !."> per cent, of my diabetic cases an increase of body fat preceded the onset of diabetes, and in (10 per cent, of my fatal cases abnormalities of fat metabolism resulting in acidosis caused death. Disordered fat metabolism was first associated with diabetes when phlebotomy (bleeding) \vas common by the observation of the milkiness of the blood of diabetic patients. Severe diabetes was, and is, practically the only disease in which lipemia is frequent enough to be of special significance. The milky appearance of the serum and the "cream" which arose from it on standing indicated that the substance producing the phenomena was fat. When bleeding 1 fell into disuse, nothing more was learned regarding the nature of this fat until nearly one hundred years later, in 190.'>, Fischer observed that the quantity of cholesterol in diabetic blood was abnormally high, and this finding was confirmed by Klemperer and 1 inbcr; they made a similar observation with regard to lecithin. The percentage of fat reported was extreme, and in one case Klemperer found 20 per cent.; figures of I o to 20 per cent, were not unusual. Among my cases there was but one of lipemia, and in this the fat amounted to 9.4 per cent., but it should be observed that my figures are based almost wholly upon patients undergoing treat- ment with restriction of fat, and after fasting fourteen hours. In comparison with the extreme values of Klemperer, the increase of 0.4 per cent, in Case No. 7S(i appears only moderate, but in reality it, represents an increase of twenty times the average normal value. No lipemia (cloudiness or milkiness of the plasma) was observed in 2.S of my cases while under treatment, although many of them were clinically severe. In fact, lipemia has been observed by Pro- fessor Hloor in only 2 cases of diabetes, Case No. 7.-, .17 .L'l Diabetic average' :;l i .52 .59 .43 .36 .30 .46 .29 .36 .20 Normal average ' J > Norn. a] extremes . .37 .39 .34 L'7 r> .30 .21 17 .L'ti .42 .:!.". .'Is .22 .10- '"> .23 .20 .17 .2-1 BLOOD LIPOIDS 101 The patient who showed the greatest increase of total fatty acids in the plasma save for lipemia was Case No. OS.'-}, who was obese and developed severe acidosis upon restriction of carbohydrate and only after eighteen days of treatment became sugar- and acid- free. On the other hand, a clinically mild case may show considerable abnormality, due probably to diet, as in 102 St. (p. 420, Bloor), whose increase in total fatty acids was 00 per cent. His tolerance for carbohydrate was l.)0 grams. Between the mild and severe cases were all degrees of gradation in the blood lipoids, but in general the more severe the diabetic condition the more marked was the abnormality in the blood lipoids. Next in importance perhaps to the absolute increase of the blood lipoids was the fact that the relations between the lipoids were practically the same as in normal individuals. A certain tendency may, however, be noted which is significant in view of the conditions in lipemia namely, the tendency of the "total fatty acids" to increase out of proportion to the other constituents as total fattv acids . shown bv the values tor the relation - - . V. in the plasma. lecithin The occurrence of increased amounts of lecithin and cholesterol has led to the belief that the increase was due to degeneration of tissue cells, setting free their lipoids, but analyses of various tissues have*shown that the lipoid content of the tissues in diabetes is not abnormal. Increased mobilization of stored fat as the result of the partial starvation has also been offered as an explanation, but here again the evidence does not bear out the assumption. In the first place, the stored fat contains only traces of "lecithin" and choles- terol, and in the second place, though fat at onset diabetics generally become thin later, and have very little stored. fat; also even com- plete starvation does not necessarily mean increased blood fat. A third interpretation considers the increase due merely to an accumulation of food fat which the organism can no longer burn, and in the light of the knowledge that in diabetes the fat-burning mechanism is probably deranged, and of the recent discovery that lecithin and probably cholesterol (as ester) are steps in normal fat metabolism, this seems to be the most reasonable explanation. Lecithin, however, except for lipemia, varies with the fatty acids in all cases, but not as constantly as does the cholesterol. In Case Xo. "NO with lipemia, the values for lecithin were not greatly increased, as shown by a comparison with the lecithin values in this same patient when he did not present lipemia. Cholesterol runs parallel with the total fatty acids in all cases, including lipemia. Therefore the determination of the cholesterol alone in the plasma should give valuable information regarding the lipoid content of the blood in diabetes. 102 FACTORS IX TltKATMEXT OF DIABETES MELLITUS It will be seen that the normal variations for lecithin and choles- terol were so great as to overlap the variations for the lecithin and cholesterol in diabetic blood. On the other hand, the lowest figure obtainable for total fatty acids in diabetic blood was at the upper normal limit for these acids in normal blood, but this held good only for the plasma. The total fatty acids therefore afford the surest indication of a change in the blood lipoids. The absence of lipemia in the L'S cases undergoing treatment, although many of them were seven*, is a striking fact, and probably to be accounted for by the partial or complete fasting. Evidently lipemia does not result from inanition. The two eases with lipemia were on an unrestricted diet. In both the lipemia disappeared when they were put on a fat-free diet. The origin of the lipemia, in these' eases at least, appears to be the fat of the diet, and a consideration of the attendant conditions indicates that a similar explanation would account for many of the reported instances of lipemia in the past. The dietary treatment of diabetes up to very recently has been to withhold carbohydrate and make up the caloric deficiency with protein and fat. The resulting overwork of the fat- burning mechanism would be expected to cause, in some cases at least, a partial failure of the mechanism with an accumulation of fat and its metabolites in the blood. A contributing factor is the lack of available carbohydrate without which complete burning of the fatty acids does not seem to be possible. The high lipoid values where there is no lipemia may probably be regarded as an earlier manifestation of the same partial failure of the fat metabolism. The mechanism is still working in an approximately normal manner, since the relations between the lipoids are normal, but the accumu- lation of the metabolites has begun. In the actual lipemia the mechanism has fallen behind with its work and the raw materials have accumulated until they can no longer be carried in solution in the plasma. That even at this time all parts of the mechanism have not failed is evident from the increase of cholesterol, which keeps pace with the fat. That part of the process, however, which has to do with the formation of lecithin from fat has apparently failed, and it is possible that this failure may be tin* underlying cause of ihe lipemia. I nder former conditions of treatment the chances of recovery of a diabetic with lipemia were generally regarded as poor, as might be expected if the above explanation ot the cause of lipemia were true, since the continued diet of fat in- volved the overloading of an already failing mechanism. In the two cases reported here the lipemia cleared up on a fat-free diet, so that even a severe lipemia apparently need not be serious il the fat of the diet be controlled. The importance of the modification of the treatment to include control of the fat of the diet is obvious. BLOOD LIPOIDS 103 Occasionally lipcmia has been reported in diabetes where the blood lipoids were not sufficiently above normal to warrant it. As throwing light on this peculiarity the observation was made several times by Bloor that a clear diabetic plasma would become milky on standing at room temperature for twenty-four hours. Bacterial action cannot be excluded but since no such change was found to occur in normal plasma or in the plasma in other pathological conditions, he believes the development of the milkiness to be due to an inherent peculiarity ot diabetic blood probably some unstable combination which is broken up on standing. Since many of the examinations on diabetic blood reported in the literature were made on samples which had stood twenty-four hours or longer, the development of the lipemia postmortem was not unlikely. From what has been said above it is evident that as yet the significance of variations in the total quantity of the lipoids or the difference of the lipoids in diabetes is not wholly clear. In an attempt to bring the analyses thus far made into relation with fasting, the quantity of fat in the diet during the preceding day, the carbohydrate balance, the influence of acidosis and coma, as well as the duration of the disease, the following tables have been compiled. The effect of fasting upon the blood lipoids in diabetes for one or two days respectively is shown in Table 52. It will be noted that the values were lower for the second than for the first day of fast- ing, but the cases are too few in number to justify deductions save that they remained above the normal average. TABLE 52. IXFLXJEXCE OF FASTIX<; ox BLOOD LIPOIDS. Total fatty acids. Lecithin Cholesterol, Date. Case Days ems. in 100 c.c. finis, in 100 c.c. gms. in 100 e.e. 1916. No. fast- ing. Whole blood. Plasma. Cor- puscles. Whole blood. Plasma. Cor- puscles. Whole blood. Plasma - pushes. Apr. 1C) 1025 1 0.50 0.60 . 38 0.31 . 25 0.38 0.22 . 26 0.17 Feb. 2 974 1 .61 .60 . 62 . 35 .24 .41 23 .28 .18 Jan. IS Si 1 . 60 .66 .52 .47 .42 .52 .40 .57 .22 Feb. 10 7G5 1 .60 .70 .48 i .50 .48 .52 .44 . 65 .21 Av. (4) 1 . 58 . 64 0.50 0.41 . 36 . 46 . 37 0.44 . 20 Mar. IS 1011 2 53 0.66 . 35 0.41 0.34 . 50 0.31 0.39 0.20 The influence of the quantity of fat in tne diet during the pre- ceding twenty-four hours is shown in Table 53. It will be noted that with the increase of fat in the preceding day's diet of the diabetics there was in general an increase of blood fat, but this did not invariably take place. Undoubtedly some other factor beside the 104 FACTORS IN TREATMENT OF DIABETES MELIJTUS TAHLE 53. INFLUEN'CE OF DIKTAUY FAT i\ PRECEDING TWEXTY-FOUH Ilorus ox BLOOD LiroiDs. Total fattv : -ids, I.c "itliin. Cholesterol. (iins. hit nis. in 100 c.c. jrnis. in 100 c.c. triiia. in 100 f.C. No. (if in diet ot . _ cases. preceding 1M hrs. Whole 1.1 1. I'hisma. Cor- ' Whole .,, : for- ! Whole puseles. blood. ' ia ' puseles. blood. Plasma. Cor- puscles. 12 . . ()' 0.51 0.57 0.41 0.35 ( 1.27 0.4.-) 0.27 0.34 0.20 5 . . 1 20 .01 .74 .4.-, .42 . 34 .51 . 32 .41 22 f> . 21 -10 .43 .47 . 35 . 3 i .23 .42 .2.") .28 .20 2 . . (12 C,7 .61 . i 2 .47 .42 .3N 1 .47 .3!) .50 .28 3 . . Ofi His . 53 .02 .41 .41 .34 .40 .32 .40 .21 fat in the diet influenced the lipoids in the blood. This is illustrated in Table f>4, which shows the influence of the carbohydrate balance in the diet during the preceding twenty-four hours upon the blood lipoids. In reality, this table shows the effect on the lipoids of the severity of the disease. It is true that when the carbohydrate balance was 2 to 40 grains the blood fat was higher than when the carbohydrate balance was at a lower level. However, the supposition is undoubt- edly correct that when the patient had so high a carbohydrate balance, his diet in the immediate past contained far more fat than when the balance was more restricted. This might not hold true for patients living upon a constant diet outside the hospital, but in the hospital where the endeavor was constantly made to increase the diet, the statement is correct. TABLE 54. IXFU:I:.NCK OF ( 'AKBOHYDUATI: BAI.AXCK OF PRECEDING TWENTY- Forit Horns UPON THE BLOOD LIPOIDS. Cholesterol, Kins, in 100 r.c. O..V.) O.tis 0.4!l 0.41 O.3.") 0.50 0.31 0.40 0.20 to -1 5 . 1 S . 1 2(1 . 20 10 The influence of acidosis upon the blood lipoids deserves atten- tion, because it is from abnormalities in the fat metabolism that (>0 per cent, of our patients die. Table 5") was constructed to show this feature, but it shows that among these eases no uniformity exists. BLOOD L IPO IDS 105 TABLE 55. THE RELATION OF ACIDOSIS TO THE BLOOD LIPOIDS IN DIABETES. Total fatty acids, Lecithin, Cholesterol, \o of Alveolar Kins, in 100 P.O. Kins, in 100 P.P. finis, in 100 c.c. air, CO 2 . [ nun. UK. Whole , Cor- blood. A ia ' pusclrs. Whole blood. Plasma. Cor- puscles. Whole blood. Plasma. C<>T ,' puscles. Av. of V 14-24 0.55 0.01 0.4C. 0.3S 0.29 0.50 . 20 0.29 0.21 Av. of 4 1 20 .47 .53 ' .40 .34 .25 .48 .24 .2las "' a - pScles. Whole blood. ,,,.,,., Cor- Whole ' ma ' puscles. blood. Plasma. 51 ) . 57 . 45 0.34 0.26 0.44 D.25 0.29 53 .02 .41 .41 .30 .48 .30 .47 53 .01 .43 .39 . 33 .48 .31 . 43 54 . 04 . 42 .36 .27 .48 27 .31 Cor- 1 Two analyses on one of the cases. 2 Note that cases between 1 and 2 years are included in thi.- 100 FACTORS IX TREATMENT OF DIABETES MELLITUS It is encouraging that the blood lipoids bear so little relation to the duration of the disease, for this faet is an argument against loss of tolerance for food in the course of years; another argu- ment in the same category is the amenability of lipoids to diabetic treatment, as shown by Case No. 7X(>, p. 200. The relation of the blood lipoids to prognosis is shown in Table 58, in which the fatal cases are grouped together according to the duration of life subsequent to the blood analyses, and the living cases with their duration up to February 1, 1917. It will be seen that this table affords no basis by which one can predict the out- come of the disease from the condition of the blood lipoids. TABLE ,58. THE RELATION OF THE BLOOD LIPOIDS TO PROGNOSIS. Dim No. of Yrs. Total fattv acids, cms. in 100 c.c. fJIUi Whole blood. lecithin . in 100 Plasma. c.c. Cor- puscles Cholester KIIIH. in 100 '1, c,.e. Cor- puscles. Mos ! Whol( ' Alos - blood. Plasma. Cor- puscles. Whole blood. Plasma. Fatal. 3 . 4 . 1 . 3 1-4 0.47 7 . 4s S . 00 9 : 59 0.51 ..')4 .70 07 0.41 .41 .48 .48 . 34 .31 . 50 .42 0.23 0.47 0.24 0.20 0.21 . 23 .48 .37 . 20 .52 .49 .44 .34 . 05 .45 .20 .21 .20 0.20 Av.ofll fatal Living. 1 cases 0.52 0.58 0.43 . 30 0.29 '0.47 0.28 9 . 34 3 . 10 0.47 . 54 0.37 . 32 0.23 0.43 0.27 0.33 0. 19 4 . 11 .50 . 57 .41 .30 .29 .43 .2S . 35 .20 . 1 .01 . 75 .40 . 3S .33 .47 .29 .38 .21 S . i 1 1 .48 . 50 .44 . 3!) .31 .48 .31 .38 .21 Av.of21 livinjz cases . 52 0.00 0.43 0.37 0.30 0.40 . 29 0.37 0.21 The quantity of cholesterol in the blood of diabetic patients has also been compared with the quantity of blood sugar on the same date. The data found are shown in Table 59, which is so arranged that the mildest cases conic first and the severest last. It will be seen that in general the cholesterol and blood-sugar values stand in close relation in the milder cases of the disease, but that as the disease advances in intensity the divergence is great. The table brings out, however, that the cholesterol tends more to increase with the severity of the disease than does the blood sugar. :>. The Metabolism of Fat in Diabetes. When carbohydrates cease to be in whole or in part metabolized in the body the con- dition becomes evident by the appearance of sugar in the urine and the symptoms are so manifest as to be appreciated by physician 1 Duration until February 1, 1917. BLOOD LIPOIDS 107 cincl patient alike. When fat ceases to he metabolized in a normal manner no striking evidence of it is afforded, and both patient and doctor continue to journey along in innocent oblivion of its exist- ence, and hence fat is often a greater danger to a diabetic than carbohydrate. Just as increased quantities of sugar may exist in the blood without being burned, so can increased quantities of fat accumulate, and with the latter as with the former it is possible that the mass action (Lusk) of the increased quantities may stimulate and accomplish this combustion. Already Murlin and Riche 1 have shown that fat injected directly into the circulation can be oxidized at once. Heat production promptly rose with the injection of .'> per cent, emulsion of lard oil, and the respiratory quotient fell as the fat became concentrated in the blood. TABLE 59. -THE COMPARISON OP THE CHOLESTEROL AND BLOOD SUGAR IN THE BLOOD OF DIABETIC PATIENTS. Case No.a Whole blood. Blood sii(iar, Cholesterol. per cent. Kins, in 100 e c. Case No. Wh( Blood su<;:ir per eent. le blood. Cholesterol. Kms. in 100 e.c. 500 . . . 0.17 0. 19 960 . . . 0.23 0.28 90S . . . 0.21 . 23 974 . . . 0.19 . 23 1020 . . . . 20 0.21 S21 . . . 0.34 0.23 970 . . . 0.16 . 25 1029 . . . . 20 0.40 1007 . . . 0.13 0.20 969 . . . 0.21 . 40 1028 . . . 0.25 0.30 951 . . . 0.10 0.37 914 . . . 0.18 0.34 810 . . . 0.16 0.40 010 . . . 0.20 0.31 786 . . . 0.29 0.37 9S3 . . . 0.29 0.30 1004 . . . 0.41 0.21 979 . . . 0.17 0.23 1005 . . . 0.42 0.26 100S . . . 0.35 0.31 1005 . . . 0.27 0.26 032 . . . 0.17 0.33 705 . . . 0.31 0.44 032 . . . 0.16 0.25 705 . . . 0.20 . 35 0:52 . . . 0.19 0.31 705 . . . 0.17 0.44 1025 . . Lost . 22 1011 . . . . 23 0.31 900 . . . 0.24 0.22 990 . . . 0.40 0.26 900 . . . 0.21 0.20 Many problems are now ripe for investigation, and only to be solved by methods of study of the respiratory metabolism. Nearly all of the cases reported in this section whose blood fat has been determined have been investigated also with the respiratory ap- paratus, and I trust will be published as a Carnegie Publication before the end of the year. For the present I can do no better than to summarize Dr. Allen's remarks in his Harvey Lecture upon the metabolism of fat. Allen points out "that in diabetic patients or animals alimentary 108 FACTORS I.\ T RE ATM EXT OF DIABETES MELLITUS hyperglycemia is more pronounced than normal, but the effect on the respiratory quotient is less, and in severe eases may he absent altogether, and this is one of the best evidences of deficient com- bustion of carbohydrate in diabetes. The case with fat is different, for diabetic patients and even totally depancreatized dogs always burn fat readily." lie also says no experiments have ever indicated any difficulty on the part of the diabetic in attacking the fat, molecule. Patients and suitable diabetic animals often go along on a certain level of marked hyperglycemia, without glycosuria, and evidently burning some carbohydrate. They seemingly require a higher "pressure" of sugar in the blood in order to accomplish the combustion of sugar. In the moderate hyperglycemia ordinarily present in severely diabetic patients, Bloor saw evidence of a similar need of increased fat "pressure" in the blood in order for the cells to burn fat. There is opportunity to test this idea with respiration experiments. In contradiction to the prevalent belief of normal fat assimilation in diabetes, investigation will probably show that a certain level of lipemia does not have equal metabolic influence in non-diabetic and in diabetic lipemic conditions. It will very likely be found that the effect on the gaseous exchange is slower and of less degree in diabetic lipemia, corresponding to the known facts concerning hyperglycemia in the milder cases, so that an alimentary lipemia of 2 or '.} per cent, in a normal animal may represent a greater activity of fat metabolism than much higher blood-fat values in an animal with diabetic lipemia. Also, it may be found that the metabolic effect varies among diabetics in proportion to their susceptibility to lipemia, and conceivably may not be fully normal in any diabetic. Studies of this sort will throw light on the ability of the diabetic to attack the fat molecule; they may help to show why fat feeding seems sometimes neither to strengthen nor build up a patient; they will indicate what significance may be assigned to lipemia in the question of metabolism in diabetes; and by completing the proof that lipemia is due to deficient assimilation rather than increased mobilization of fat (even if increased mobilization sometimes occurs), they may contribute an analogy in support of the dominant belief that the hyperglycemia is primarily due to deficient assimilation rather than increased mobilization of sugar. TOTAL METABOLISM IN DIABETES MELLITUK 109 G. TOTAL METABOLISM IN DIABETES MELLITUS. 1. The Varying Metabolism in Severe Diabetes. The total or more specifically the basal or gaseous metabolism of diabetes attracted the attention of Pettenkofer and Yoit in 1X07. For a period of twenty years thereafter no worker ventured to take up the problem, but then various investigators entered the field, and in the last few years, with the multiplication of apparatus for metabolic studies, considerable attention has been devoted to it. 1 The chief interest in the problem has been afforded by the question as to whether the metabolism of a diabetic patient is less, greater, or the same as that of normal individuals. Whereas Pettenkofer and "\ oit came to the conclusion from their researches that the diabetic patient gave off less carbon dioxide and con- sumed less oxygen, it is highly probable that their experiments hardly justified this inference. Magnus-Levy, in 1905, in an excel- lently conducted study of 4 diabetic patients, showed an increased consumption of oxygen by the patient per kilogram body weight. A moderate number of other investigators obtained results, but these were so divergent in character that the question remained still an open one in 190X. The problem was then taken up afresh with the writer's private patients in connection with Professor Francis G. Benedict of the Nutrition Laboratory of the Carnegie Institution of Washington, situated in Boston. For the most part the investigations have been carried on in the course of the routine hospital treatment. Professor Benedict and I felt that the lack of uniformity in the results of other observers might be due to the fact that different types of diabetes were studied, and therefore our attention was directed almost entirely to the metabolism of severe diabetes. Like so many problems which have given rise to sharp discussion, it now appears as if there were two sides to the shield, for it is possible to show in the same diabetic patient that at one time the metabolism is increased and at another time it is below the normal. The story of the experiments of Benedict and myself bring out this point very clearly. Between 1908 and 1912 we studied the metabolism of '22 severe diabetics (Table (>0). Upon comparing these individuals with twenty normal individuals of approximately the same size and weight, we were able to say that the average metabolism of the diabetic group exceeded that of the normal group by from 15 to 20 per cent. The increase in the consumption of oxygen and the increase in the heat eliminated went hand in hand. The increase in the excretion of carbon dioxide was present, though less, but this 110 FACTORS IX TREATMENT OF DIABETES MELLITUS TABLK GO. COMPARISON OF TIIK CARBON DIOXIDIC ELIMINATED AND OXYGKN ABSORBKD BY DIAHKTIC AND NORMAL INDIVIDUALS IN Exi'KRIMKNTS WITHOUT FoOD. 1 (Bed Calorimeter and Respiration Apparatus.) Normal. Subject. '? J !y '>~ T u. ^5 O >' U _:5z Subject. C - f \ Ji a Severe kilos cms. c.c. c.c. kilos cms. c.c. c.c. diabetes: Case A 51.0 171 1 3 . 22 4.22 A. F. G. 53.9 175 1 3 . 30 3 . 88 10.1 171 1 2 . s<) 4.08 T. M. C. 48.5 100 is 3.25 3.80 Case B 41.4 158 5 3.10 4.29 Miss J. 43.3 157 3 3 . 30 3 . 74 Miss A. C. 42.0 105 2 3 . 24 3 . 99 Miss F. \V. 40.5 157 2 3 . 7s 4 . 57 Mrs. S. C. 37.4 155 2 3 . 20 3.77 Case C 02.7 100 6 2.81 4.12 H. H. A. 02.2 104 31 2.89 3.51 S. A. R. 01. 1 105 13 2 . 90 3.02 55.5 100 7 3.01 4.35 Dr. P. H. 55.2 104 9 2 . 89 3 . 55 C. H. II. 55.1 109 ' 9 3.14 3 . 72 Case D 48. S 173 3 3.11 4.19 T. M. C. 48.5 100 IS 3 . 25 3.80 C'ase G 07.1 178 1 3 . 22 4.3S M. A. M. 00.0 177 54 3.12 3.68 Case II 52.4 159 7 3 . 09 4.05 Miss B. 52.2 158 2 3.31 3.91 Case 1 10.0 170 1 4 . 33 0.03 T. M. C. 48.5 100 18 3.25 3 . SO Case .1 52.9 171 1 3.48 4.54 C. II. II. 55.1 109 9 3.14 3.72 ( 'ase K 59.1 180 1 2.84 4.04 Dr. S. 58.5 181 5 2.02 3 . 34 55.0 ISO 3 3.31 4.05 II. F. T. 57. S 179 44 2. SO 3 . 32 Case L 03.0 183 1 3.44 1.11 F. P. C. 03.2 185 i 4 2.71 3.47 00.5 183 1 3.08 1.23 K. II. A. (Hi./ 1S2 10 2.U2 ,1. R. 00.0 1S2 13 3.01 3 . 02 3 . 03 Case X 31.5 140 7 4.11 5 . 57 Case () 52.0 173 3 3.00 4.24 Miss F. 52.4 108 2 3.09 3 . 07 <" 'ase P 10.0 173 5 3 . 75 5.3-1 T. M. C. 48. 5 100 is 3.25 3. SO Case Q 51.7 10s 2 3.30 4.37 V. G. 54.3 102 19 3 . 00 4 . 33 A. F. G. 53.9 175 1 3.30 3 . 88 Case R 55.3 isl 1 3 . 5 1 4.90 IF F. T. j 57.8 179 14 2 . SO 3 . 32 ( 'ase S 5V 177 1 2 . 01 3.07 II. F. T. 57. S 179 14 2. SO 3 . 32 C'ase T 51.4 180 2 3 . 29 4.51 H. F. T. i 57. S 179 44 2.80 3 . 32 A. F. G. 53.9 175 1 3.30 3 . ss C'ase I." 39.5 100 5 3 . 80 5.20 Miss A. C. 42.0 105 2 3 . 24 3.99 Mi>s I'.. W. 40.5 157 2 3 . 78 4.57 Mrs. S. C. 37.4 155 ; 2 3.2(1 3.77 Case V OO.d 173 1 3.42 4.72 11. 1',. F. 00.0 173 7 3. 19 3 . 84 Moderately severe and lij-dit diabetes Case M 82.1 172 3 242 3.14 ().!'. M. s5 . s 171 2 2.41 I'rof. C. S3 .0 109 3 2.40 Case W 59.5 Kil 2 2.75 3.77 Miss B. \V. 59.4 102 2 3.03 Case X 04. (J 170 1 2 . s2 3.79 I). M. 01.0 171 2.89 Av. of mod- 2.00 3.57 Av. of nor- .. .. ..2.09 erately mal sub- >evere and jects ! li^ht cases TOTAL METABOLISM IX DIABETES MELLITUS 111 could l>e explained from the naturally lower excretion of carbon dioxide which would occur upon the fat-protein diet of the diabetics. As a reason for this increase of metabolism we called attention to the severe acidosis of these patients, and cited other experiments in the laboratory by Iliggins, which showed that the metabolism of a normal individual with an artificial acidosis was also raised. (See Table 61.) TAULE Gl. INFLUENCE OF AN EXPERIMENTALLY INDUCED ACIDOSIS UPON A NORMAL INDIVIDUAL. 1 Carbon dioxide Oxygen eliminated absorbed per miii., per min., Respiratory Average Sept. c.c. c.c. quotient. pulse-rate. 7. Carbohydrate-free diet 2 200 299 0. 07 84 8. Carbohydrate-free diet 198 280 0.71 75 9. Carbohydrate-free diet 195 272 0.72 72 11. Normal diet since Sept. 9 noon 210 243 0.87 69 Peabody, Meyer and DuBois, 3 in a study of cardiac and cardio- renal cases, also noted an increased metabolism in the two showing the most marked acidosis, but in two others with only a slight acidosis the metabolism was also high. Our results were confirmed by Holly, Leimdorfer, (irafe and Wolf, but were interpreted differently by Falta, who took part in three of the first experiments and also by Lusk, 4 though the latter acknowl- edged the increase but considered it less in degree. In 1914-1915 the subject was again approached by us, but with patients under- going fasting treatment. The results obtained were so at variance with our earlier observations that at our invitation Dr. DuBois, working in the Russell Sage Laboratory at Bellevue Hospital, came to Boston and discussed these with us, and to our mutual surprise it was found that the observations which he had just completed upon one of Dr. Allen's patients were similar to ours in that they showed that the metabolism of a severe diabetic patient when freed from acidosis changed from above normal to below normal, with a gradual rise in respiratory quotient. 5 Whether the acidosis is the sole factor in regulating this alteration in the metabolism, it would be unsafe to say. It is true that coincident with an acidosis arti- ficially produced in three normal individuals by a non-carbohydrate diet, Iliggins, Peabody and Fitz (i observed a rise in the metabolism as indicated by the oxygen consumption. However, the protein metabolism was also high and Lusk 7 inclines to the belief that 1 Loc. cit., Carnegie Pub., 176, p. 127. 2 Carbohydrate-iree diet begun after noon meal, September 5. 3 Peabody, Meyer and DuBois: Arch. Int. Med., 1916, xvii, p. 9,sO. 4 Lusk: Johns Hopkins Bull., 1915, xxvi, p. 10. 5 Allen: Metabolic Studies in Diabetes, Now York St ite Journal of Medicine, 1915, xv, p. 330. Higgins, Peabody and Fitz: Jour. Med. Research, 1910, xxxiv, p. 263. 7 Lusk: Am. Jour. Med. Sc., 1917, cliii, p. 40. 112 FACTORS IX TREATMENT OF DIABETES MELLITUS it is the increased protein metabolism rather than the acidosis which accounts for the phenomenon. Since H)12 Benedict and I have studied the metabolism of about 77 more cases of diabetes and we hope before the end of the present year to publish our results. A most careful study of the metabolism of three patients with severe diabetes has also been made by Allen and DuBois. 1 They found, in a study of the metabolism of these three patients with severe diabetes, according to comparisons of the surface area by the new "linear formula," in contradistinction to those of weight and height, that "an increase of the basal metab- olism in severe diabetes is generally absent or slight. During fasting the metabolism fell to 20 per'cent. below normal. The level of the metabolism in diabetes is the resultant of a number of forces; for examr^e, increased destruction of protein and perhaps other processes tending to increase metabolism, and undernutrition, muscular relaxation (as in prolonged confinement in bed) and other possible conditions tending to diminish metabolism. Accord- ing as one or the other of these groups of forces predominate, a higher or lower metabolism may be expected in any individual case of diabetes," and in an earlier statement occurs this sentence: "The percentage of increase may be fully as great as claimed by Benedict and Joslin, if the metabolism of diabetics is compared with that of other equally undernourished subjects, or if the same patient is com- pared at different times when active diabetes is present or absent." (ireeley- has pointed out the analogy between the steadily decreas- ing severity of diabetes as age progresses and the decrease in the metabolism. " The curve of basal metabolism coincides exactly with the curve of severity in diabetes, being greatest in infancy and child- hood and least in old age. The severity of a mild case of diabetes is at once increased by the incidence of any accident increasing basal metabolism (fever, infections, goitre)." It is of the utmost interest, however, that these observations have been made, and they should all'ord Professor Xaunyn great satisfaction, because he had the clinical acumen to discern that severe cases of diabetes treated by his precise and successful methods brought the patients to a condition where the metabolism was below normal, which he recorded in the statement that severe dia- betics would often live on less calories than the normal individual. The clinical importance of eliminating acidosis is apparent, for until this is done the diabetic patient must be supplied not only with the food which a normal individual would require, but, in addition, with enough to make up for calories lost in the form of acid bodies in the urine, and probably with 10 to 20 per cent, additional because of his increased metabolism. 1 Allen ;ii,d ])uR>i-: Arch. Int. Mod.. 1010, xvii. p. 1010. Grocloy: Knsiun Mcd. ,-md Sur-:. .lour.. liMii, dxxv, i>. 7.V3. TOTAL METABOLISM IN DIABETES MELLITUS 113 I -c c C d J- O be - xi> t- CON t^ l~ '. O) X c o X" C 71 CO ; c b- : t- : t- : t- co o o o CO O-M ~* -r -H co ;coco C CO 2 -rt^^s? o o o c o _ < "2 t- ,2 T I o -.: < 7) CM o oi ' -r co i.o L.O i.~ t^ t-^ X i-O L.O * 7 C C-. X C-. t^ o C X t^ -^ -t i rt C 1 ! ^J ^1 "N ^^ ^H ^H CO -t CO C>] C-l C^ "M ^^ C t " 5;- d ~ 114 FACTORS IX TREATMENT OF DIABETES MELLITUS The marked variations in the metabolism of diabetes is well exem- plified by a ease studied by (ieyelin and DnBois. 1 A young man, aged nineteen years, weight 1 7- pounds, began to grow thin November 1, 191.1, and on Xovember 20 weighed about, 1.10 pounds. Before he came to the authors, carbohydrate had been greatly restricted and protein (and presumably fat) much increased, with resulting seven 1 acidosis. Between December 7 and 11, he was fasted and .13 to 1 14 grams sodium bicarbonate were given daily. lie was then fed (see Table 02), alkali continued and the metabolism studied. From the table it can be seen that: (1) the metabolism varied from 73 calories per hour 'Ml calories per kilo body weight per twenty-four hours) to 43 calories per hour (23 calories per kilo body weight per twenty-four hours) in the course of a few weeks; 2 (2) that a dextrose-nitrogen ratio in excess of Lusk's 3.0.1 : 1 ratio, was obtained on three successive days, to wit: .'5.97 : 1, 4.01 : 1, 3.87 : 1; (3) that the nitrogen in the urine was extreme, being 29. S grams on the second day of fasting and MS. 27 grams even when 99 grams of protein were ingested; (4) that the acidosis was extreme, for the /3-oxybutyric acid eliminated amounted on one day to S7 grams. The case is remarkable in all of the above particulars and will always remain a classic in diabetic metabolism, but to me it is far more interesting from other points of view and my interpre- tation of it is as follows: A fat young man develops diabetes rather acutely. Such cases usually are quickly amenable to fasting or preparatory treatment consisting of a non-fat diet (see Case Xo. 923, Table 212) but if placed upon a protein-fat diet with little carbohydrate, as was this case 1 , develop severe acidosis which is not easily overcome unless taken early. Fat patients are especially prone to develop such an acidosis (see p. .'544). On December 11-12 a diet with 42 grams carbohydrate, IS grams protein and IS grams fat lowered the D:X ratio. It is true the acidosis increased, but this may be explained by large doses of sodium bicarbonate (?). At any rate, when protein and fat were increased on the following three days, the severity of the case increased and the phenomenon noted above occurred. Is it not probable that if fat had been decreased in the diet at the very beginning of treatment, acidosis would have been avoided? Did not the sodium bicarbonate act harmfully by setting free /3-oxybntryrie acid which previously was innocuously combined? Had not the case been intrinsically a comparatively mild ca-e. as was later shown by a toleration of 100 grams carbohydrate, and a youthful individual, death from coma must have resulted. As it was, the case survived (a) an initial nearly non-carbohvdrate diet; (/;) the increase of fat in the diet when 1 Goyolin ami DuBoU: Jour. Am. Mod. A.ssn.. HUG. Ixvi, p. 1.582. 2 Gophart. Auli. DuBoN and Lu-k. Arch. Int. Med., 1917, xix, p. DOS. discuss this case more in detail. TOTAL METABOLISM IX DIABETES MELLITU8 llo acidosis was well under way; (V) the setting free of enormous quantities of /3-oxybutyric acid through the use of alkalis, and (d) a furuncle, which though small, undoubtedly increased the severity of the case (see Case No. (ill), p. 314), and recovered when (1) fat was eliminated from the diet and protein in moderate quantities given, (2) followed after two days by fasting. Was it not in this ease, as in so many of my own in the past, though with less fortunate terminations, the hand of man that made the diabetes severe, just as later the hand of man made it mild. Such cases I have seen only too often (see Case Xo. 47.'!, p. 300), but I do not see them when the precautions for preparatory treatment described on pp. 307-311 are followed. Though perhaps not so intended by the authors, I believe no case on record presents a stronger argument against a non-carbohydrate diet in the early treatment of diabetes, the harmfnlness of soda, the unreliability of an exact D:X ratio even in the hands of the most competent, the sudden disappearance of acidosis when a little protein is given and fat omitted from the diet and finally, that acidosis, brought on by restricted carbohydrate and increased fat, can be readily produced in a mild diabetic and when set free with alkalis can change the mild type to that of the severest on record. 2. Observations upon the Respiratory Quotient in Health and Diabetes. The relation which the volume of carbon dioxide pro- duced by an individual bears to the volume of oxygen required during the same interval of time constitutes his respiratory quotient. This has been found to be dependent upon the character of the material in the body which is oxidized or burned at this time. An exami- nation of the composition of the carbohydrate molecule will show that it contains sufficient hydrogen to unite with all the oxygen present during its oxidation. Consequently, for each volume of oxygen used in the oxidation of carbohydrate a volume of carbon dioxide will be produced and the respiratory quotient of such a carbohydrate as glucose (C 6 II 12 O 6 ) will therefore be 1. It matters not whether the oxidation takes place rapidly outside of the body in a flame, or less obtrusively in the body during twenty-four hours. Protein, on the other hand, does not contain sufficient oxygen for the oxidation of the hydrogen atoms contained in its molecule. As a result, in the burning of protein, oxygen must be used not only for the carbon in the molecule, but for the hydrogen as well. The denominator of the fraction is tliib> increased, and the final quotient of protein must be less than 1 and is O.Sl. The protein molecule is made up of many component parts and while the respiratory quotients of these parts vary greatly, yet for protein as a whole the above quotient (O.Sl) holds. With fat a similar condition exists to that in protein, only there is still more hydrogen present to require oxygen, so that the amount of oxygen necessary 116 FACTORS IN TREATMENT OF DIABETES MELLITUS for the combustion of fat is still greater, and as a result the respira- tory quotient falls to 0.71. The respiratory quotient of alcohol is still lower, and is 0.(>7. /3-oxybutyric acid, which can be taken as the chief one of the group of acid bodies formed in diabetes, lias a respiratory quotient of O.S9, diacetic acid of 1 and acetone of 0.7"), so that one will not go far astray to take O.S9 as a common respira- tory quotient for these three acid bodies. The respiratory quotient of an individual can be determined by measurement of the quantity of carbon dioxide exhaled and the oxygen absorbed. When this is done information is obtained concerning the character and total amount of the combustion taking place in the body. Since the urinary nitrogen gives us a definite idea of the quantity of protein metabolized, if we calculate what this represents in terms of carbon dioxide and oxygen, and subtract it from the total carbon dioxide exhaled and the total oxygen absorbed we have left the combustion derived simply from fat and carbohydrate. Knowing the respiratory quotient produced when fat and carbohydrate are oxidized as well as that of the in- dividual, it is possible, by computation, to determine the share which these two variables have taken in the total metabolism. TABLE (13. THK RESPIRATORY QUOTIENT (K. Q.) OF A FOOD is OBTAINED BY DIVIDING THE VOLUME OF CARBON DIOXIDE PRODUCED DURINU ITS OXIDATION BY THE VOLUME OF OXYCEN ABSORBED. Carbohydrate: C 6 H 12 O 6 + 6() 2 = GCO 2 + GH 2 O R. Q. Oxygen is required for oxidation of the carl ion alone Volume ()CO 2 produced Volume G() 2 absorbed Fat: Oxygen required for carbon and a large quantity of hydrogen 0.71 Protein occupies an intermediate position O.X1 Alcohol (C 2 II e O) O.K7 B-oxybutyric acid (C 4 H 8 () 3 ) O.X!) Diacetic acid (C 4 II 6 O 3 ) 1.00 Acetone fC 3 II 6 O) 0.75 The respiratory quotient of normal individuals twelve hours after a meal has been determined by Benedict, Kmmes, Roth and Smith, and is summarized in Table (J4: 1 TABLE G4. RESPIRATORY QUOTIENT AND TOTAL METABOLISM OF NORMAL INDIVIDUALS AT REST AT A PERIOD TWELVE HOURS OR MORE AFTER THE LAST MEAL. Average Calorics respiratory per kilo per Individuals. quotient. twenty-four hrs. S!) men O.S3 2f).,5 (IS women O.S1 24.9 If the fast is prolonged the respiratory quotient will fall because the individual is forced to draw wholly upon fat and protein for 1 Benedict, Emmes, Roth and Smith: .lour. Biol. C'heni., 1914, xviii, p. 1,'iO. TOTAL METABOLISM IN DIABETES MELLITUS 117 nutritive material, and it will be remembered that these have a respiratory quotient of 0.71 and O.Xl respectively, in comparison with carbohydrate, which is 1 . The respiratory quotients of the individual studied at the Nutrition Laboratory 1 who fasted for thirty-one days, are shown in the following Tables Go and 06. TABLE 65. THE RESPIRATORY QUOTIENTS OF A MAN DURING A PROLONGED FAST. Calories per kilo, body Respiratory weight per Period. Time. quotient. 24 hours. Preliminary period Fourth day before fast 0.81 33 Third day before fast . 89 32 Second day before fast 0.89 29 First day before fast 0.82 27 Period of fast . . Days 1-5 of fast 0.77(av.) 29 Days 6-31 of fast 0.72(av.) 26 After period . . Second day after breaking fast 2 0.78 Third day after breaking fast 2 0.94 TABLE 06. QUANTITIES OF PROTEIX, CARBOHYDRATE AND FAT OXIDIZED BY FASTING MAN AT NUTRITION LABORATORY. 3 Quantities oxidized. Calories per Respiratory quotient. Carbo- kilo per Non- Protein, hydrate, Fat, twenty-four Period of fast. Actual. protein. gnis. gms. gins. hours. 1st day ,78 0.76 43 69 135 30 2d day . . . 75 0.74 50 42 142 30 3d day . . 74 0.74 68 39 130 29 4th day . . . 75 0.71 71 4 136 28 5th day . . . 0. 76 . 72 63 15 133 28 6th to .'Ust day av. 72 . 70 53 O 4 114 26 It will be seen that the respiratory quotient became approxi- mately that of fat within six days and remained there for the balance of the period; that it rose sharply within two days after breaking the fast, and upon the third day reached 0.94, showing that the individual must have taken an extraordinarily large quantity of carbohydrate food. The respiratory quotient differs little from that of normal in- dividuals in mild cases of diabetes when the urine is free from sugar and the carbohydrate in the diet large. The respiratory quotient of these same mild cases of diabetes will be lowered by fasting or by the withdrawal of carbohydrate, as shown above in the case of the normal fasting man. Evidence is thus afforded that the limited quantity of carbohydrate in the diet in cases of severe diabetes is responsible to a large degree for the low respi- ratory quotient which such patients exhibit. Magnus-Levy called 1 Benedict: A Study of Prolonged Fasting, Carnegie Inst. of Washington, No. 203. 2 Twelve hours after food. 3 Determined from the daily metabolism, the urinary nitrogen and the calculated non-protein respiratory quotient. 4 Actually a total of 32 gms. carb. were burned during the sixth to thirteenth day inclusive, and later none. 118 FACTOliX IX TREAT MK^T OF DIAfiKTKX MKLUTVS attention to this, and so have other observers. It, is well exempli- fied by the change in tlie respiratory quotient, of Case Xo. 71 1. This patient, with only moderate acidosis, became sugar-free upon April 1C), 1914, following fourteen days of treatment. On December )>, 11)1 1, she reentered the liospital with 1.1 ])cr cent, of sugar, but under fasting treatment became sugar-free after the omission of four meals. The respiratory quotient on successive days is shown in Table (>7: TAHLF. 07. ILLTSTHATKIN OF KALI- i\ RKSIMUATOKY QroTiF.vr OF Mn.n DlAHKTIC. 1 Case 71 1. I'Vmalr. Ajic. t hid y-oiti'ht years, nine niontlis. \\oitiht, f>:> kilos. f + + + + + + + + 10 25 L5 40 45 10 15 45 GO 7 15 .).) 100 (i !).-<. :; . . . 4 ;; . . .") (i . t; 7 . 7s. 10-11 . It will be seen that whereas the respiratory quotient was 0.7X on entrance, due undoubtedly to the oxidation of some of the carbo- hydrate ingested, though much at the same time was being lost in the urine, this rapidly decreased to 0.7-'! under fasting followed by a low carbohydrate diet. Yet. this woman could not be con- sidered a severe case of diabetes. The quotient was low simply because she was living almost exclusively upon a fat protein diet. The problem of drawing inferences from the respiratory quotient in diabetes is complicated by the fact that much of even the little carbohydrate which is given to a diabetic patient is lost in the urine. The patient really approaches the condition of the fasting man in that he is living exclusively on fat aim protein, although in this case not that of his own body. If all the carbohydrate ingested is lost in the urine, his respiratory quotient has been computed to be 0.72 upon the assumption that, b~> per cent, of the metabolism is due to protein and N"> per cent . to fat . P>ut there are other com- plications. Occasionally cases of diabetes are seen where the sugar in the urine exceeds that of the diet, and speculation at once arises as to the soum of this excess of sugar. Magnus-Levy has pointed out that if the sugar in the urine amounted to (10 grains and the protein in the diet to 100 grams, the additional quantity of oxygen which would be demanded lo form this amount of sugar out of protein would lower the respiratory quotient to 0.70. The situa TOTAL METABOLISM IN DIABETES MELLITUS 119 is still further complicated by the presence of unoxidized acid bodies in the urine, amounting frequently to 20 to 40 grains and occasionally to GO grains calculated as /3-oxybutyrie acid. The amount of oxygen consumed in the formation of these bodies for (8-oxybutyric is far richer in oxygen than are protein and fat would again lower the quotient, and it lias been calculated by Magnus-Levy 1 that the respiratory quotient of a case of diabetes presenting GO grams of sugar in the urine for 100 grams of protein in the diet, and excreting 20 grams of /3-oxybutyric acid, would fall as low as ().()(). For convenience, these figures are summarized: The respiratory quotient of the fasting man at the Nutrition Laboratory was 0.72. The calculated respiratory quotient of a normal individual who is burning 15 per cent, protein and 85 per cent, fat is also 0.72. The theoretical respiratory quotient of a diabetic individual excreting all the carbohydrate ingested, as well as GO grams of glucose for each 100 grams of protein in the diet, is 0.70. The theoretical respiratory quotient of the diabetic individual excreting GO grams glucose for 100 grams protein eaten and 20 grams /3-oxybutyric acid as well, is O.G9. These calculations presuppose that the sugar and /3-oxybutyric acid excreted were formed during the same twenty-four hours, but who knows whether this is the case? The theoretical non-protein respiratory quotient of a case of diabetes living upon fat and the non-carbohydrate part of the protein molecule, as calculated by Lusk, is also O.G9. TABLE OS. THEORETICAL KESI>IRATORY QUOTIENTS (FROM MAGNUS-LEVY). Respiratory Diet. Protein, 100 gin. / 100 X 4.1 == 410 \ Carl)., f>07 gin. \ .W X 4.1 = 282.") / Protein, 100 gin. f 100 X 4.1 == 410 ) Fat, 2">0 gni. \ 250 X 0.3 = 2325 J Loss in urine Sugar, 00 gin. (.00 X 4.1 = 240) Sugar, 00 gin. / 00 X 4.1 = 240 \ B-oxy. acid, 20 gin. \ 20 X 4.7 == 94 / Table GS shows the theoretical respiratory quotient, which should be reached under varying conditions of diet for a normal individual, and the changes which theoretically are present under special conditions in diabetes. Figures of this type have dominated the discussion of the metabolism in diabetes from the start, and whenever experiments have not produced figures comparable with 1 Magnus-Levy: Ztschr. f. klin. Mod., 1905, Ivi, p. S3. Diet. Caloric.- 2735 120 FACTORS I.\ TREATMEXT OF DIABETES MELLITUS these they have often been considered erroneous. We are taught to believe that diabetic patients are not severe unless the respiratory quotient is ().C>9. It is questionable, however, whether the experi- mental data at our disposal enable us to say that our theories are backed up by the results which we obtain. If one looks over the lists of respiratory quotients obtained in successive periods with any variety of respiratory apparatus or calorimeter, he will be shocked at the discrepancy and is forced to the belief that any argument based on a change in the respiratory quotient of one point is unjustifiable, and any argument which is based on a change in the respiratory quotient of two points really rests on a very slender thread. A change of three points is, however, deserving of con- sideration, but modesty should rule in conclusions which are to be drawn from any given set of experiments. The average respiratory quotients in experiments with the severe diabetics without food which were studied by Benedict and myself during 190S to 1912 are given in Table (>9, and for comparison other figures are also added which show the respiratory quotients obtained by various other observers: TABLE 09. RESPIRATORY QUOTIENT IN SEVERE DIABETES. Respiratory Year. Observers. Cases, quotient. 1S94 Weintraud and Laves: Ztsehr. f. physiol. 1 0.70 ('hem., 1S94, xix, p. (Mi. 1S97 Nehring-Schnioll: Ztst-hr. f. klin. Mecl., 1S97, 2 0.72 xiii, p. f>9. !!()") Magnus-Levy: Ztsclir. f. klin. Mod., l'.M)r>, 2 0.71 Ivi, p. Mi. 1!)07 Mohr: Ztsclir. f. exper. Path. u. Thorap., 1 0.72 l'.i!)7, iv, j). 910. 190s 1911 Benedict arid Josliii: ( 'ariie-rio I list, of Wash- 19 0.73 iiurton, Publications I'M and 17(1, 1910, 1912. 1912 Holly: Doutseh. Arch. f. klin. Mod., 1912, S 0.74 cv. p. 494. 1912 Grafo ami Wolf: Deutsch. Arch. f. klin. 3 0.74 Mod., 1912. cvii, p. 201. 1912 1914 Benedict and .Io>lin 7 0.73 The uniformity of the respiratory quotients of the severe diabetics studied prior to 1914 would lead one to believe that the ordinary respiratory quotient of severe diabetes was ().7o. To our surprise, in November, 191."), with one of our first severe diabetics which underwent fasting treatment we noted that with the continuance of the fast the respiratory quotient had a tendency to reach the neighborhood ot'o.so. Experiments at the Russell Sage Laboratory, to which reference has already been made, conducted by Dr. DuBois upon one of Dr. Allen's private patients, suggested a similar TOTAL METABOLISM IX DIABETES MELLITI'S 121 condition. In other words, whereas the normal individual during a fast exhibits a respiratory quotient based upon the combustion of protein and fat alone, the severe diabetic during fasting showed a respiratory quotient which could only be accounted for by the combustion of notable quantities of material other than fat and protein. That this material was not protein became evident, because the amounts of nitrogen in the urine excreted during these periods were not abnormal. An explanation why the severe dia- betic shows no acidosis during a fast, in contradistinction to the normal individual, is suggested by this increase in the respiratory quotient. Several explanations for this increase in the respiratory quotient of fasting diabetics are available. During fasting the diabetic may be able to draw upon sources of carbohydrate in the body which the normal individual cannot. Furthermore, the diabetic has in the body undoubtedly more carbohydrate stored than we have hitherto supposed, and the supposition must be entertained that the diabetic may actually have more carbohydrate in some form in the body than exists in the normal individual. A third supposi- tion for the increase in the respiratory quotient is that considerable quantities of acid bodies have accumulated and that with the improvement of the condition of the patient during fasting these are burned. It will be remembered that /3-oxybutyric, diacetic acid and acetone all have relatively high respiratory quotients, namely: O.S9, 1.00 and 0.75 respectively, and therefore the oxida- tion of a small quantity of these substances would markedly raise the respiratory quotient. Which of these suppositions is correct will be eventually known because of the improved methods of estimating carbohydrate and acid bodies in the blood, fluids and tissues of the body, and also by the help which is afforded from the estimation of the carbon dioxide tension of the blood. If the rise in respiratory quotient is due to oxidation of acid bodies one would expect that in the course of the improvement of the severe diabetic, while still on a protein-fat diet, the respiratory quotient would again fall. 3. Observations upon the Pulse. Surprisingly few statements are made in the literature upon the pulse-rate of diabetic individuals. In the earlier series of experiments conducted by Benedict and myself a cursory examination of the pulse-rate indicated no marked difference between that of normal individuals and of diabetics. In our second series of experiments, conducted during 1911-1912, however, we found, upon summarizing our observations upon the pulse-rate, that it was distinctly elevated. 1 1 Carnegie Pub., No. 176, p. 86. 122 FACTORS I.\ TREATMEXT OF DIABETES MELLITU8 The average maximum and minimum pulse-rates for t'ne normal subjects used for comparison were 7f and ."H respectively, while for the diabetics the average maximum was SI and the average miniinnm (>"> beats per minute. In preparing these tables care was taken not to include any records obviously affected by excitement or other untoward incident in connection with the experiment, three records approximately constant, usually being taken as indicat- ing the maximum or minimum. The values given can therefore be considered as representing the average normal variations in pulse- rate to be expected for normal individuals under the conditions of experimenting as employed in the Nutrition Laboratory, and for patients with severe 1 diabetes. It will be seen, therefore, that both the minimum and maximum pulse-rates of normal individuals are markedly lower than those of diabetics. TAHI.K 70. PCI.SK-IIATK or DIAHKTIC Srn.iK(Ts. I ) . 52 (': . . 09 II . . 64 I . . . 117 J , . . 00 K . . 7:? L . . .->4 M 01 \imiim. Subject. Minii) KG x - 72 o . . >i S4 p . . f>- 5S Q <>] 74 H . . 6J 90 s 5^ 122 T . . 6: S4 r . . - S3 v . . . . (>: 73 \Y , . 9( M) X 7- SI It. must furthermore be remembered that in practically every instance the normal individual led a much more active life, had much greater muscular tune, and was more restless than were the diabetics. This makes the difference in pulse-rate all the more TOTAL METABOLISM IX DIABETES MELLITCS 123 striking, since one would ordinarily expect a somewhat higher value with the more active and restless normal subjects than with, the weak, sick diabetics, disinclined t:> any extraneous muscular motion. An examination of Tables 70 and 71 shows no regularity in either the minimum or maximum pulse-rates when individual diabetics are compared with normal individuals, since some of the diabetic subjects show a minimum pulse-rate considerably lower than the average" minimum pulse-rate of the normal subjects; and similarly, the normal individuals in some instances show a pulse-rate con- siderably higher than the average maximum pulse-rates of the diabetics. On the whole, however, a distinct tendency toward an increased pulse-rate in diabetics is here clearly shown. This increase is fully in conformity with the increased metabolism which was noted with these severe diabetics. Yet in this discussion of the pulse-rate it should be stated that it is questionable whether suffi- cient information with regard to normal pulse-rate has accumulated to indicate that a variation of 15 per cent, above normal could be recognized. While the metabolism may be 15 per cent, above normal and be accurately measured, since normal metabolism is fairly well established, a 15 per cent, increase in the pulse-rate could not be so easily identified. The intimate relationship between pulse-rate and metabolism in resting subjects is evident in nearly every new publication on metabolism and shows that today the pulse-rate is taking on a new significance; it is not at all impossible that the degree of aeidosis and the degree of metabolism may be at least approximately estimated in diabetics by a careful examination of the pulse-rate. The attention of the physician i.s directed to the importance of the pulse-rate in diabetes and particularly to the value of changes in the pulse-rate in the same patient during the progress of the disease. The practice first established by Prout of reporting the pulse-rate with the chemical data of each experiment is recommended to all experimenters. It is hoped that physicians will study this question further. Of course a few observations upon the pulse-rate during the twenty-four hours would be of little value, because these are so easily affected by extraneous causes. To be of value the observations should be at frequent intervals of the day and several counts should be taken in succession. For example, it was not uncommon in our experiments for the pulse to be counted forty times in a single experiment. The count was made 1 with the use of a stethoscope which the patient wore during the whole period, and the observations were made without the patient knowing when they were being obtained. 4. Observations upon the Weight. Edema. -Perhaps no one gross observation made during the course of diabetes mcllitus is of greater significance and causes greater alarm, both to patient and 124 FACTORS 7.V TREATMENT OF DIABETES MELLITUK to the physician, than the persistent loss in body weight. On the other hand, slight changes in body weight which may accompany dietetic alterations or the ingestion of sodium chloride and sodium bicarbonate are looked upon as material gains and are thus liable to be misunderstood by the patient. To interpret intelligently these changes, it is necessary both for the physician and for the patient to realize the factors affecting the body weight of normal as well as pathological cases. Few realize that the normal individual is continually undergoing changes in body weight throughout the twenty-four hours. Even during sleep it has been shown that a man of 85 kilos loses 30 grams per hour, and a woman of 05 kilos 20 grams per hour. With exercise this is, of course, greatly increased, and may amount to 0.4 kilos for a foot-ball player during one hour and fifteen minutes of active exercise. Edema is a common source of error and it is important to recognize it as a cause of gain in weight in diabetes. Patients may seem to be gaining when in reality they are losing weight because of insufficient diet. The edema occurred most frequently in former years following oatmeal clays and the administration of alkalis, but now is common with fasting diabetics of severe type and is apparently related to the large quantity of salt which they ingest with broths and vegetables. The edema may become extreme and one of my patients (Case Xo. 922), whom 1 had not seen for months, called in a laryngologist and barely escaped tracheotomy for edema of the larynx. This quickly disappeared with the omission of salt and a diet of water and a few oranges. The patient later entered the hospital, became sugar-free and developed a tolerance for 49 grams carbohydrate, 09 grams protein and 143 grams fat. Diabetic patients should be weighed, preferably naked, before breakfast, and after the urine has been voided, for a patient fre- quently voids a pound at a time. The average weight of clothes of men and women falls between 8 and 10 pounds. Women's clothes weigh approximately 2 pounds less than the clothes of men. The table of height and weight for normal men and women with clothing and for diabetic patients will be found on page 55, and on page 50 is given the height and weight of children between the years of one and fifteen. (a) Water Content of the Body. A factor which should be taken into consideration in interpreting changes in body weight is the fluctua- tion in the water content of the body. It should be realized that the average man at rest without food oxidizes per day about 75 grains of protein, 25 grains of glycogen and 200 grams of fat a total of 300 grams of water-free, organized body tissue. It can readily be seen, therefore, that with the subjects at rest, large and rapid changes in weight must be due not to the oxidation of organic TOTAL METABOLISM IN DIABETES MELLITUS 125 material, which amounts to only 300 grams per day, but to large excretions of water. Under certain conditions it is possible for the body to retain considerable quantities of water, and conversely, to be deprived of considerable amounts of water that would normally be retained. Since about 00 per cent, of the body is water, any change of water content may result in material gains or losses in body weight. A man, weighing, for example, 05 kilos, may have an absolute water content of 39 kilos, so that a relatively small change in the percentage of water in the body may produce a change in body weight of 1 kilo. (6) Influence of Fat and Carbohydrate Diets upon Weight. Remark- able changes in the weight of normal individuals will also occur, if the proportion of fat to carbohydrate is altered, although the caloric value of the diet remains constant. A diet rich in carbohydrate brings about an increase in weight, whereas a diet of exactly the same number of calories, though chiefly made up of fat, lowers the weight. These changes undoubtedly are due simply to the retention of water by the tissues upon a carbohydrate diet and loss of water upon a fat diet. Such changes appear reasonable because the storage of 1 gram carbohydrate in the body demands the retention of 3 grams of water, 1 1 gram of protein would appear to require the same amount, and 1 gram of fat requires only 0.1 gram of water. These changes are well illustrated by the following table : TABLE 72. CHANGES IN WEIGHT WITH FAT AND CARBOHYDRATE DIETS. CARBOHYDRATE DiET. 2 Food and drink. Body Gain ( +) Solid matter, Water, Total weight, or loss ( ), Date, 1904. gms. gms. gms. kg. gms. Apr. 16 . . .. 75.086 16-17 . . 970 3577 4547 75.443 +357 17-18 . . 966 3553 4519 75.414 - 29 18-19 . . 966 3491 4457 75.269 -145 FAT DIET. Apr. 19-20 . . 750 3108 3859 74.319 -950 20-21 . . 745 4150 4896 73.480 -839 21-22 . . 747 4152 4899 72.528 -952 Average gain per day, carbohydrate diet, +61 gms. Average loss per day, fat diet, 914 gins. Water stored per day, carbohydrate period, +165 gms. Water lost per day, fat period, 906 gms. It is important to bear in mind the effect upon weight which must occur when a carbohydrate-free diet is prescribed, for other- wise the loss of the few pounds which is bound to ensue might cause undue apprehension or be interpreted as loss of tissue. 'Zuntz: Biochem. Ztschr., 1912, xliv, p. 290. 2 Carnegie Pub. No. 176, p. 93. 126 I' ACTORS J.\ TREAT. ME XT OP DIABETES MELLITUS An increase in weight following a marked increase of carbohydrate in the diet is strikingly illustrated in severe diabetic patients under the oatmeal treatment. I nder these conditions the weight may rise 4..") kilos during one or two days. It is not uncommon to observe that edema develops during the course of an oatmeal cure. It is significant that some of these cases show little or no carbohydrate in the urine. There will probably be general agreement to the statement that the gain in weight following the sudden introduction of large quantities of carbohydrate is to be explained by the storage temporarily, perhaps, of carbohydrate in the body, and along with this, as has been pointed out, marked quantities of water will be retained. That this storage or delay of excretion is accentuated in the case of the presence of diseased kidneys is common knowledge. Barrenscheen 1 showed that excretion of lactose was delayed upon the day following an oatmeal cure. ( c] Influence of Sodium Chloride upon Weight. The quantity of salt in the diet also affects the weight. For example, in the study of an individual upon a salt-free diet consisting of the whites of IS eggs (210 calories), 120 grams olive oil (10SO caloric's) and 200 grams crystallized sugar (SOO calories), total 2090 calories, or 30 calories per kilogram body weight, the weight fell from 70.2 kilograms on the first day to (ik ( .) kilograms on the thirteenth day, as will be seen from the following table: TABLE 73. Loss OF WKK;HT ('OIXCIDKNT \vrni A SALT-FREE PIF.T.- Day. 1 . . 470 1720 1012 .29 4 . 00 2 . . 550 1810 1010 .29 2 . 52 3 . . 500 1430 1012 1 .28 1 .SS 4 . . 290 930 1017 1 .20 O.S7 ,j 290 1100 1013 .43 0.09 . . 515 1170 1012 1 .04 0.48 7 . . 200 S50 1015 1 . 15 0.40 S . 125 1000 1013 ( ).7S 0.40 9 . 290 1100 1011 ( ).95 0.20 . . 200 800 1015 ( ).S9 . 22 1 . . 200 050 101S ( ),70 0.22 2 2 1 5 510 1023 ' ( ).79 0.17 3 170 500 102:; ( ).80 0.17 ro.2 The subject was then put upon a free diet and three days later the weight had risen to OH kilograms. ! Barri'ii>chciMi: Hiorhom. Zt.-rhr.. lit 12. \\xix. p. 4 ">'.). - Guodall and ,Jn~lin: Experiments with Ash-free Diet, Arch. Int. Med., 190S, i, p.Glo. TOTAL METABOLISM IN DIABETES MELLITUS 127 Tf such notable changes in weight occur with normal individuals, it is evident that one must be extremely cautious in interpreting changes which occur in the weight of diabetic patients. In illus- tration I would cite the gain in weight which sometimes takes place during fasting, due, undoubtedly, to the retention of water, and as 1 was able to demonstrate in Case No. 5 18, this was brought about by the patient having taken a mineral water containing considerable quantities of salt. (See p. lioo.) None of my other patients have shown such extensive variations in weight in so short a time as did this individual. (See also Case No. "Go, p. :>14.) (d) Influence of Sodium Bicarbonate upon Weight. The administra- tion of sodium bicarbonate is frequently followed by a gain in weight. Thus, in Case Xo. 220, the changes in weight during the adminis- tration of sodium bicarbonate were as follows: TABLE 74.- Date. Nov. -GAIN ix WEIGHT COINCIDENT WITH SODIUM BICARBONATE. ADMINISTRATION OF Sodium bicarbonate gms. Body weight, kilos. Date. Sodium bicarbonate gms. Body weisiht, kg. . IS . 1 Nov. 7 . . . 20 50 . 7 . 48.6 8 . . . 20 51.5 . 40.0 9 . . . 20 52.4 . 4S . G 10 . . . 20 53 . 3 . 20 49 . 3 11 . . .. 20 53 . 3 Iii order to show that this gain in weight was not directly due to the alkali, but rather to retention of salt, the weights of another diabetic patient, Case Xo. 135, were taken while upon a salt-free diet. TABLE 75. ABSENCE OF GAIN ix WEKIIIT COINCIDENT WITH ADMINIS- TRATION OF SODIUM BICARBONATE WHEN THE DIET is SALT-FREE. CASE Xo. 135. Diet. Salt-free. Trine. -/ 5 = 7; '5 ;. tt ;_ M : ~ ^ ^ ~ c~* ^ >> - f ~ ^ *- 1 - w K tl _/ : ^ -r i ti | -r tt "~ '-' -^ 5 s =' C3 -j " ti *- *j 3" '<-! _, ~ w -X c ~ ^-- Q ^ - - i * - > * X < cb. - o -/. Jan. 26 I 135 110 185 3500 3720 21.8 4.2 7.9i 20 4.4 8.2 160 27 135 110 185 3500 3940 19.6 4.3 7 . 8 29 4.5 6.3 165 2s 135 110 185 3500 3210 20 . 5 4.4 7.3 24 4.6 5.9 160 29 135 90 155 3500 3210 19.2 4.. 1 7.3 26 4.2 4.8 163 30 25 135 ! 70 185 3500 3190 10.3 3.5 8.7 33 4.1 1.6 146 31 25 120' 60 95 23 5370 4600 19.1 4.3 12.6 51 5.1 2.3 146 Feb. 1 37 130 100 130 45 5250 4050 is. 7 3.3 10.7 39 4.3 2.0 137 o 52 70 60 95 45 5370 3510 16.0 3.5 10.2 37 3.9 2.1 121 3 . . 15 . 15 30 45 800 360 15.0 .. 86 128 FACTORS IN TREATMENT OF DIABETES MELLITUS It will be seen that while upon, the salt-free diet the weight steadily fell and, despite the administration of sodium bicarbonate later, no increase in weight occurred. This observation has been confirmed by Levison. 1 The explanation of the usual gain in weight of diabetic; patients following the use of sodium bicarbonate was pointed out by Cioodall and Joslin 2 some years ago. Apparently the adminis- tration of sodium bicarbonate, by favoring the excretion of large quantities of retained acid bodies, leads to irritation of the kidneys, resulting in their inability to excrete salt in the normal manner. If the salt in the diet is restricted, there is less to be retained and consequently no gain in weight results. Fischer 3 explains the phenomena in another way: "Sodium chloride retention is not due to an inability of the kidneys to eliminate it, but to a change in the proteins (and other colloids) of the body as a whole. Sodium chloride retention does not lead to edema, but the changes which lead to edema and to sodium chloride retention are the same, consisting, in the main, of an abnormal production and accumulation of acid in the body." I might here make the clinical observation that a salt-free diet in diabetes is inadvisable. It is noteworthy that patients during the period of coma markedly lose weight. Edema, which may be present just prior to coma, disappears during coma; in fact, I remember to have seen but one patient in coma who showed edema. The severe diabetic during coma utilizes apparently all possible liquid in the tissues to aid in the excretion of toxic bodies. The importance of maintaining sufficient fluid in the body deserves emphasis. Hodgson 1 has thoroughly appreciated it for years and endeavored to make his patients retain large quantities of fluid. These remarkable changes in weight suggest several lines for investigation: (1) Does the mere retention of fluid in the body lead to the retention of carbohydrate? (2) Does the retention of fluid in the body lead to a better utilization of carbohydrate? ('>) Is the reverse of these two propositions true? (4) Might it not be possible by a series of experiments designed to increase the content of water in the body to throw light upon the capacity of the body to store carbohydrate? (">) May not rapid losses in the weight of severe diabetic patients be an early guide to approaching coma? (e} Weights and Losses of Weight in Diabetic Patients.- The percent- age of loss in body weight of some 200 of my cases was computed for me by the Nutrition Laboratory. The average loss previous to the first visit was 10 per cent., and between this time and the last 1 Levison: Jour. Am. Mod. Assn., 191"), Ixiv, p. .'ii'O. 'Ibid.: 190s, li, p. 71' 7. 3 Ihid.: 191.J, Ixiv, p. 325. 'Hodgson: Jour. Am. Mod. Assn., 1911, Ivii, p. 1LS7. TOTAL METABOLISM IN DIABETES MELLITUS 129 observation a further loss of 3 per cent, occurred. This is good evidence that these cases came almost uniformly late for treatment. In this series of cases the highest reported weight of a patient was 118.5 kilograms and the lowest 13.4 kilograms. It is most important that patients should understand the signifi- cance of changes in their body weight, and should not be too much depressed or elated by losses or gains in the same. It should be explained to them that loss of weight by design during the beginning of treatment is very different from a loss of weight which occurs through ignorance while the disease is progressing unfavorably. (/) The Loss of Weight Prior to and during Coma. The loss of weight of patients immediately prior to and during coma has always appeared to me to be great, but from the nature of the case it has been difficult to determine this point. A hint of this was afforded by noting the marked loss in weight which occurred in Case No. 135, from whose diet salt was excluded, but a still better example was that of Case No. 513, already referred to on page 127, who lost 35 pounds in the eleven days preceding the third day before death in coma. Further observations of this character should be made for I doubt not that they will show surprisingly large losses of weight by the diabetic individual preceding death by coma. Too much emphasis, however, must not be laid upon these data until they are compared with the data of non-diabetics dying from a variety of causes. This apparent loss of weight during coma has always interested me deeply. Obviously it is due to a desiccation of the body, and in conformity to it can be placed my experience of not having seen a patient who has edema develop coma. (() grams nitrogen are obtained in contrast to half this quantity for the normal individual, ruder modern treatment the nitrogen excretion of a diabetic patient falls to the neighbor- hood of 4 FACTORS IX TREATMENT OF DIABETES MELLITl'S convicted of larceny, if not of perjury. The reputation of these workers will allow no such interpretation. Xo one believes today that, the human 1):X ratio exceeds .'>.(>."): 1. Higher ratios are illustrations of irregularities in excretion of the two substances. That 4 cases should have recovered with such ratios is confirma- tory of this conclusion. X'ew light upon the ]):X ratio has recently been offered by Janney, 1 who has worked out an exact method for the determina- tion of the formation of glucose in the animal body. By feeding various forms of animal protein to fasting phlorhizinized dogs, he demonstrated that f)8 per cent, of these proteins went over into glucose. By these direct experiments and other proofs which cannot be entered into here, he concludes that the actual D:X ratio is o.4 : 1. If one would determine the dextrose-nitrogen ratio accurately in patients, the following requisites should be fulfilled: (1) an exclusive fat protein diet or fasting; (1 ) ) surroundings which make errors in diet impossible; (8) a period of observation of at least seven days to exclude the washing out of stored carbohydrate; (4) a constant (not falling) I) : X ratio of .'!.('>."): 1 for the last three of the seven days, and (">) several daily blood-sugar determinations to furnish some proof, inadequate though it be, that the sugar in the urine has not come from that left over in the blood. (i. The Storage of Carbohydrates in Diabetes. It is well known that following a period of fasting large quantities of carbohydrate can be administered without immediately appearing in the urine. The best illustration of this is von Xoorden's oatmeal treatment. Thus Case Xo. 'U4 (see p. 410) showed a positive carbohydrate balance of .li'l) grains when undergoing an oatmeal cure under the direction of Professor von Xoorden, although he never after this cure became sugar-free, despite a rigorous diet, save for occasional days. A more spectacular demonstration is the severe diabetic of Klemperer- who took 100 grams of glucose in divided portions during twenty-four hours without its appearing in the urine. Almost as impressive is that of Case Xo. 7S.~), a boy, aged seventeen years, who came to me in the twentieth month of the disease. 3 By con- sulting Table 79 il will be seen that only 7 grams of sugar appeared in the urine following an intake at one time of SI grams levulose, although by observations before and after the tolerance was known to be nearly ////. I low large a quantity of carbohydrate it is possible for the body to store is really unknown, but its importance is none the less 1 Janney: AIM. Jour. Med. S<>. 1917, .-liii. p. 14. '-' Kleuiporor: Therapip dor (ienemvart, I'.Ul. In. p. 447. ;i For other examples see p. 3<8. TOTAL METABOLISM IX DIABETES MELLITUS 135 evident. Unless the amount of stored carbohydrate is known it is unjustifiable to say that the carbohydrate excreted represents a part of that ingested during the same twenty-four hours. All data with reference to the I):X ratio are confused by our ignorance of stored carbohydrate. The influence of carbohydrate so stored in the body upon carbohydrate assimilation or retention, but not necessarily utilization, is also great. Whatever virtue the oatmeal cure possesses, all agree that it depends in major part upon preceding starvation, which has tended to exhaust the carbohydrate depots of the bodv. TABLE 79 Case Xo. 785. Mule, aged seventeen years. Output. Intake. Days Vol., Diac., SuKar, Nitrogen Annn onia. Carl)., Protoil , Fat, Alcohol Calories. c.c. Kins. >;nis. K"is. gl IS. Kins. Kins. gins. Kins. 1-7 1080 1 + II 2 T.s 17 58 127 9 1506 s 960 + 7 0.5 0. 69 90 3 21 30 3 735 9 390 ++ "> 4.3 o. 35 20 63 110 9 1385 10 1175 + 3 8.4 0. 74 20 03 1 1 9 13S5 ((/) Storage of Carbohydrate as Glycogen. ( 'arbohydrate is stored in the body in various ways, but most of it is supposed to be in the form of glycogen, and this is about equally divided between the liver and muscles. An old estimate of Bunge that the body con- tained 400 grams carbohydrate, is roughly approximated by experi- ments upon fasting men and professional athletes doing severe work without food. This figure may be taken as a fair average, but there are enormous variations. This statement is based upon glycogen which has been shown to be burned in calorimetric experi- ments; it does not exclude the possibility of some glycogen still remaining in the body. Experiments on fasting men show that they may burn from 93 to 232 grams of glycogen in the first three days of a fast. 4 In diabetic patients the quantity of glycogen is universally considered to be far below this amount, but Frerichs 5 found, upon puncturing the liver of two diabetics, a small amount of glycogen in one and a considerable amount in the other, and Kiilx found 4 Benedict: The Influence of It 1907, Pub. 77, p. 404; Benedict: Washington, 1915, Pub. 203, p. 2 5 Frerichs: ('it. Xehring and Si Pfiugcr's Archiv, 1870, xiii, p. 207. inition on Metabolism, Carnegie lust, of Wash., V Study of Prolonged Fasting, Carnegie Inst. of 51. hmoll: Ztschr. f. klin. Men 1 ., Is97, xxxi, p. 59; 136 FACTORS IX TREATMENT OF DIABETES MELLITUX 10 to 12 grains glycogen in the liver of a diabetic who had been foi a long time on a diabetic diet. Examinations of the tissue removed from the livers of living diabetic patients also show appreciable quantities of glycogen, and it is the experience of pathologists that the organs of diabetic patients contain more than traces of glycogen. It is most unfortunate that no data exist which enable us to deter- mine what this minimum is. It is quite conceivable that although it might be extremely small at any one moment, a small quantity might be continuously formed and destroyed, and the sum of these small quantities reach a substantial amount in twenty-four hours. The recent work of Kelly 1 throws new light upon the problem, lie points out the striking contrast between the constant presence of glycogen in the liver of human diabetes and the very small quantity which is found in the severe diabetes of depancreatized dogs, yet even in the latter the power of the liver to form or deposit glycogen is shown when levulose is administered. If a milder form of diabetes is produced in the dog more glycogen remains in tin- body and there is a closer resemblance to human diabetes. Whereas with total removal of the pancreas there was only 0. 00") per cent, of glycogen in the liver; with partial removal there was 0.3 per cent, of glycogen, even though S to 10 per cent, of sugar remained in the urine. By microscopic examination so considerable a quantity as this appeared small. Woodyatt 2 found that the muscle of a case of severe human diabetes formed even less lactic acid than that of fully phlorhizinized dogs. This suggests an impaired power to dissociate glucose on the part of the diabetic muscle, since such muscles are bathed with an abnormally high quantity of sugar which if available should yield more lactic acid than is found in the muscle of phlorhi/in diabetes. Kleiner 3 finds in agreement with other observers that large amounts of sugar injected intravenously into a normal dog disappear from the circulating blood in about ninety minutes after the end of the injection and only about 00 per cent, are excreted in the urine. The same quantity will leave the blood in nephrec- tomi/ed animals. This occurs even after intravenous injection into dead animals. An increase in carbohydrate of the muscle tissue was found and there was also evidence of the formation of polysaccharides in the muscle tissue. (h) Storage of Carbohydrate as Blood Sugar. Sugar is also stored in the body in the form of blood sugar. The normal quantity of sugar in the blood of healthy individuals varies between 0.07 and 0.1 1 per cent., and for convenience in calculation may be considered 1 H.'lly: /tschr. f. exp. Path. a. Therap., 1914, xv, p. 464. - Woodyatt: Jour. Mini. Chcin., I'll:*!, xiv, p. 411. 3 Kleiner: Jour. Kxp. Med., I'JIO, xxiii, p. "()7. TOTAL METABOLISM IN DIABETES MELLITUS 137 0.1 per cent. This rises quickly after a meal rich in carbohydrates, but soon falls to its former level. Of 191 observations upon 72 of our diabetic patients, the percentage of blood sugar varied from 0.45 to 0.07 per cent. 1 But the blood of these diabetic patients does not behave like that of normal individuals following the ingestion of food. It is true that the percentage of sugar rapidly increases following a carbohydrate meal, but it does not as rapidly fall, and in my own experience most diabetic patients, even after prolonged fasting, show values for blood sugar which are far above normal. Certain types of diabetic patients namely, those with disease of the kidneys are expecially prone to maintain high percentages of sugar in the blood for many days after their urines have become sugar-free. It is impracticable to consider that the percentage of blood sugar is maintained independently of the other tissues in the body because: (1) the percentage is so unstable; (2) there is no constant relation between the sugar in the blood serum and the sugar in the total blood ; and (3) because the capacity of the blood for storage of sugar is so slight. Confirmation of the above statements has been afforded by the ingenious experiments of Woodyatt and his co-workers to be later described (see p. 255). If we assume an individual of 70 kilograms body weight and con- sider that 7 per cent, of his weight is made up of blood, we have 4.9 kilos of blood of which the sugar content is 0.1 per cent. This would amount to 4.9 grams, even taking the highest for the normal individual, and should we take the highest figures we have encoun- tered even after the administration of food with our recent diabetic patients, namely, 0.45 per cent., the total quantity of sugar stored in the blood would not be far from 22 grams, a trifle more than two-thirds of an ounce. Falta 2 has called attention to the slow return of the blood sugar of diabetic patients after a carbohydrate meal to its former sugar level. Kleiner and Meltzer 3 have also beautifully shown this same difference to exist between normal and depancreatized dogs. Whereas the sugar in the blood of normal dogs increases fourfold, namely, from 0.20 per cent, to 0.79 per cent., following the injection of 4 grams of dextrose per kilogram body weight, and of depancreatized dogs threefold, from 0.38 per cent., before to 1.19 per cent, after the injection, the blood sugar of the former returned nearly to normal at the end of an hour and a half, while diabetic dogs then showed O.SG per cent. It is significant that in these experiments the quantities of sugar excreted in the urine were practically the same. Interesting as these figures are from this point of view, they are still more interesting from another 1 In a later case, Xo. 1015, it amounted to 1.37 per cent. See p. 88. 2 Falta: Med. Klinik, 1914, x, p. 9. 3 Kleiner and Meltzer: Proc. Soc. for Exp. Biol. and Med., 1914, xii, p. 58. i\ TUEATMKXT or DIAHKTKX MELI.ITCS point of view. It is impossible to account for all the sugar ingested by adding together the sugar found in the blood and that in the urine, \\here did the sugar go? One may say it, was burned, and this possibility, though not probability, must be admitted in the normal animal, but no one would contend this to be wholly the case in the depancreati/ed animals. Here again \\oodyatts work oilers new methods which we all hope will throw light upon this problem. At the Nutrition Laboratory we have been able to carry those experiments to their logical conclusion, for we have had the oppor- tunity to determine the respiratory quotient following the adminis- tration of levulose to severe diabetic's. In Table Si will be found a report of some of these experiments upon the effect of levulose when administered to severe diabetic patients m amounts of 2..">1 grams, 2.42 grams and 1 .'.).") grams per kilogram body weight. In the first and third cases there was no increase in the respiratory quotient. A considerable portion of the levulose was probably excreted in the first case, but in the others little or none. The explanation of this difference in behavior in the storage of levulose among the three patients appears to be that the first cast 1 had not fasted beforehand and that the third had been on a lo\v carbohydrate diet for a long- time; this is confirmed by the second case, which also excreted little of the levulose, for it was administered following a period of strict, dieting. At the present time Benedict and I are compiling our experiments with levulose. These: are extensive, and as yet our conclusions have not been drawn. They will be reported shortly in a monograph to be issued by the Carnegie Institution of Wash- ington. (<") Influence of Rate at which Carbohydrate is Administered.- it should also be recorded that in all these .'! cases the levulose was given at one time and not in divided doses throughout the twenty- four hours, as in Klemperer's test, and yet the carbohydrate was stored by the body in two approximately as well. This gives added emphasis to the possibility of the presence of an empty storehouse for carbohydrate in the body. I also have evidence that the gradual administration of carbohydrate is of little value, provided the body is not prepared to retain it. Following etherization, a patient. Case Xo. SOS, while fasting for the first twenty-four hours, was sugar-free, but on the next day, though only 2 grains carbohydrate per hour were administered, he excreted practically all of it, though formerly his tolerance amounted to .">() grams carbohydrate. Com- pare the behavior of Case Xo. C> ( .>7 after two operations, when different anesthetics were employed. (See p. 112.) It is not strange that carbohydrate is probably better tolerated at breakfast than at anv other time in the dav. TOTAL METABOLISM IX DIABETES MELLITUX 139 (d) Other Possible Storehouses for Carbohydrate. The small amount of glycogen in the body and the still smaller quantity of blood sugar represent an amount of carbohydrate far too lo\v to account for the phenomena above described in diabetes. Other sources for storage of sugar in the body must be sought, and this has been emphasized by Ivar Bang. 1 If we should assume that the percentage of sugar was the same for all the fluids in the body asin the blood, certain amounts of sugar might be stored in this manner, ^hile such an assumption is not wholly justifiable, it has some basis, for we know that sugar exists in the spinal fluid of diabetics as well as in other fluids. In normal individuals Dr. Jacobson tells me that he has not found it so closely to follow the blood but the opposite was true in his cases of diabetes mellitus. Hopkins 2 has observed in diabetic patients that the reducing substances in the spinal fluid were but slightly less than the blood sugar of the same individuals. These varied from a minimum of 0.074 to a maximum of 0.623 for the spinal fluid and the corresponding values for the blood were 0.077 and 0.600. Notable percentages of sugar, not very different from those in the blood, have been found in pleuritic and ascitic fluids, and Husband 3 found even 0.7 per cent, in theamniotic fluid. Yet granted that the assumption of uniform distribution of sugar in the fluids of the body is correct, we cannot increase our storage capacity very much in this way. For example, assuming the total quantity of fluid in the body as 00 per cent, of the body weight of 70 kilograms, we have 42 kilograms of body fluid, from which we must deduct 4.0 kilograms already reckoned as blood. This leaves us a remainder of 37.1 kilograms of fluid in the body, and using the highest figure (0.45 per cent.) for blood sugar which we have personally encountered, the quantity of sugar in this mass of fluid would be only 107 grams. This is not enough relatively to explain Kleiner and Meltzer's experiment and far less those of Woodyatt. (e) Body Protein and Carbohydrate Storage. Another source for the formation, though perhaps not for the storage of carbohydrate in the body, has long been recognized in protein. The close con- nection which is maintained between protein and carbohydrate in diabetes would make a clinician with modest chemical knowledge seek for some combination of carbohydrate in the protein molecule some arrangement by which a portion of the sugar molecule could be stored in protein or attached to protein and given up as occasion arises, ju>t as water is squeezed out of a sponge. Good chemists (and I have asked many) assure me that it is impractical to consider 140 FACTORS IX TREATMEXT OF DIABETES MELLITUS such a combination, and that one must consider that even with glueoproteids sugar can only be extracted from the protein molecule when the molecule itself is disintegrated. If disintegration does take place then (50 grams of carbohydrate may be derived from each 100 grams protein. The large quantity of movable protein and fat in the body suggests a large carbohydrate reservoir also. Few realize how large this quantity of movable protein is. It has been shown by Albert Mueller, of Vienna, 1 that by overfeeding it is possible to add 210 grams of nitrogen to the body, and this would be the equivalent of 1200 grams of body protein, in turn the equivalent of (5.3 kilograms of muscle tissue. Conversely, it has been shown by Professor Benedict 2 that a similar quantity can be removed. This movable protein amounts to about one- third of the total body protein. The readiness with which fat can be increased and decreased in the body is so universally recog- nized as to be overlooked. If we conceive that it is possible for a severe case of diabetes to form (50 grains of carbohydrate from 100 grams of protein, then from the movable protein of the normal individual, which has just been shown to be 12(50 grams, it would be possible to form 7.">(5 grams of carbohydrate. 7. Utilization of Carbohydrate in Diabetes. The demonstration of the utilization of carbohydrate in diabetes is as scientifically difficult as it is apparently simple. Nothing would appear easier than to say that when a patient ingests 50 grams of carbohydrate in the diet and excretes 2o grams of sugar in the urine the balance of 2") grams was utilized. I low complicated the situation actually is can be seen by: (d) A study of what constitutes the carbohydrate balance. (b) The influence of the changing weight of the patient. (c) The capacity of the body for storage of carbohydrate. (d) The difficulties attendant upon the use of the respiratory metabolism as an index of combustion of carbohydrate. (e) Finally, the effect of exercise upon the carbohydrate utiliza- tion. Before entering upon a discussion of each of these topics in detail I quote the closing sentence of my Harvey Lecture : :i " In conclusion, it is gratifying to be able to record that the recent experimental evidence confirms the old clinical view that the severe diabetic still retains power to utilize a portion of the carbohydrate of his diet, small though it may be, and that herein lies renewed hope for the success of treatment." 1 Mueller: ZcntralU. f. d. (lc>annn!.> I'h p. 017. -T;irMO4, upon this point. (d) Respiratory Quotient, (treat hopes were entertained that the changes in the respiratory quotient following the ingestion of carbohydrate would definitely show whether any was utilized. It i^ true that the many experiments at the Nutrition Laboratory tend to point to a utilization of a small quantity of carbohydrate, even in the severest forms of diabetes. When it is considered, however, that even the utilization of 24 grams of carbohydrate in twenty-four hours at the rate of 1 gram an hour could not be readily shown by methods of studying the respiratory metabolism now in use, it is evident that proof of utilization of carbohydrate needs further study. 1 nquestionably, the respiratory metabolism is the most accurate means of proving carbohydrate utilization, but it is a method which at present cannot be expected to show whether small quantities of carbohydrate are utilized or not. Tables (SO S3 ilhi>trate the effect of the ingestion of various carbohydrates levu- losc, oatmeal, potato- upon the respiratory metabolism as shown by changes in the respiratory quotient. Whereas comparatively slight changes in the respiratory quotient following the administra- tion of these carbohydrates to severe diabetics were observed, still the general tendency points to a slight utilization of the carbo- hydrate. It is noteworthy, however, that Cases No. MM- and 441, with positive carbohydrate balances of 100 grams or more for the day of the experiment gave in the former instance a respiratory quotient of 0.71 before and immediately following the administra- tion of oatmeal, and in the latter 0.70 before and 0.71 directly afterward. In future experiments upon the utilization of carbohydrate, it must be borne in mind that the rapidity with which diiferent foods enter into the metabolism varies. Thus Higginshas shown upon normal individuals that sucrose, lactose and levulose begin to be burned quite as soon as alcohol if not sooner; glucose and maltose are not utilized as food as soon as the other sugars or alcohol, approximately twenty to thirty minutes elapsing before their combustion plays an important part in the metabolism. 1 TOTAL METABOLISM IN DIABETES MELLITl'S 143 TABLE 80. -EFFECT OF LEVTILOSE AND OATMEAL ox A SEVERE DIABETIC. Case '.V.V2. Female. A^c-d thirty-seven years. Weight 40 kilos. Date. March 31, 1011: 10.50 . . . Fasting 11.23 . . . Fasting 12.10 172 271 0.03 40 12.44 ... 1S4 201 0.70 44 1.30 ISO 210 0.73 42 2.05 172 235 0.73 40 2.2S ... .. 171 250 O.OS 42 2.53 ... .. 166 240 0.00 40 April 2, 1011: S.I 7 . . . Fasting S.45 . . . Fasting 0.14 . . . Fasting Oat me 10.13 ... .. 103 234 0.70 30 10.30 . . .. 107 22S 0.73 30 11. OS 177 23S 0.75 40 12.12 170 230 0.74 39 1.2S . . . ,. 154 200 0.75 35 2.37 103 200 0.7S 36 TABLE SI. EFFECT OF LEVULOSE UPON THE RESPIRATORY QUOTIENT OF DIABETIC PATIENTS. Carbo- hydrates. SiiKar in Respiratory Duration, Month of preceding urine quotient. months, observation. day, Levulose, 24 hours, Before After Case. Dead. Kins. Kis. Kins. levulose. levulose. 332 . 2S 23 100 = 552 . . . 32 IS 30 100 3 7S5 . . 23 20 20 SI 1 7 No increase. TABLE 82. -EFFECT OF POTATO o.v RESPIRATORY QUOTIENT IN" SEVERE DIABETES. Carbohydrate intake. Suwir in Respiratory Case. Duration, Months, I'rceediiiK Test day, urine 2 1 quotient. \Io. months, observed.- day. KIIIS. K'ns. hours, gins. Before. After. 03 - 705 ... 7 3 15 22 20 0.74 0.73 85 SOO , 6 3 10 03 O.OS 0.71 ' Si gins, lovulose and later gins. carl), as vegetables 00 gins, total. 2 4S gin-, carl), as potato; 10 gins. carh. as oatmeal; 5 gins. carl), as cream, total 63 gins.; later in day, 22 gins. carh. as potato and vegetables; also one egg and 30 gins, butter. 3 60 gms. carh. as potato; later in day, one egg. butter, (i gins. carl), as vegetables. 144 FACTORS IN TREATMENT OF DIABETES MELLITUS TABLE 83. EFFECT OF POTATO ox THE RESPIRATORY QUOTIENT OF A SEVERE CASE OF DIABETES. Case SOI). Male. Weight, 02 kilos. C( > 2 per O 2 per ( 'alories per Blood min., min., Respiratory kilo, per .sujjnr. Date. Condition. c.c. c.c. quotient. 24 hours. per cent Dec. 22, 1914: 9.25 } f 150 223 0.70' '241 9.54 Fasting . \ 150 224 0.07 i-O.OS 24 j- 24 10.22 j [loo 228 . OS J 10 15 14 Potato - ti() tz us. curb. 10 55 18 1 1 59 .... 181 257 0.71 1 28 12.22 10S 252 0.07 0.09 27 27 172 250 ().0 ( .) ~>- 3.00 .... 170 233 . 73 ' 20 ' 3.20 157 227 0.70 0.72 25 \ 25 3 54 100 231 0.72 25 J 4.45 . 0.19 TABLE 84. EFFECT OF OATMEAL ox THE RESPIRATORY QUOTIENT OF SEVERE DIABETIC. Case 773. Female. Weight, 40 kilos. Date. Condition Oct. 10, 1914: ,x.OO Fast iny; 11.00 . . . . Oct. 13: MM) Fasting 11.00 . . . . Oct. 19: 9.00 . . 146 ( (at meal o 42 ^ms. carl). 178 C0 ? per <>2 per Calories Hl< min., niin., Respiratory per kilo suj. c.c. c.c. quotient. per 24 hrs. per 212 0.09 30 249 0.72 43 138 189 0.73 33 0. 101 1 ar. cent. Fasting . . . 135 195 Oatmeal o SO urns. carl). 12.00 107 237 Oct. 20: After breakfast Diet contained 15 gins. carh. October 9 and October IS. 34 40 (c) Effect of Exercise. The effect of exercise upon the utilization of carbohydrate has excited interest for many years and has recently been advocated by Allen. Yon Noorden, referring to this subject, writes: "\\ide individual differences exist with regard to this matter. It is found that some patients can tolerate much more carbohydrate when taking plenty of exercise; with others this is not the case, muscular work doing them more harm than good." 1 At the Nutrition Laboratory a small series of experiments were conducted by Benedict and myself and published in !!)!(). - The evidence collected was not enough to warrant general conclusions, 'Von Xoonlen: Diabetes Mellitus, 10. 15. Treat Company, 1905, p. 177. 2 Benedict and Joslin; Metabolism in Diabetes Mellitus, Carnegie. Pub. 130, p. 217. TOTAL METABOLISM IN DIABETES MELL1TUS 145 but insofar as it went, it tended to show that during exercise there was a slightly increased combustion of carbohydrate, as shown by a rise in the respiratory quotient and also a rise in the total metabolism. TABLE So. -EFFECT OF OATMEAL ox THE RESPIRATORY QUOTIENT OF SEVERE DIABETICS. For text, see page 142. Duration. Date. Carbohydrates ingested. Respiratory quotient. Sugar in urine, gnis. Carbo- hydrate intake for entire 2-1, gms. Carbo- hydrate balance for entire 24, gins. Case Xo. Onset to coma, mos. Month of test. Day preced- ing, gms Before test, gnis. Fast- ing. After oat- meal. 194 34 31 Sept. 22 15 0.74 42 15 -27 23 15 100 + 0.71 0.71 50 165 + 115 24 165 0.72' 19 15 -4 246 15 11 Aug. 9 50 40 0.71 . 67 124 9 9 13 Oct. 29 65 60 0.68 0.70 100 125 +25 30 125 0.71' 93 65 -28 25-31 71 . 69 102 72 -30 281 19 17 Dec. 1 15 0.75 69 135 +60 2 135 29 . 76' 58 45 -13 3 45 0.76 38 30 -8 332 28 13 May 19 100 25 0.73 15 in 3 hrs. 26 95 0.73 3 in 3 hrs. 24 April 2 ? 52 0.74 0.74 97 26 June 2 ? 48 0.71 . 69 36 330 132 127 May 18 20 0.73 26 45 +19 21 45 25 . 75 31 45 +14 441 11 9 Sept. 29 15 75 . 70 0.71 65 165 +100 Oct. 9 15 73 . 69 ) 79 561 33 23 Feb. 7 60 . 75 31 60 +30 8 60 116" 0.71 0.74 128 185 +57 9 185 200 0.72' 0.72'- 209 205 -4 10 200 . 76- 102 60 -42 591 50 44 April 10 ? 0.74 63 30 -33 11 30 . 73 37 15 -22 12 15 80 . 70 0.70 85 165 +SO 13 165 80 0.73 1 . 69 1 77 165 +88 15 40 0.69 29 9 773 20 18 Oct. 8 115 70 0.70 176 165 -10 10 15 47 . 69 0.72 95 130 +35 13 50 . 73 84 50 -34 19 15 80 0.70 0.70 97 115 +18 74G 22 3 18 7 65 28 . 73 93 65 -28 9 15 50 . 73 0.71 87 163 +76 10 165 0.72 1 35 25 -10 15 165 80 0.74- 96 165 +69 7S6 17' 14 Nov. 12 15 60 . 69 . 73 62 +02 10 1 Respiratory quotient taken following an oatmeal day 2 Respiratory quotient taken subsequent to two oat me 140 FACTOh'S IX TREAT. MEXT OF DIABETES MELLITl'S Nehring and Schmoll 1 concluded that muscular activity in dia- betics was also at the expense of glycogen from the fact that it was possible in many instances to reduce the sugar excretion by con- trolling the muscular activity. This is also fully in accord with the experiments made by Mohr 2 in which he found with a diabetic dot; that there was a noticeable increase in the respiratory quotient during walking, and therefore he concluded that diabetics can burn but little glycogen until there is the most urgent need for it. No other respiratory experiments upon the etl'ect of exercise; have been made by Benedict and me, but the following clinical observations are of interest: ( "ase No. 1 1 2, a doctor, and a good observer, noticed that the quantity of sugar in the urine always decreased while on hard camping trips in Maine, although the diet contained the same, or probably more, carbohydrate than when he was at home. Case No. 22, a Harvard professor, said to me in 1901 that "mental work makes sugar, manual work burns it up." Case No. 511 shows sugar Avhen hard at work in the city, but when quite as occupied with mental work in the country, but with more exercise and a similar diet, shows no sugar. Since my attention has been directed to this question by Allen's experimental work, I recall instances of diabetics who were able to carry on severe muscular labor, although the diabetes was of extreme severity. One striking case of this type was a railroad inspector, who traversed several times a day alone in a handcar many miles of railway, and yet maintained fair health, much to the astonishment of his physician and myself. 1 was always impressed in former days of treatment by the better results frequently obtained by ambulatory as compared with the hospital treatment at that time, provided the same degree of attention was given to the details of the diet and hygiene of the patient. According to the result of recent work it would seem now as if this was in part due to the individual maintaining his ordinary vocation, and thus undergoing considerable muscular work. Finally I have looked up 29 cases of diabetes of fifteen or more years' duration, and in many of these it is apparent that considerable activity has been maintained throughout life. Recently Allen and DuKois' have* recorded a slight rise in the respiratory quotient of a severely diabetic patient undergoing mild exercises during two respiration experiments. Like Benedict and myself, however, they draw no positive deductions. Allen 1 has found that diabetic dogs with a known constant limit for tolerance of carbohydrate or protein upon vigorous exercise 1 Nchrinir and Srlimnll: Zisrhr. f. klin. Mrd.. I '.107. xxxi. p. .V.) '.)_'. - Molir: L<>f. fit., p. '.Kill. Sec p. IL'O. :; Allen and DuBois: Luc. fit., p. Klin. 'Allen: Ijo-toii Mi-d. and Sur^. .lour., l'.M.">, clxxiii, p. 74^. LOSS OF ENERGY IN THE URINE OF DIABETICS 147 in a treadmill showed a marked increase in tolerance, as demon- strated both by the sugar in the urine and in the blood. lie further says that dogs which have for months regularly shown glycosuria whenever they were given 100 grams of bread, on exercise became able to take 200 grams of bread as a regular daily ration without glycosuria. With patients, he has observed results sufficiently favorable to warrant recommending exercise as an addition to treatment, and lie goes on to say that in a patient free from glyco- suria with a persistent hyperglycemia, one fast day with exercise may reduce the blood sugar as much as several fast days without exercise. Dr. Allen was kind enough to call this to my attention in the summer of 1915, and I am able to confirm his statement. In particular I have noted that patients have retained strength and appeared better during fasting when moderate exercise was allowed than when they remained abed. May it not be that the good results of treatment which are often obtained with children are in part due to the fact that they exercise far more than adults? On October 22, 191"), I observed that Case No. 925, aged eleven years, onset one month previously, was tired out after playing tennis five minutes, although eating 13 calories per kilogram body weight, but on October 25, without essential increase in diet, he played tennis for half an hour without being tired. H. THE LOSS OF ENERGY IN THE URINE OF DIABETICS IN THE FORM OF SUGAR AND ACID BODIES. Diabetic patients lose a considerable percentage of the energy of their food by the excretion of sugar and acid bodies in the urine. The enormous draft upon the food supply is obvious when we realize that occasionally even as much as 680 grains of sugar (2720 calories), and 100 grams acid bodies (500 calories) may be so excreted. This is also illustrated by Case No. 235. TABLE 8(5. CASK 2:>f). MALE, A<;ED TWENTY-SEVEN YEARS AT ONSET ix 1901. No HEREDITY. WEICIIT, (>"> KILOS, FEBRUARY 17, 1909. Substance. Grains. Protein . . 100 X 4 Carbohydrate no X 4 Fat . '. . . 220 X 9 Alcohol 30 X 7 2810 807 2810 - 807 = 200)5 net calories. The diet thus furnished only (2003 -f- 05) 31 calories per kilo instead of an apparent (2810 4- 05) 43 calories per kilo body weight. 14S FACTORS IX THEATMEXT OF DIABETES MELLITUH ('use No. oil, weighing 50.4 kilograms, also illustrates the groat loss of potential energy in the urine, as is shown from the following urinary ana lysis: December 25 and 2(5, I'.H 1 ,47(M)e.e. ; specific gravity, 1027; diaeetie acid, -\ | \-; (8-oxybutyric acid, .14. X; nitrogen, 14.5; ammonia, total (>..'>; sugar, copper reduction, 1 XX grains; rotation, !('() grams. Diet: carbohydrate,!.'!"); alcohol, 45; carbohydrate balance 55; XaHCO.i, 2i>; body weight without clothing, 50.4 kilos. It will be seen that 1XX X 4 = 752 calories were lost in the urine in the form of sugar, and 54. X X 5 = 274 calories in the form of /3-oxybutyric acid. Case 1 Xo. 205, voided on October 2o-24, 11)00, approximately 10 liters of urine, containing GXO grams sugar, equivalent to 2720 calories. I. GLYCOSURIA IS THE MOST TRUSTWORTHY SYMPTOM. In most chronic diseases there is no criterion by which the success or failure of treatment can be readily estimated. Such is not the case in diabetes. One can toll when treatment is successful, for the patient should be free from sugar and acid and be happy and vigorous. While the twenty-four-hour quantity of sugar in the urine is not an absolute measure of diabetes, still in the vast majority of cases it is an accurate index. Taken alone the quantity of sugar eliminated is not of great significance, still less the percentage of a single specimen, but when compared with carbohydrate intake it is possible* to determine quite definitely the condition of the patient. I should deplore that anyone 1 from the above statement should lay undue- stress upon this sign, because 1 the strength, the 1 weight, the 1 me'iital attitude, the pivse-ne-e or abse-nee of complications, and the- acidosis are all important, but it remains true- that this is the 1 erne 1 tVaturo of the 1 disease which is of almost mathematical accurae-y, though we 1 often err in thinking of it alone 1 . The' sugar in the- urine" of diabe'tic patients usually varies diree-tly with the- quantity of carbohydrate-forming mate'rial in the 1 die't, to a lesseT e-xtent with the 1 protein, but is uninfluenced by the- fat, save- as that influence's the total calorie- intake. A change of did is shown in the 1 urine within a few hours. J. THE NATURE OF DIABETIC ACIDOSIS AND ITS RELATION TO COMA. 1. Acidosis in Normal Individuals.- If a healthy individual live-s for three siicceive days upon a carbohydrate-free die't, the 1 urine- vended upon the subsequent morning will show a re-action for dia- evtie- acid with the feme- chloride. This is evidence of a type- of acid intoxication which has been termed bv Xaunvn " acidosis." Acidosis NATURE OF DIABETIC ACIDOSIS 149 is represented by three bodies: /3-oxybutryic acid (CII 3 .CIIOII.- CILCOOII), diacetic acid (CH 3 .CO.CH 2 COOH), and acetone (CII 3 .rO. CII 3 ), all of which are excreted in the urine, and the latter in the breath as well. Acidosis is still more simply produced in normal individuals by fasting, but it is of a milder type. Thus Benedict's subject at the Nutrition Laboratory constantly showed an acidosis during his fast of thirty-one days. 1 welcome the opportunity to call to the attention of physicians the records of this case of fasting which are more complete than those reported for other fasting individuals. This subject took only 900 c.c. of distilled water daily by mouth and during the thirty-one days of the fast lost 13.25 kilograms or 21 per cent, of his normal weight. Of this quantity, 55 per cent, was computed as due to water, 2 per cent, to stored glycogen 13 per cent, to protein and 27 per cent, to fat. The remaining 405 grams (3 per cent.) was attributed to loss of mineral salts. The blood-pressure fell 30 mm. mercury. In the four days prior to the fast the pulse-rate varied between 82 and 70, during the first four fasting days between 68 and 02, and in the last four days between 59 and 57. So soon as the percentage of carbohydrate in the com- bined carbohydrate and fat of the body metabolized fell below 12 percent, acidosis was marked but eventually no carbohydrate was burned and yet the acidosis markedly decreased. From the first to the third day from 10 to 11) per cent, of the heat was derived from glycogen, but between the third and thirteenth days this fell to 1 and 3 per cent, and after this day entirely ceased. The acidosis in this man was demonstrated by estimations of the carbon dioxide tension of the alveolar air, the ammonia-nitrogen, and the /3-oxybutyric acid. During these thirty-one days the lowest point reached by the carbon dioxide tension was 26.8 mm., the highest ammonia-nitrogen was 1.91 grams upon the sixteenth day, the /3-oxybutyric acid reached (i.9 grams upon the twenty- fourth day. (See Table 87 on pages 150 and 151.) It is of interest that the acidosis during prolonged fasting in a healthy individual does not notably increase with the continuation of the fast, but it is still more interesting, in the light of the behavior of acidosis in fasting diabetics to be described later, that it should not at length entirely disappear. The effect upon the acidosis of prolonged fasting in obesity is not known, but during the first four days of the fast of a fat young woman Folin and Denis 1 found that there was a notable increase. This is shown in Table 88. 1 Folin and Denis: Jour. Biol. Them., 1915, xxi, p. 183. ( 150 ) >.CO r^ co x OM 1 ^"S X t^ O CO c o > c-i ,~ en -7 t>. ^5 >v C3 <*< ~ t^ C - 7 CO _ >i~ .27 6 ^. o t^ >.>i S '-.? -t c ^5 t^ ~ co" *" _ ST. C ^ X ^2 R t^ i c 2* - ^ j x <"- M ' X *r co ~: -, C^l tr r~ r--t * * ^ t~- c^r ^c ' < - X S = : -i t^ tc 1.1 -^ s ox I-H --c x s co r-i i .i x c-i H rt COOOOC O x --r; c-t f f co c x co EH Sit- CO t^ X "~ 1^ ~] . t- _ . . . . . _. M . c; c: M i^ .1 ... __._._ I < ~? "< ~-c~_, M - r! - ^ ..^ r3 >~ C-l C -? O i-l C; t~ VC: C-l t^ CO l^ S I- -H C-l 2TI'~ O i 2 "C~ 2 "" _ ' - ' '~ _'._''_' _ ' '""_: ri M c-i -co x _ ^ 2 i- CO CO ?l ~ J^ "1 C-l C~ a 152 FACTORS IX TREATMEXT OF DIABETES MELLITUS TABLE 88. THK Acinosis OF A FAT WOMAY nrui.\<; THREE PERIODS OK FASTI \<; (Foi.i.v AND DEXIS). Day. cms. 1 0.01 cms. 0.27 c I) 11 230 is. Remarks. Feeling well. 2 .OS 1 .12 2.00 7:5 150 5.2 Slight headache. 3 . 10 1 . .J7 17.01 1 87 508 21. Severe headache. 1 . ss 2. 10 is. 47 ~2 50 ( 05 49. 5 Heada< he, nausea, and ( lizziness. 15 31 ISO Feeling well. 10 .02 37 200 Feeling well. 17 o:; 1 .17 0. 17 53 335 30 Feeling well. IS 35 1 . 10 5.44 I 01 505 32. Slight headache, nans a. 10 .-1(1 1.15 13.54 1 50 055 45 Headac he, nausea, and ( lizziness. 24 50 1 15 Feeling well. 25 37 100 Feeling well. 20 04 0.37 0.1S 51 210 GO . Feeling well. 27 20 1 . 30 17.34 SI 300 24. Headai he, nausea. The subjective symptoms disappeared as if by magic as soon as the patient began to partake of food. A single piece of toast with a cup of coffee at once; restored the patient and kept her perfectly cheerful for several hours. The marked acidosis presented by this fat woman is in distinct contrast to the slight acidosis obtained in Benedict's normal sub- ject, and suggests the importance of fat in bringing this about. Xo less striking is the tendency of the acidosis to become less during successive periods of fasting and later this will be seen to be of great therapeutic significance. The amount of acidosis thus far described is much below that produced when an individual is placed upon a diet from which carbo- hydrate is wholly excluded. In the classical experiment of Gerhard and Schlesinger \) grams of /3-oxybiityric acid were demonstrated in 1 .">()() c.c. of urine while they were living upon a fat-protein diet, and Benedict and I 1 obtained S grams of /3-oxybutyric acid and 4.07 grams ammonia in the urine of a healthy student who had been upon such a diet for three days. If the quantity of fat in the fat-protein diet, is increased there is a simultaneous increase in the excretion of acetone and /3-oxybutyric acid. Thus Satta- found the following changes during four days when a normal individual was taking L'OO to MOO u-rams of fat and I'OO to LV>0 "Tains of meat: NATURE OF DIABETIC ACIDOSIS 153 TABLE 89. ACIDOSIS OF A NORMAL IN-DIVIDUAL n>ox A FAT-J-ROTKIX DIET. Day. Acetone. /3-oxybutyric. 1 ... 0.0(52 0.84 2 ... . 0.0(50 0.73 :$ 2.5.50 :i..5 4 3.110 14.70 Landergren 1 still further increased the acidosis by combining muscular work with the diet. I5y this means not only was the carbohydrate in the diet excluded, but the carbohydrate stored in the body in the form of glycogen was rapidly depleted, and so not available for counteracting the acidosis. Forsner, 2 by forcing up the limits of fat in the diet, obtained an excretion of 42.8 grams of acid bodies in one day, and even when the diet contained 40 grams of carbohydra-te the artificial acidosis remained as high as 32.3 grams. Aside from the acid bodies eliminated in the urine, mention has been made of acetone excreted from the lungs, but this is probably not very considerable. Thus, it is estimated that a normal individual will excrete 0.013 gram acetone in twenty-four hours from the kidneys, but only 0.0038 gram from the lungs. After the adminis- tration of 90 grams of oleic acid to a fasting man I found 0.847 gram of acetone in the breath and 0.185 gram in the urine during twenty-four hours. 3 Sassa 4 found traces of /3-oxybutyric acid in health in the blood of men and animals amounting to 0.01 to 0.02 per cent. In diabetic coma the percentage may reach eight times the normal figures. Just as Landergren was able to increase the production of acid bodies by depleting the glycogen storage, through muscular exercises, so S. R. Benedict has succeeded in doing the same with phlorhi/in. In an experimental period of about two weeks, during which phlorhizin was injected to prevent the oxidation of sugar through the breaking down of protein the subject excreted 32 grams of pi-oxybutyric acid per day, and during the whole period over 300 grams. 5 Experiments of this nature have been interpreted as showing the similarity of the acidosis of normal individuals to that which is seen to occur in diabetic individuals, and furthermore that the absence of the combustion of carbohydrate is an important factor in acidosis. AVhether such a conclusion is justifiable has been recently placed in doubt by experiments upon severe cases of diabetes 1 Landorgren: Xord. mod. Ark., 1910, ii, p. 1. 2 Forsner: Arch. Skaiulin. f. Phys., 1910. xxiii, p. 305. 3 Joslin: Jour. Mod. Research, 1904, vii, p. 433. Sassa: Bioehem. Ztschr.. 1914, lix, p. 3f>2. 6 Benedict: Proc. Soc. Exp. Biol. anil Mod., 1914, xi, p. 134. 1.V4 FACTORS IX TREATMENT OF DIABETES MKLLITUS with little or no tolerance for carbohydrate, who, upon fasting become entirely free from acidosis. The question, however, is not wholly settled, because our present methods of determining the respiratory quotient are not sufficiently exact to absolutely exclude the possibility of the combustion of small quantities of carbohydrate, for instance at the rate of 1 gram of carbohydrate per hour for the twenty-four hours. I'ntil lately it has appeared safe' to say with Landergren that the diabetic acidosis is due not so much to a loss of power to burn carbohydrate, but rather to the fact that the diabetic patient has no carbohydrate to burn. Today we cannot be so dogmatic, for the patient may have carbohydrate to burn of which we have not been aware, and it is certain that the fat metabolism plays a more prominent part in the acidosis problem than has been supposed. (a) Adaptation of the Body to a Non-carbohydrate Diet and to Acidosis. The Kskimos have always been a pu/xle to those 1 who explain the presence of acidosis as due to the absence of carbo- hydrate in the diet. It is possible, however, that the Kskimos obtain more carbohydrate 1 than we have thought, and, furthermore, the quantity of protein in their diet is sufficiently large to furnish notable quantities of carbohydrate. It has been repeatedly shown that large quantities of protein will restrict the development of acidosis in normal individuals on a protein-fat diet. Another instance of adaptation is furnished by the experiment of the fasting man at the Nutrition Laboratory. During the progress of his fast the acidosis increased for a time, then was practically at a stand-still, and decreased in the latter days of the fast. 1 It is reason- able to conclude that the body of the fasting man adapted itself to the changed conditions and in some way restricted the output of acid bodies. This may be the explanation of the lack of acidosis among the Ivskimos. Finally the case of the fat woman studied by Folin and Denis, and already cited on page 1.12, beautifully illus- trates this point. This power of adaptation is certainly an important factor to bear in mind, because it shows that by training it may not be dangerous to keep a diabetic patient on a non-carbohydrate diet for a prolonged period. Changes in the carbohydrate and water content of the diabetic organism occur so rapidly that they are often overlooked, and yet these may be of great importance in influencing the acidosis. The quantity of carbohydrate which ordinarily must be oxidized to prevent the appearance of acidosis in the healthy individual is approximately KH) grams. This statement in this form is really incomplete for it is necessary to know the amount of fat which is NATURE OF DIABETIC AC I DOM K 155 simultaneously ingested. The quantity of protein metabolized is also of great significance, for from every 100 grains of protein approximately GO grams of carbohydrate may be formed. Acidosis may disappear when the body has been trained to live upon a diet containing less carbohydrate, but with this exception the presence of acid bodies in the urine accompanies a lowering of combustion of carbohydrate in the body. It makes no difference whether the lowering of combustion is due to the lack of carbohydrate in the diet, to fasting, starvation, or to lack of carbohydrate which occurs in the malnutrition of cachexia or the increased metabolism of fever. The absence of food, together with disturbances in the metabolism of fat, probably accounts for the frequent acidosis in the gastro- intestinal diseases of children. (6) Source of the Acid Bodies. The source of the acid bodies is evidently not in carbohydrate. Of the remaining two foodstuffs, it has already been shown that an increase in the administration of fat leads to an increased acidosis, and there is no question but that fat is the chief source of /3-oxybutyric acid. Each molecule of a higher fatty acid, as it is broken down to a lower, provided it has an even number of carbon atoms, leads to the production of one molecule of /3-oxybutyric acid, though upon this point there is not complete unanimity. As a matter of fact, the only fatty acids which are present in the body are those containing an even number of carbon atoms. It has been estimated that out of 210 grams of fat, 72 grams of (3-oxybutyric acid may be produced. 1 Notable quan- tities of /3-oxybutyric acid, however, can also be formed from pro- tein. Thus, from 120 grams protein I->Gto40 grams of /3-oxybutyric acid may theoretically be obtained. It is interesting that those amino-acids of the protein molecule which lead to the production of /3-oxybutyric acid do not produce sugar, and conversely, that those which lead to the formation of sugar produce no /3-oxybutyric acid (c) The Percentage Relation of Acid Bodies to One Another. The. interrelation of acetone, diacetic acid, and /3-oxybutyric acid to one another must be very intimate. Netibauer 2 considers that there is a reversible action between diacetic acid and /3-oxybutyric acid, and that these readily change back and forth with one another. It is thought by some that there is a fixed proportion between these acids, varying with different patients but constant in the same patient. Folin 3 is inclined to doubt the existence of preformed acetone in the body, but explains its presence in the urine as a decomposition product of diacetic acid. At any rate, acetone 1 Magnus-Levy, in Spcz. Path. u. Therap. inn. Krunk., Kraus u. Brugsch, Berlin, 2 Nenbauer: Verhnrd. <1. Kong. f. inn. Mod., 1010, xxvii, p. 5GG. Folin: Jour. Biol. Chem., 1907, iii, p. 177. Ml) FACTORS IX TRKATMKXT OF DIAHETES MEELITCS seldom constitutes over a small percentage of the total acidosis, and acetone and diacetic acid combined not over )>() per cent. (tl) Mode of Elimination of Acid Bodies. Practically all the acid bodies are eliminated by the kidneys. Only one of these acetone - is excreted by the lungs, and the amount must be small. In experi- ments upon a healthy man following the feeding of oleic acid, I obtained an excretion of O.M7 gram acetone calculated for the 24 hours in the breath. The method of estimation of the acetone in the breath was not very satisfactory, and it would be advantageous to institute a series of experiments of this nature on a larger scale. The oilier two bodies are excreted as salts and even as free acid in the urine. Magnus-Levy 1 found that the concentration of these acids in the urine seldom rose above 1.5 per cent., and lie never encoun- tered out of < oma values greater than (>.(> per cent. It is evident, therefore, that the elimination of the acid bodies is closely con- nected with the quantity of water excreted. It is also dependent upon the available alkali, for it is universally recognized that in severe acidosis the quantity of /3-oxybutyric acid in the urine rises when alkalis are given. 2 I do not, however, remember data in the literature which show an increased elimination of acid bodies following the administration of alkalis to a normal individual with experimental acidosis. Such would be 1 valuable. It would be wrong to consider, however, that the only method the body had by which to free itself of acid was by the kidneys. The body has a most efficient pathway in the lungs through which carbonic acid is constantly removed from the body. In fact, so soon as this acid begins to increase in the tissue's, it also rises in the blood, this stimulates the respiratory center with resulting hyperpnea and the excess of carbonic acid is removed by the increased ventilation. The total quantity of carbon dioxide* for the twenty-four hours is not increased save for the increase due to the increased metabolism which accompanies acidosis, and indeed the percentage of carbon dioxide in the alveolar air is actually diminished because it is diluted in consequence of the increased ventilation. '2. Acidosis in Diabates. --(a) Similarity to that in Normal Indi- viduals. The acidosis of diabetic individuals differs in no particu- lar from that of normal individuals except in degree, and even this limitation may be overcome in a normal in r hvidu:il placed on a carbohydrate-free diet with a forced feeding of fat and much exercise. Whereas in normal individuals the acidosis seldom exceeds a few grams, in diabetic individuals it often reaches hundreds of gran>. Thus, C'ase No. -I, male, onset at fifteen vears of age, NATURE OF DIABETIC ACIDOSIS 157 without diabetic heredity, excreted three years and two months later, in three successive days of coma, 4o7 "Tains /3-oxybutyric acid as calculated from the excess of bases in the urine or )>">(> grams acid bodies, calculated from (3-oxybutyric acid and diacetic acid extracted from the urine. Just as in normal individuals the presence of these bodies implies non-combustion of or combustion of insufficient carbohydrates, in diabetic individuals the presence of these bodies implies non-combustion of or combustion of insuffi- cient carbohydrate material. But it is easier for'a diabetic individual to acquire an acidosis, because most of the carbohydrate which he eats is lost to his metabolism. If the diet of the diabetic individual is unrestricted, he often eats so large a quantity of carbohydrate- forming material that enough of it is oxidized to prevent the occur- rence of these bodies: Thus Case No. 295, male, onset at sixteen years, seven years later excreted 10, 000 c.c. of urine containing GSO grams of sugar, and yet failed to show a positive reaction for diacetic acid. When the carbohydrates in his diet were restricted to even 280 grams, acidosis appeared; when the diet was still further restricted, the acidosis became extreme. Just as normal individuals vary according to the ease with which an acidosis is produced and the extent of the same, so do diabetic individuals. Very likely this will eventually be explained by the extent of the carbohydrate and fat storage of the different individuals, as well as by the storage of alkalis and water which plays such an important factor in the elimination of acid bodies. (6) Extent of Acidosis in Mild, Severe, and Extreme Cases. The acid bodies which make up the acidosis of diabetes are constantly being produced and constantly being excreted. A moderate acidosis represented by the excretion of 5 to 10 grams /3-oxybutyric acid, the elimination of 2 grams ammonia, or a fall of carbonic acid in the alveolar air to 4 per cent., or 29 mm. mercury may or may not be harmful to an individual. So far as my experience goes such moderate quantities of acid products are of little significance in cases of long duration, but this is not true of those patients who have recently acquired an acidosis for the first time. This does not prove that these bodies are not harm- ful, but that they are a useful sign of a condition which ought to be remedied for fear it will grow worse, rather than that they are actually harmful. When the quantity of acid is present in double this amount, diabetic patients invariably show that they are burdened with disease, but it would be unfair to attribute this state to the acidosis alone, and not to the neglected disease itself. In former days, when the acidosis of patients was not controlled, over and over again 1 saw diabetic cases who carried an acidosis of 4 grams ammonia for years, with only gradually declining health. 158 FACTORS IN TREATMENT OF DIABETES MELLITUS Even so extreme a ease of diabetes as Case No. 344, eight years after onset showed 51.5, 52.0, 52.0, and .14. grams /3-oxybutyric acid (if calculated by excess of bases 4o7 grams) already mentioned as excreted by Case Xo. 4, during three days of coma is nearly the maximum found in the literature. This is equivalent to the elimina- tion of o grams per kilo body weight daily for three days for an individual weighing 50 kilograms, or if the acidity is expressed in terms of hydrochloric acid approximately 1 gram per kilo. Xaunyn estimates the quantity of /3-oxybutyric acid in the tissues of a patient near coma at between 200 and .')()() grams. /3-oxybutyric acid circulates in the blood combined with alkali as a salt, but the acid is excreted in the urine to a large extent as free acid and the base 1 retained in the body. It is only when much alkali is given that the large quantities of /3-oxybutyric acid, such as were excreted by Case Xo. 4 during coma, are found. The formation of acid never goes on to such a degree that the blood shows an acid reaction. Such a condition is incompatible with life. In fact, so constant is "the reaction of the blood that a change from the reaction of ordinary tap water, which is more alkaline than the blood, to that of distilled water, which is much more acid than blood, would be fatal." If the blood were acid, the carbonic acid would be displaced from its combination with an alkali by the stronger /3-oxybutyric acid and set free in every cell of the body, and no alkali would be available to combine with it and take it back to the lungs for elimination. The normal blood and respiratory exchange is represented by the following formula: 2XaIICO ;i = Xa 2 CO 3 + CO, + II,(). The above formula clearly shows how easy it is for /3-oxybutyric acid, which is stronger than carbonic acid, to sei/e upon free alkali and thus hamper the removal of carbonic acid from the tissues. Henderson, at the meeting of the Association of American Physicians, in Washington, in 19 Hi, said: "Any modification of the normal equilibrium between acids and bases within the organism whereby the power to neutralize acid is diminished is to be regarded as a condition of (letdown. "What is essential and common to all conditions of acidosis is a depletion of the alkali of the body. This involves at least a diminu- tion of the bicarbonatcs of the blood, and in severe cases it probably XATUIiE OF DIABETIC ACIDOMS 159 involves the draining away of very large quantities of alkali from many sources. "Nothing is simpler than the process by which this condition is established. If an acid is poured into an aqueous solution of carbonic acid in equilibrium with the air, to which a certain amount of a bicarbonate has previously been added, the acid will react, according to its concentration and avidity for base, to a greater or less degree with the bicarbonate, forming in due amounts its own salt and free carbonic acid, which must escape into the air, since the solution is already in equilibrium with the carbonic acid of the atmosphere. Thus, for instance, a solution of sodium bicarbonate to which half of the equivalent amount of hydrochloric acid has been added will in the course of time contain just as much free carbonic acid as it did before and just half as much bicarbonate. Now the laws governing the equilibria between acids and bases determine the fact that it is in like manner chiefly bicarbonates which react with acids introduced into the blood, and when the resulting carbonic acid has been liberated by the lung the result is very similar to that of the simple chemical experiment. It is to be observed, however, that the respiratory process tends, under these circumstances, to eliminate more than the newly liberated carbonic acid; the tension of carbonic acid in the blood is thus diminished nearly in proportion to the diminution of bicarbonates, and since the hydrogen-ion concentration is proportional to the ratio of the free carbonic acid to the bicarbonates, the degree of alkalinity of the blood is unchanged by the introduction of acid hence the theory that the hydrogen or hydroxyl ion is the hormone of respiration but tlie eijiiilibriuni is changed. "The condition of acidosis, simple though it may be in its essential features, is physiologically of the greatest complexity; for the acid- base equilibrium is involved in a great variety of other equilibria, such as those of osmotic pressure and volume; it influences the distribution of electrolytes between corpuscles and plasma, the state of colloids, hence perhaps the union of oxygen and hemoglobin; it modifies the activity of enzymes; and it is probably involved in many undiscovered phenomena. It modifies, moreover, the general metabolism and the activities of lung and kidney. Here are prob- lems enough to occupy many years of research, and I have no doubt that the results of these researches will advance the science of medicine." The method by which the blood retains its alkalinity is admirably described by Rowland: 1 "The important constituents of the blood so far as the regulation 'Rowland: Bull. Johns Hopkins Hospital, 1916, xxvii, p. 03. 1(K) FACTORS IX TREATMEXT OF DIABETES MELLITUS of the reaction is concerned are (a) sodium bicarbonate, occurring both in the plasma and in the cells, (/;) the acid and alkaline phos- phates of potassium, found almost entirely within the red blood cells, and (r) the proteins. "Considering the blood first as a solution of bicarbonates: A la rue amount of acid (carbonic- acid) is constantly being formed in the tissues. It. must be removed by the lungs, but first it must be transported to the lungs by the blood. This stream of acid which, with an adult, in the course of the day, is the chemical equivalent of several hundred cubic centimeters of concentrated hydrochloric acid, is sufficient to render acid any ordinary solution and keep it permanently acid. If this should happen in the blood, life would of course be impossible, but owing to the laws that govern the reaction of solutions of weak acids and their salts, the solutions of bicarbonate are able to take up the quantity of the acid, carbon dioxide, without appreciably undergoing a change in reaction. Thus there can be transported from the tissues to the lungs and so con- tinuously eliminated from the body, a very large amount of acid. This steady escape of acid is accomplished with no harm and with no strain upon the organism. The respiratory center is adjusted to assist in the removal of the carbon dioxide. If there were no respirations and circulation were continued, eventually the carbon dioxide concentration would be the same in the tissues, in the blood, in the air and in the pulmonary alveoli. " But. the respirations lower the concentration in the lungs and thus allow the carbon dioxide to escape from the tissues where the concentration is highest, by the blood where the concentration is lower, to the air in the lungs where the concentration is lowest. The respiratory center is extraordinarily sensitive to the slightest alteration in the reaction of the blood toward the acid side, so that an increased production of carbon dioxide in the tissues, such as occurs, for instance, with muscular exercise, and the resultant slight excess in the blood is answered by an increased ventilation of the lungs which removes the carbon dioxide, thereby bringing the reaction of the blood back to normal. Other acids, whether formed in the body or introduced from outside, produce a similar effect. They displace the carbonic acid from the sodium bicarbonate and set the carbon dioxide free. This excess of carbon dioxide is removed by the increased pulmonary ventilation, leaving a neutral >alt, sodium oxybutyrate, or chloride or what not to be removed by the kidneys. Such a mechanism allows relatively huge amounts of abnormal acids to be at once rendered innocuous and removed; for instance, XallCO, + IIC! = Nad + IW + <'(),. The hydro- chloric acid is neutralized and the resultant sodium chloride is removed by the kidneys while the carbon dioxide is given off by the lungs. NATURE OF DIABETIC ACIDOSIS 161 "Henderson 1 calls the carbonates of the blood the first line of defense. Thus, dyspnea, more properly hyperpnea or increased pulmonary ventilation, under abnormal circumstances, is an agent of the greatest value in ridding the body of carbon dioxide and thus keeping the reaction within normal limits. It may also be remarked that hyperpnea is the best of all the evidences of acidosis to be obtained by physical examination alone. It may almost be said that hyperpnea means acidosis. " If the bicarbonates of the plasma were the only method of defense of the body the organism would succumb to acidosis as soon as the bicarbonate was depleted by the excretion of neutral salts through the kidneys; every molecule of an acid would rob the body of a molecule of bicarbonate. The second mechanism here comes into play and is that by which acids may be removed, leaving behind part of the base with which they have been combined, this base being available for further neutralization. The elimination is by the way of the kidneys. These have the capacity to excrete an acid urine from a nearly neutral blood. They remove acid phos- phate and save base with each molecule of acid phosphate that they excrete. Thus, although alkali is eliminated in the urine, it is much less than would be the case without this specialized-kidney activity, and can readily be replaced under normal circumstances by the alkali of the food. For instance, with the introduction of a foreign acid Xa 2 IIPO 4 + HC1 = XaCl + XaII 2 PO 4 the hydrochloric acid is neutralized, the sodium chloride and acid sodium phosphate are excreted by the kidneys or the following reaction may take place -NasPHO* + HoO + CO 2 = XaII,PO 4 + NaHCO 3 . By this method the sodium bicarbonate reserve of the body is renewed. "Henderson and Palmer showed the magnitude of alkali sparing very prettily by titrating with alkali the acid urine back to the normal reaction of the blood. The alkali spared was found in normal subjects to vary in terms of tenth-normal alkali, between 20(3 and 800 c.c. This is equivalent to saying that the kidneys eliminate from 200 to 800 c.c. of tenth-normal acid in twenty-four hours." (c) Safeguards of the Body against Acidosis. Two of the chief safeguards of the body against acidosis have been already discussed, namely, the removal of carbonic acid by the lungs without depleting the body's store of alkali, and second, the reaction between a molecule of disodium phosphate and a molecule of acid by which the sodium bicarbonate of the blood is conserved and yet large quantities of acid are eliminated. The glycogen storage in the liver and muscles of the body was considered Xature's chief safeguard against acidosis until it was shown, as tile result of the fasting treatment of diabetes inaugurated 1 Henderson: Am. Jour. Phys., 190S, xxi, p. 427. 11 1(>2 FACTORS IX TREAT ME XT OF DIAKKTEX MELUTCX by Allen, that acidosis usually disappeared in these individuals while fasting. It is hard to give up this hypothesis even now, although the low respiratory quotient of these patients indicates very little carbohydrate being burned. The anomaly becomes greater when it is considered that ;i healthy individual, fasting or on a non- protein dic-t, develops an acidosis, whereas a severe diabetic treated in exactly the same manner loses an acidosis. There is surely something here which is not understood. 1 It would appear that the diabetic individual is able to manufacture and utili/e some sub- stance which, in its combustion, is able to prevent acidosis, but yet a material which is not available to the normal individual, and the fact that as the aeidosis decreases the respiratory quotient rises above that to be expected upon a fat-protein diet is favorable to this view. These considerations show the importance of a determination of the amount of carbohydrate stored in the diabetic organism. In considering carbohydrate storage one must not neglect to include the large amount of carbohydrate which is available when the protein molecule breaks up. This is we'll illustrated by Case No. 51;!, Avith onset at thirty-three years, who finally succumbed to multiple carbuncles, septicemia, and diabetic coma, three years later. 2 During the last four days of life, while fasting, due chiefly to his inability to retain food, he excreted 142.4 grains nitrogen. This would represent the metabolism of SOU 3 grains protein and the setting free of 5.'54 grams carbohydrate. As a matter of fact the dextrose-nitrogen ratio in this case was ;!.():!, showing Nature's fruitless struggle to conquer an acidosis. Another safeguard against acidosis exists in the large amount of alkali stored in the body. Not only are sodium and potassium present in considerable quantities, but in emergency the organism can fall back upon the calcium and magnesium of the bones. Cerhardt and Schlesinger 1 originally pointed out that these alkalis met, an attack of severe acidosis by changing their path of excretion from the bowels to the kidneys, in this manner removing /i-oxy- butyric acid. In a study of the urine; of Case No. 4 during coma, 5 Mc( 'rudden found for me unusually large quantities of calcium and magnesium excreted in the urine. Whereas 40 per cent, of the magnesium is normally excreted in the t'eces, in Case No. 4 during three days 0.55 gram was found in the urine, but only 0.15 in the feces. Similarly, 2.0.'! grams calcium appeared in the urine, but 0.75 gram were eliminated by the intestinal tract. 1 See AYoodyatt's paper read before the Association of American Physicians, Mill), for a fresh discu-sion of this point. .lour. Am. Mod. Assn., l'.)l('), Ixvi, p. HMO. - See p. :r>:i '< s."> 1 ram> if 1-^ram nitrogen is taken to represent (i grains hoili/ protein the more correct factor. See Benedict. Carnegie Pnli. L'IKi, p. -100. '('.eihardi and Schloin^er: Arch. f. Path. 11. I'har., ISO!), xlii, p. 100. r ',Io-lin: Jour. Med. Research, 1'JOl, i, p. :i()(i. NATURE OF DIABETIC ACIDOK1S 1(>3 Another important, and indeed a most remarkable means of defense of the body against aeidosis lies in its power to excrete nitrogen in the form of ammonia instead of urea, as in health. How efficiently Nature works is evident when it is realized that 1 gram of ammonia (NII ;! molecular weight 17) can neutralize five times as much /3-oxybutyric acid as 1 gram sodium bicarbonate (XaIICO 3 molecular weight . . 19 1 7.2 NO') . 4S 12 10.0 3S7 . . U l.(J*i ,S23 . 4S 1 G.7* 391 . . 50 fill raced *30 . 45 17.0 to:} . . 55 0.1*- S5S . 05 1.5 105 . . .'() 0.0 SOI . 5 1 4 2 2*9 JOG . . 10 11 0.2 803 . 32 30 4.0 129 . . -.0 o 0.4 807 . 52 20 3.7 103 . . 20 10.0 1 3.1 1003 . 4 1 3 1 . 1 07S . . 10 1 5.4 1013 . 43 -12 2.0 GS1 . . 31 5 3.3 1010 . 30 12 1.9 GS9 . . 13 12 9.2 1019 . 00 12 1 .0*1 091 . . 53 1 5.0 1031 . 45 30 13.0 092 . . 32 1 3.5 1035 . 8 1 0.7 097 . . 19 1 S.S 103S . 37 30 3.0 70.'5 . . 3,'i 2 7.3 1012 . 59 1 0. 1*3 704 . . 10 14.5 1015 . 55 4 1 .0 705 . . 32 1 2 S.5 1017 . 00 12 + 7.0 700 . . 22 4 1 .5 1002 . 53 -12 15.0 712 . . -'9 1 2.s 1071 . 14 1 1 .0 713 . . 37 5.0 1072 . 03 -12 S.O 710 . . 11 3 1.9 107s . ' 4s 1 2 . 9 7-> k > 10 5 3. 1 1095 . 47 1 3 . 3 7'*5 32 1 . 2* 4 : 109S . . 40 24 7.0 72s 511 3.5 1100 . ',0 1 0.9 730 . . 37 13.3 14133 . 50 3 0.4 731 . . 7 10 l.S*i 137 . .14 2 0.3 74* . . 11 3.3 115 . 00 1 7 7 719 . . 11 19.0*=" 157 . 5S 30 0.0 750 . . 3 0.2* 15s . . , 5s 19.0 759 . . 10 12.7* 1 102 . j 5 4 1.4 700 . . 4 1 1.3* 1 211 . 70 -12 4.0 77S - . j . 33 1 0.1 225 . 43 12 1 .0 * Signifies fatal. ' C'oma. 4 Hemiplegia. fj Pneumonia. 8 Carbuncle. 9 Tuberculosis. 2 C.antzrene. 6 Cancer stomach. 10 Cardiac. 3 Enlarged prostate. 7 Abdominal infection. 11 Under 1 year. THE EXAMINATION OF THE URINE 169 (3) Other Causes. The other causes which led to the examination of urine were those usually recorded in the symptomatology of diabetes, namely, polyuria, polydipsia, loss of weight, weakness, pruritus, pains in back or legs, but almost never polyphagia. 2. The Mixed Twenty-four-hour Quantity of Urine should be Examined. The mixed twenty-four-hour quantity of urine should be examined. Many cases of diabetes show sugar in the urine at only one period of the day. The urine which 1 have found most apt to contain sugar is that voided after the noonday meal. This might be explained by the hearty meal, the absence of exercise following it, and the fact that this is the business period of the day, hence the patient is under excitement. Case No. 30 showed 0.3 per cent, of sugar in the urine on March 20, 1902. Gradually the specimens of urine voided became sugar-free, but that after lunch began to be constantly sugar-free November 7, 190S. The last urinary examination of this case was December 17, 1914, at which time the patient reported he was taking 100 to 125 grams carbo- hydrate, and the urinary nitrogen amounted to 13.2 grams. The urine least apt to contain sugar is that voided on rising. This is not strange, for the urine is voided several hours after the last meal upon retiring and the subsequent urine is a fasting urine. No patient with or without diabetes is to be considered sugar-free unless the twenty-four-hour specimen of urine is proved free from sugar. Case No. 978 showed no sugar in a single specimen, but in a twenty-four-hour amount there was 0.0 per cent. In illustration of the above I would mention Case No. 473, who showed 7 per cent, of sugar in January, 1912, but the urine was pronounced sugar-free in the preceding November and December, although the patient was losing weight rapidly in those months. Inves- tigation disclosed that only the morning specimen of urine had been examined on either occasion. Case No. 4S8 was accepted by an insurance company who depended upon the examination of a specimen of urine voided before dinner. He was refused shortly after by a second insurance company, who found sugar in the specimen voided after dinner. 3. The Volume of Urine in Twenty-four Hours. The quantity of urine frequently coincides with the quantity of sugar eliminated and von Noorden 1 gives the following table: TABLE 92. VOLUME OF URINE, SPECIFIC GRAVITY AND PERCENTAGE OF SUGAR COMPARED. Specific Sugar c.c. gravity. percentage. 1500- 2,500 1025-1030 2-3 2500- 4,000 1030-1030 3-5 4000- 0.000 1032-1040 4-7 6000-10,000 1036-1040 6-9 1 Von Xoordeii: Loc. cit., p. 125. 170 1'ItIXE, BLOOD AM) ItESI'l If AT/OX IN DIABETES But the volume of urine may give little index of the seventy of diabetes. ( 'ases of diabetes deeipiens are most common. ( 'ase Xo. S at onset of treatment showed 1030 c.c. of urine with 5 per cent, of sugar, and twelve years later 112") e.e. of urine with 2.4 per cent, of sugar. Case Xo. 340 showed o.S per cent, of sugar in iS(')f) c.c. of urine, ('ase Xo. 3~)l> showed o.S per cent, of sugar in 103") c.c. of urine. In other words, the twenty-four-hour quantity of urine may be normal and yet contain a large amount of sugar. A patient's report that he is voiding a normal quantity of urine, therefore, is no excuse for the neglect of a urinary examination. However, it is unusual for the urine to be normal in quantity unless the patient is sugar-free. There arc other exceptions to Table 92. Severe cases of diabetes passing through a period of restricted diet show a steady and daily diminution in the quantity of urine consistent with the restriction of carbohydrate, provided acidosis is absent. On the other hand, if acidosis appears, it is occasionally found that the volume of urine fails to decrease and may rise. The quantity of urine may be quite independent of the amount of sugar which is excreted. The emaciated Case Xo. 1190 was sugar-free yet voided over 3000 e.e. urine for days and the only cause found was increased nitrogenous metabolism despite a low protein intake. (See p. 379.) Case Xo. 9S2 was also sugar-five, but his increased volume was explained by the abnormal intake of salt. (See p. 284.) TABI.I: 9o. CIIAHT OF CASK Xo. 347. Dat Urii 1010 .!' ilv 11 : 1000 SI. + 11 12o() ll.S Hi 1 .">( H ) (I is 1000 L'O 1 .">( K ) 21 i:>7.~> ' The volume of urine of Case Xo. 3-1* changed little as the sugar decreased, despite renal involvement. Onset March 17, 1909 (acute), first seen -Inly 11, 1910. Blood-pressure at that time 170; the arteries hospital upon the patient's own initiative THE EXAMINATION OF THE URINE 171 showed moderate sclerosis. The heart extended well out toward the anterior axillary line, the liver was three fingerbreadths below the eostal margin and the spleen was palpable at the eostal margin. The patient became sugar-free quite promptly without the de- velopment of acidosis, as will be seen from Table 93. Soon after leaving the hospital he omitted alkali and restricted his diet still further, and on August 3, 1910, cardiac failure, accompanied by an increased quantity of albumin and acidosis appeared, and he died on August 19, 1910. This case illustrates the necessity of care in the restriction of the diet of patients with impaired renal functions. A nearly normal quantity of urine, even with a large percentage of sugar, when patients first come for treatment is usually a favor- able prognostic sign. On the other hand, a marked polyuria is by no means always an unfavorable omen. Case Xo. 5X5 shows this. Onset of diabetes at forty-nine years of age, first seen at fifty-two years and eleven months on February 13, 1913. Table 94 shows that a large volume of urine with a high percentage of sugar in a woman with diabetes of three years' duration need not necessarily be regarded unfavorably. TABLE 94. CHART or CASE Xo. 585. VOLUME OF URINE LARGE, YET COURSE OF CASE FAVORABLE. Date, Volume, ': Specific c.c. gravity. Diacetic Nitrogen, acid. gnis. Ammonia. Sujiar, total total Kins. finis. 1913 Feb. 21 ... ftfoO 1044 500 Mar. 9 ... 2250 1030 1 20 23 . . . 2000 1033 OS July 1 ... 2115 1040 140 AUK. 4 ... 2000 1025 - 11 ... 1500 1020 13 IS ... IGsO 1027 2 Nov. 4 ... 2000 1030 SI. + 18 1.30 27 1914 Summer Successfully shipwrecked off the const of Ireland. 1916 Jan. 17 ... 1000 1020 19 ()' 1017 Jan. 2 ... 2000 1029 14 50 The greatest volume of urine in twenty-four hours in comparison to the weight of the patient was voided by Case Xo. 1151, who developed diabetes at the age of three years and came under my care at the age of ten years weighing IS. (> kilos in October, 191C>. During the first fifteen hours at the hospital the volume of urine 1 Carbohydrate in diet, 40 grams 172 1'ItIXK, BLOOD AXD REXI'I h'ATIOX IX DIABETES THE EXAMINATION OF THE URINE 173 was 7200 + c.c., and for the following twenty-four hours 7000 c.c. The course of the ease is shown in Fig. 2. It will be noted that during the first sixteen hours the little boy was in the hospital he voided 7200 e.c. of urine, or 39 per cent, of his weight. If one should calculate on the same basis the total twenty-four-hour quantity of urine, it would amount to 10,SOO c.c., or 58 per cent, of the body weight. The volume of urine should be expressed in cubic centimeters. This enables the percentage of the twenty-four-hour amount of sugar to be most readily calculated. An ounce of urine is actually 29. () c.c., but I usually reckon it in clinical work as 30 c.c., for the errors in collection of urine more than offset the trifling error in the equivalent. One quart of urine is equivalent to 940 c.c. From experience with patients, I believe accurate enough figures are obtainable, except for scientific experiments, if we consider a quart of urine 1000 c.c., and I am content with this rule because the error comes in reporting too little rather than too much urine. Naturally, such methods are absolutely barred when accurate work is being done. At the New England Deaconess Hospital bottles graduated to 50 cubic centimeters are exclusively employed. These obviate the necessity of measuring the urine. 1 strongly advise the use of the metric system, both in recording the urine and in computing the quantity of sugar, and, indeed, for the diet as well. Most patients are glad to adopt it. The avoirdupois system involves too much labor. I recall few, if any, instances of a physician who was accus- tomed to record the volume of the urine and quantity of sugar by the avoidupois system who knew the total amount of sugar voided by his patient in twenty-four hours. How could such a physician estimate the quantity of carbohydrate in the diet in grains to the ounce? 4. The Specific Gravity. The specific gravity of the urine in diabetes is usually high, but Case Xo. 3X showed sugar in the urine when the specific gravity was 1007. A low specific gravity, there- fore, is no more excuse for neglecting to examine the urine for sugar than is a normal quantity of urine. The specific gravity of the urine has taken on increased significance since it has been shown that variations in the specific gravity during the day indicate very satisfactorily the functional power of the kidney. Perhaps no test of renal function is of so great value for the general practitioner or so simply performed. The fixation of specific gravity is well illustrated in Table 95. The patient, Case No. 10X0, was a man, aged sixty-nine years, with piostatic obstruction, and the following test was made 90 days subsequent to the removal of the prostate gland by Dr. A. L. Chute. It will be seen that during the whole twenty-four hours 174 r/i'/A'/',', HL(H)I) AM) RKM'IRATIOX IN DIAHKTKH tlie specific gravity varied from lOOSto 1012. Along with the figures for the 1 specific gravity are included the>se > for salt and nitrogen, and it will be encouraging te> the practising physician te> observe that the 1 information disclosed by the- spe-e-ifie' gravity was ejuite 1 as valuable as that obtained by these- other analyses, which required so much time. The constancy in the percentage of salt and nitrogen in the urine, however, is most striking. For further details concern- ing this case see page 211. TAIU.K '.to. Tin; FIXATION OF Si-iccinc (!KAVITV. CASK No. 10X0, p. 17)). Time. Volume, c.c. Oct. :;i, ntiti. Salt. Nit,- Specific .KO... ^r:ivity. Per cent Grams. IVr cent,. Crams. X to 10 A.M ' I-'50 1012 ().4. - { ; 0.71 0..',2 O.G7 10 to 12 " .... 1:50 loos o.r>:; 2.0 o.:-!2 t.3 12 to 2 I-..M 200 1012 0.4 1.2 O.ol 1 . 1 2 to 4 " .... I.").', 1010 0.4 ().! 0.57 0.9 4 to (i " .... 210 1010 0.42 1.4 0.4.S 1.1(5 to X " .... 270 1012 0.47 1 .4 0.47 1.2fi x to x A.M 1 1-">0 1012 0./17 4.1 0.47 5.4 Totals .... 2.~>7.~> 12.77 11.70 5. Tests for Glucose. (Dextrose, Cell^Oe). It is the ])resence or absence of glucose in the urine 1 which is important rather than the percentage of sugar, though a knowledge of the latter is essential for careful work. The quantity of sugar in the urine should be recorded in per cent, and in grains for the twenty- four hours. The clinic upon diabetes given by Friedrich vonMiiller at the Boston ( 'ity Hospital, in which he illustrated the total amount of sugar voided by the patient by exhibiting an equivalent amount of cane-sugar, was most instructive. It is not the percentage of sugar which is important in the urines of diabetic patients, it is the "total quantity of sugar eliminated in the twenty-four hours. (a) Qualitative Tests. Many of the 1 qualitative tests for glucose; are 1 excellent, and nearly all have 1 the 1 advantage that although sugar is pre-se-nt in the 1 urine' of normal individuals, they fail to demonstrate its presence unle-ss the 1 sugar exists in a greater than normal amount. I ntil recently I have found Fehling's te-st to be 1 most generally useful. Lately, 1 have used it less because 1 the Benedict test reejuircs a single- solution, keeps indefinitely and the 1 re-action offers less chance 1 for e-rror. I'nfortunately it is a little- too dedicate- and repeatedly traces of sugar in supposedly normal urines have been reported. It will be- interesting to note- whether it is adopted as a routine 1 te-st by life insurance companies. THE EXAMINATION OF THE URINE 175 (1) ' Fehlin r/'.v Test. In performing the test, 3 to 5 c.c. of equal quantities of the eopper solution and the alkaline solution are mixed in a test-tube and thoroughly boiled. If no reduetion takes place one-half as much urine as the reagent employed is then added and the whole boiled vigorously again. A yelhnv or red precipi- tate indicates the presence of sugar; a greenish precipitate may or may not indicate sugar. Occasionally substances in the urine other than sugar reduce the copper upon prolonged boiling, but this is so exceptional that I consider it far safer to boil the solu- tion a second time, and when in doubt, to repeat the test without boiling. (2) Benedicts Test. 1 Five cubic centimeters, a trifle over one teaspoonful, of the Benedict solution, are placed in a test-tube and 8 to 10 drops (not more) of the urine to be examined are added. The mixture is then heated to vigorous boiling, kept at this tem- perature for three minutes, and allowed to cool spontaneously. In the presence of glucose the entire body of the solution will be filled with a precipitate, which may be greenish, yellow or red in tinge, according to whether the amount of sugar is slight or con- siderable. If the quantity of glucose be low (under 0.3 per cent.), the precipitate forms only on cooling. If no sugar be present, the solution either remains perfectly clear, or shows a faint turbidity that is blue in color, and consists of precipitated unites. The chief points to be remembered in the use of the reagent are (1) the addition of a small quantity of urine (S to 10 drops) to 5 c.c. of. the reagent, this being desired not because larger amounts of normal urine would cause reduction of the reagent, but because more delicate results are obtained by this procedure; (2) vigor- ous boiling of the solution after addition of the urine, and then allowing the mixture to cool spontaneously, and (3) if sugar be present the solution (either before or after cooling) will be filled from top to bottom with a precipitate, so that the mixture becomes opaque. Benedict (personal communication) states that the test as per- formed above will detect glucose in as low concentration as 0.01 to 0.02 per cent, provided the urine is of low dilution. The formula and directions for preparing the Benedict solution follow; they should be strictly adhered to in preparing the solution: Copper sulphate (pure crystallized) 17.3 Sodium or potassium citrate 173.0 Sodium carbonate (crystallized) (one-half the weight of the anhydrous salt may be used) 200.0 Distilled water to make 1000.0 1 Benedict: Jour. Am. Med. Assn., 1011, Ivii, p. 1193. 170 URINE, BLOOD AX I) RESPIRATION IN DIABETES The citrate and carbonate are dissolved together (with the aid of heat) in about TOO e.c. of water. The mixture is then poured (through a filter), if necessary, into a larger beaker or casserole. The copper sulphate (which should be dissolved separately in about 100 c.c. of water) is then poured slowly into the first solution, with constant stirring. The mixture is then cooled and diluted to one liter. This solution keeps indefinitely. Folin 1 comments instructively upon the various qualitative tests: "The sensitiveness of the various alkaline copper solu- tions employed in testing for sugar depends probably most of all upon how little cuprous oxide they are capable of holding in solu- tion. The merit of Stanley R. Benedict's qualitative test for sugar is due in large part to its solvent effects on cuprous oxide, although it possesses the additional advantage of being but slightly reduced by creatinin. The creatinin disturbs not only by its reducing power, but also by its property of holding relatively con- siderable traces of cuprous oxide in solution." (o) Folin'x Test. A qualitative reduction test for sugar in normal human urine.- "This test depends (1) on the use of an uncom- monly sensitive alkaline copper solution, and (2) on the fact that in the presence of copper alkaline pierate, solutions are not reduced by sugar. The importance of the last-named consideration lies in the fact that it permits the removal of the creatinin and other substances with picric acid before applying the reduction test. "The reagent is made up in two solutions: "A. 5 grams of crystallized copper sulphate are dissolved in 100 c.c. of hot water and to the cooled solution are added GO to 70 c.c. of pure glycerin. " B. 125 grams of anhydrous potassium carbonate are dissolved in 400 c.c. of water. "One part of the glycerin-copper solution (A) is mixed with two parts of potassium carbonate solution (B). Only small portions should be mixed at a time, as the reagent (after mixing) does not keep but undergoes gradual reduction. "The test is made as follows: To about 10 c.c. of urine in a test-tube or small flask add about 2 grains of picric acid and about 2 grams of good quality bone-black (Kahlbaum's or Merck's blood charcoal), shake for five minutes, and filter. "Concentrated urines, which give the most trouble in testing for sugar, contain from )> to 5 ing. creatinin per cubic centimeter. By the above procedure the creatinin content is reduced to prac- tically nothing- at the most a few hnndredths milligrams per cubic centimeter being left in the filtrate. Bone-black has very 1 Folin: .lour. Hi 2 Folin: Ibid, THE EXAMINATION OF THE URINE 177 strong absorbing properties for the picrates of creatinin. By allowing the urine and picric acid to stand for a longer time (half an hour or overnight) the addition of bone-black may be omitted if desired. The filtrate in that case will contain about 0.1 mg. per cubic centimeter, a quantity too small to interfere with the test for sugar. "Add 1 or 2 c.c. of the creatinin-free filtrate to about 10 c.c. of the freshly mixed sugar reagent in a large test-tube (together with a pebble or two to prevent bumping) and boil with constant shaking (the shaking is desirable to avoid bumping and is necessary to prevent superheating and consequent reduction of the reagent on the sides of the test-tube) for one and a half minutes. If the sugar present is considerable (above the normal variations), a typical reduction is obtained. If the trace of sugar is smaller, but still rather large, the whole solution will become turbid as in Benedict's test. If no such turbidity is produced and the boiling mixture remains clear, transfer it at once (i, e., while still very hot) to a centrifuge tube and centrifuge for one or two minutes. Typical red cuprous oxide such as obtained with pure sugar solutions will be found in the bottom of the centrifuge tube below the green crystalline potassium picrate which usually forms as the liquid cools. "Some copper reagents made as described above give a slight cuprous oxide sediment when boiled alone, i. e,. without any added sugar or urine. When that is the case the reagent must be boiled and centrifuged once before using it for the test." A method for the determination of small amounts of sugar in the urine has recently been devised by V. C. Myers. 1 (1)) Multiple Qualitative Tests. The performance of a quali- tative test for sugar in the urine requires scarcely three minutes, but when there are ten to twenty urines to be examined the amount of time consumed is considerable. Time thus spent is wasted. With this in mind my assistant, B. H. Ragle, 2 undertook to examine simultaneously several urines qualitatively for sugar by utilizing a water-bath with perforated top. Later he found that Prof. Victor Myers and Dr. Morris Fine had already described a bath filled with saturated solution of calcium chloride for a similar purpose. 3 (These writers also describe various practical methods for routine urinary examinations in hositals. 4 ) It would appear that the water-bath filled with water is equally as good. A 1 Myers: Loc. fit., p. 17. 2 Ragle: Boston Med. and Surg. Jour., 1915, clxiii, p. 746. 3 Myers and Fine: Essentials of Pathological Chemistry, reprinted from the Post-Graduate, 1912-1913, New York, p. 127. 4 Myers and Fine: New York Med. Jour., 1913, xcvii, p. 1126. 12 178 1'RIXK, BLOOD A XI) RESPIRATION IN DIABETES water-bath, with a top perforated for test-tubes, either round 01 rectangular in type is suitable. If Bent-diet's solution is used the directions are as follows: The water-bath, filled with )> cm. water, is placed over a large flame. While the water is coining to a boil about 5 c.c. of the reagent are introduced into the same number of test-tubes as there are urines to be examined. A very handy way is to have the tubes in a rack, and the solution can be siphoned into them from a bottle. Eight drops of urine are now put into each test-tube and the tubes immedi- ately transferred to the actively boiling water-bath and left for five minutes. Within thirty seconds to one minute urines containing ().") per cent, or more of sugar will have reduced the copper, and in five minutes any specimen with a pathological trace of dextrose will give a positive reaction. The method is just as applicable to Eehling's qualitative test. After the preliminary proof that the Fehling's solution is not reduced by boiling the tests are carried out in the ordinary manner and placed in the boiling bath for five minutes. The advantages of this method are that multiple qualitative tests for sugar may be made simultaneously with little labor, bumping is avoided, and test-tubes are less apt to be broken. 1 cannot too strongly recommend this simple procedure. Dr. John A. Peterson, of llingham, tells me that he lias used this method for individual sugar tests, and I have adopted it as a routine method for patients to use at home. They call it the "teapot method," because a little teapot, reserved for the purpose, is so convenient for the test and the kitchen stove is far safer than to heat a test-tube over an alcohol lamp. (c) Quantitative Tests.- All quantitative tests for glucose in the urine are as unsatisfactory as the qualitative tests are satis- factory. It is one of the chief advantages of modern treatment that the need for these tests is nearly abolished. It will be one of the disadvantages of modern treatment if we introduce a mul- tiplicity of new tests in diabetes. The simplification of the treat- ment of diabetes means everything to the practitioner and patient. The best quantitative test for sugar for physicians who do not devote unusual attention to diabetes is the fermentation test. (1) Fermcnfntion 7V.v/.- To 100 c.c. of urine of known specific gravity, one-fourth of a fresh yeast cake, thoroughly broken up, is added and the whole is set away at a temperature of S.V 3 to 5).") V. Twenty-four hours later the urine is tested with Fehling's or Benedict's solutions. If a reduction is obtained it is set aside for further fermentation. Complete fermentation having been proved the specific gravity is taken after the urine has acquired its original (room) temperature, The difference in specific gravity multiplied THE EXAMINATION OF THE URIXE 179 by 0.23 gives the percentage. In the performance of the fermenta- tion test for sugar a few crystals of tartaric acid should be added whenever the urine is alkaline. If the temperature of the urine (room) is 76 F. when the specific gravity is taken at the begin- ning and end of the test the result will be still more accurate. The physician who expends two cents of his own money for a yeast cake to perform this simple quantitative test for sugar usually knows far more about the treatment of diabetes than his brother practi- tioner who expends $10 of his patient's money fora laboratory report. The test is most valuable in determining the presence or absence of traces of sugar. At the end of twenty-four hours, if the filtered urine which previously showed a doubtful test for sugar with Feh- ling's or Benedict's solutions shows no reduction upon repetition of the test, it can be assumed that a fermentable substance, pre- sumably glucose, was present. If further doubt exists, the use of the polariscope and the phenylhydrazin tests will help to settle the question. (2) Polar iscopi/. The determination of the percentage of sugar by means of the polariscope is the most convenient of all quantita- tive tests for glucose. The test is easily performed during the patient's visit and the result so obtained can be immediately utilized for treatment. The quantity of urine required for the test is small and quantitative solutions are not needed. It is not usually recognized that in the vast majority of cases simple filtration suf- ficiently clarifies the urine to allow of its examination. This saves a few minutes' time in one examination, but the aggregate of many hours in the course of a year. Should it be necessary to decolorize the urine, use infusorial earth. Magnus-Levy has pointed out that the employment of lead acetate introduces an error. Unfor- tunately in the presence of /3-oxybutyric acid the polariscope gives too low a reading, for /3-oxybutyric acid is levorotatory, whereas glucose (dextrose) is dextrorotatory. This error may be considerable, even amounting to as much as 57 grams sugar in twenty-four hours in Case Xo. 235. To obviate this the levorotatory power of the urine can be deter- mined after fermentation, and this fraction of a percentage added to the percentage of sugar obtained before fermentation. But this calculation is not very satisfactory and, beskles, requires a long time. Whenever, therefore, /3-oxybutyric acid is present and accurate knowledge of the percentage of sugar in the urine is desired the Benedict test should be performed. In general the promptness with which the determinations with the polariscope can be made discount this error in clinical work. (3) Fehling's Test. I have little faith in this test, because it is so difficult to determine the end-point. Various modi- 180 URINE, BLOOD AND RESPIRATION IN DIABETES fieations of the tost have appeared, and at one time and another I have employed some of them only later to discard them. At present I use instead the quantitative test described by Stanley R. Benedict. Fehling's quantitative method, however, is as fol- lows: Dilute 10 e.c. of Fehling's solution (5 c.c. copper sulphate and 5 c.c. alkaline tartrates) with 40 c.c. of water and boil in a flask to test the solution. Add slowly, by means of a burette, the urine diluted with water 1 : 10. After each addition boil briskly. When the point is reached at which the blue color wholly disappears the copper is reduced. Ten c.c. of Fehling's solution are reduced by 0.05 gram of glucose, which is, therefore, the amount of glucose in the urine used. The percentage of glucose equals 5, divided by the number of cubic centimeters of undiluted urine. To get accurate results with this method, the dilution of the urine or the amount of water added to the Fehling's solution in the flask should be such that about 100 c.c. of fluid are present at the completion of the test. (4) Benedict's Test} "Like Fehling's quantitative process, the method is based on the fact that in alkaline solution a given quantity of glucose reduces a definite amount of copper, thus decolorizing a certain amount of copper solution. The copper is, however, pre- cipitated as cuprous sulphocyanate, a snow-white compound, which is an aid to accurate observation of the disappearance of the last trace of color. The solution for quantitative work, which keeps indefinitely, has the following composition: " Pure crystallized copper sulphate, 18 grams. "Crystallized sodium carbonate, 200 grams (or 100 grams of the anhydrous salt). "Sodium or potassium citrate, 200 grains. "Potassium sulphocyanide, 125 grams. " Five per cent, potassium ferrocyanide solution, 5 c.c. "Distilled water to make a total volume of 1000 c.c. "With the aid of heat dissolve the carbonate, citrate, and sulpho- cyanide in enough water to make about 800 c.c. of the mixture and filter if necessary. Dissolve the copper sulphate separately in about 100 c.c. of water and pour the solution into the other liquid, with constant stirring. Add the ferrocyanide solution, cool and dilute to exactly one liter. Of the various constituents the copper salt only need be weighed with exactness. Twenty-five c.c. of the reagent an' reduced by 50 ing. (0.050 grain) of glucose." The procedure for the estimation is as follows: "The urine, 10 c.c. of which should be diluted with water to 100 c.c. (unless 'Benedict, S. !{.: The Quantitative Estimation of Glucose in the Urine, Jour. Am. Med. Assn., Chicago, I'.Ul, Ivii, p. 1193. THE EXAMINATION OF THE URINE 181 the sugar content is believed to be low), is poured into a 50 c.c. burette up to the zero mark. Twenty-five c.c. of the reagent are measured with a pipette into a porcelain evaporating dish (10 to 15 cm. in diameter), 10 to 20 grams of crystallized sodium car- bonate (or one-half the weight of the anhydrous salt) are added together with a small quantity of powdered pumice stone or talcum, and the mixture heated to boiling over a free flame until the car- bonate has entirely dissolved. The diluted urine is now run in from the burette, rather rapidly, until a chalk-white precipitate forms and the blue color of the mixture begins to lessen percep- tibly, after which the solution from the burette must be run in, a few drops at a time, until the disap- pearance of the last trace of blue color which marks the end-point. The solution must be kept vigor- ously boiling throughout the entire titration." If the mixture becomes too concentrated during the process, water may be added from time to time to replace the volume lost by evaporation; however, too much emphasis cannot be placed upon the fact that the solution should never be diluted before or during the process to more than the original 25 c.c. Moreover, it will be found that in titrating concentrated urines, or urines with small amounts of sugar, a muddy brown or greenish color appears and obscures the end-point en- tirely. Should this be the case the addition of about 10 grams of calcium carbonate does away with this difficulty. The calculation of the percentage of sugar in the original sample of urine is very simple. The 25 c.c. of copper solu- tion are reduced by exactly 0.050 gram of glucose. Therefore the FIG. 3. Apparatus required for a simplified quantitative Benedict test (p. 182). 182 ('HIM-:, BLOOD AM) REHPI RATIOS IX DIAKETES volume of diluted urine dra\\'ii out of the burette to eil'eet the reduction contains ~>() nig. of sugar. \Yhen the urine is diluted 1 To 10, as in the usual titration of diabetic urines, the formula for calculating the percentage of sugar is the following: X 1000 -- pei'centa^' in the original sani])le, wherein ,r is tlu > number of <:ubic centimeters of the diluted urine required to reduce 2o c.c. of the copper solution. "In the use of this method chloroform must not be present dur- ing the titration. if used as a preservative in the urine it may be removed by boiling a sample for a few minute's, and then diluting to the original volume." My laboratory assistant, Miss Kvelyn Warren, suggested the use of white enamelware dishes instead of porcelain in the performance of the Benedict quantitative test. These serve the purpose admir- ably and the saving in expense is very considerable. J recommend their adoption in the laboratories of large hospitals quite as strongly as for smaller laboratories of physicians. The equipment of a diabetic nurse includes this apparatus (Fig. o) and the rules given for the performance of the test are as follows: Jii'LKs FOR NntsKs ix PERFORMING THE SIMPLIFIED QUANTITATIVE BEXEDKT TEST. Articles Required, Ten c.c. graduated pipette; small white enamelware dish, ',] inches across, 2 inches deep; sodium carbonate; talcum. The test can be performed by the aid of a kitchen gas burner. If the gas burner is not a small one and so flares up around the edges of the dish, put an asbestos plate over it or simply an iron cover. 1. Place ."> c.c. of the quantitative Benedict solution in the dish. 2. Add less than one-fourth teaspoonful of sodium carbonate. IS. Add one-half as much talcum. 4. Add about 10 c.c. water. ."). Dilute 1 part urine with parts of water unless the quantity of sugar is low. (A low percentage of sugar is shown by the qualita- tive Benedict test turning green instead of yellow, \\ith small quantities of sugar it is unnecessary to dilute the urine.) ('). Bring the contents of the dish to a boil, maintain in this con- dition and then add, drop by drop, the urine from the graduated pipette until the blue color has entirely disappeared. I poll the THE EX AM IX AT I OX OF THE URINE 1S3 first trial too much may be added, and therefore, having noted the approximate quantity of urine required to reach the end-point, invariably repeat the test as a control. Calculation. Five c.c. of the Benedict quantitative copper solution are reduced by 0.01 gram glucose. Consequently, the quantity of undiluted urine required to reduce the o c.c. Benedict solution contains 0.01 gram glucose. - X 100 = per cent. x = c.c. of undiluted urine. .r Example. Fifteen hundred c.c. urine in twenty-four hours. Five c.c. used to reduce (decolorize) Benedict solution. C l x 100 = 0.2 per cent. o 1")00 X 0.002 (0.2 per cent.) = 3 grams sugar i i twenty-four hours. Example. If the urine had been diluted with !) parts water, in other words, 10 times, the calculation would be: o c.c dilute 1 urine. = O.o c.c. actual urine. (] - X 100 = 2 per cent. 1500 X 0.02 (2 per cent.) = 30 grams sugar in twenty-four hours. (5) Peter's Test. A most accurate method for the quantitative determination of glucose in the urine has been devised by Dr. A. AY. Peters, and has been used with satisfaction at the Nutrition Labora- tory. It is unfortunately more complicated than the practitioner would care to adopt. 1 (i. Tests for Other Sugars. Other sugars than grape-sugar are occasionally found in the urine. It is rare that they cause confu- sion. It is not quite so easy to detect or to exclude them, as would appear from the description of the tests by which they are said to be identified. Of these various sugars lactose is the one most commonly encountered. (a) Lactose (('^HoaOn). i.actose in the urine may give rise to confusion in the performance of Fchling's or Benedict's tests. Fortunately, the conditions in which it is liable to occur, pregnancy and the lactation period, are usually known to the physician, and it is then not considered of significance. It has also been found in the urines of nurslings. Lactose, like glucose, reduces copper, is dextrorotatory, but it yields a characteristic osazone and does 'Sec Benedict and Jo^lin: Carnegie Publication 17G, p. N; also Jour. Am. Them. Soc., 1912, xxxiv, p. 928. 184 URINE, BLOOD AND RESPIRATION IN DIABETES not ferment with pure yeast (saccharomyces apiculatus). However, the osa/one is very difficult to obtain from the urine and ordinary yeast is not to be depended upon for the fermentation test. Jviibner's test is the most practicable. To 5 c.e. of urine in a test-tube, add 1 or 2 grains of lead acetate. Heat until bubbles appear and then add ammonium hydrate until the color changes. A deej) yellow or brown color in the solution is distinctive of lactose, whereas in the presence of dextrose the color is a cherry red. Prac- tically no test for lactose is satisfactory. (//) Pentcse (CJIioOa). Pentose occasionally is present in the urine, and probably will be encountered more frequently in the future now that attention has been called to it. It may be suspected when Fehling's test, performed in the ordinary way, fails to show reduction upon the second boiling, but later suddenly causes a partial reduc- tion. Pentose neither ferments nor is it optically active. The orcine test for pentose is as follows: To 3 c.c. of urine are added G c.c. fuming IK! and a knife-point full of orcine. Boil. If a bluish- green color promptly appears, it is characteristic of pentose. Urines containing glucose and lactose do not give this test, but urine con- taining glycuronic acid produces a similar color, although only after prolonged boiling, and even then the precipitate is never greenish blue, but more of a violet blue. (f) Levulose (C 6 II 12 O 6 ). Levulose is frequently present in the urine of severe diabetics. It is not easily identified. It is levoro- tatory, but so is /3-oxybutyric acid, which is found under similar conditions. Levulose ferments with yeast, gives a positive Feh- ling and Benedict test, and yields the same osa/one as does dex- trose with phenylhydra/in. Levulose can, however, be differenti- ated by the Seliwanoff's reaction, provided the urine does not con- tain too high a percentage of dextrose. SeliwanofY's test is as follows: To about f) c.c. of urine in a test-tube add 1 c.c. of con- centrated hydrochloric acid and a few crystals of resorcin. Heat until boiling begins. Upon cooling if levulose is present a red substance precipitates, which is soluble in alcohol. A new method for the determination of levulose in the presence of dextrose is described by Loewe. 1 (d) Maltose (C^IL-jOn). Maltose very rarely occurs in human urine, and has not been shown to be of clinical significance. Maltose is powerfully dextrorotatory, completely fermented by yeast, reduces copper slowly, and yields a characteristic osa/one. (c) Glycuronic Acid ((\\\ V} (.)t) . Glycuronic acid as such is not found in fresh urines, but conjugated glycuronic acids occurring in the urine spontaneously decompose and may cause confusion. Such 1 LoewojProc. Soc. Exp. Biol. and Mod., 1016, xiii, p. 71. THE EXAMINATION OF THE URINE 185 conjugated glycuronic acids only appear after the ingestion of chloral hydrate, camphor, menthol, turpentine, or phenol in large enough quantities to be of significance. If this point is borne in mind confusion will not arise. Glycuronic acid reduces copper and bismuth, but does not ferment. It may be difficult to detect in the presence of pentose, although one can rely on the characteristic osazone of pentose if differentiation becomes necessary. (/) Substances Found in the Urine which give Rise to Confusion in Testing for Sugar. These may be divided into two groups: those in normal urines and those in pathological urines. Of those met with in normal urines there are creatinin, especially if the urine is con- centrated, earthy phosphates, uric acid, xanthin bases, peptone, glycuronic and glycosuric acids, mucus, indoxyl sulphates, and unites. These decolorize the blue solution, but do not give the red precipitate, which cannot be mistaken. In pathological urines albumin is the most troublesome, but this is easily counteracted by precipitating the albumin by the Esbach test and subsequently filtration. In doubtful cases, precipitation of uric acid, xanthin, hypoxanthin, creatins, and phosphates may be accomplished by adding 5 c.c. of 10 per cent, solution of sodium acetate. The test is then applied to the filtrate. 7. Methods for the Determination of the Urinary Acids. (a) Qualitative Tests. (1) Diaceiic Acid (CII 3 COCH 2 COOH). The simplest method for the detection of acidosis by urinary examination is Gerhardt's ferric chloride reaction for diacetic acid. The test may be performed as follows: To about 10 c.c. of the fresh urine carefully acid a few drops of an undiluted aqueous solution of ferric chloride, Liquor Ferri Chloric!!, I . S. P. A precipitate of ferric phosphate first forms, but upon the addition of a few more drops is dissolved. The depth of the Burgundy-red color obtained is an index to the quantity of diacetic acid present. I record the intensity of the reaction as follows : +, ++, + + + , or + + + +. Attempts are sometimes made to arrive at approximate quanti- tative values of acid bodies by the depth of the Burgundy-red color obtained in Gerhardt's test for diacetic acid. Liithje pointed out the fallacies of this method and Table 96 shows how unreliable Benedict and I have also found it to be. 1 TABLE 90. COMPARISON OF DIACETIC AC-ID AND /S-OXYHUTYRIC ACID. Diacetic acid, p-oxybutyric arid, symbol. grains. 5. 7 to 11.0 + 7.0 to 14. 2 + + 8. 5 to 55. 3 + + + 13. 3 to 51.0 + + + + 17. (> to 36. 8 1 Benedict and Joslin: Carnegie Institution of Washington, Publication 130, p. 25. ISO n?I\E, HLOOI) AM) RKHPIRATIOX IN DIAIIETES It should not be forgotten that if a patient is taking salicy- lates, antipyrin, cyanates, or acetates, the foregoing test will give a similar reaction, but one that cannot be mistaken if the solution is boiled for two minutes. Diacetic acid is unstable and the color disappears if due to it, but if due to the above substances the color is unchanged. Diabetic patients often take salicylates for pain of one kind or another, and therefore one must always be on the watch for this possibility. I have seen a patient on the verge of coma who was taking salicylates, and at the start confusion arose as to whether the ferric chloride reaction was due to diacetic acid or to salicyluric acid. With the test for diacetic acid the physician must be absolutely at home. It is doubtful if one physician in five hundred will employ any better test for acidosis, and only rarely is any other test necessary. In confirmation of this state- ment 1 cite the data on p. 210 which show the close correspondence between the results of the ferric chloride test and determinations of the ( 'O-2 in blood and alveolar air. (2) Acetone (CHaCOnis). The test for acetone was the first employed for the detection of acid poisoning, but the small role which Folin 1 has shown that acetone plays in the total acidosis led me to discard it, believing it better to concentrate time upon the quantitative estimation of the acidosis than to use several qualita- tive tests. Folin demonstrated that most of the substance supposed to be acetone in the urine is really diacetic acid, and that Legal's test for acetone is really a very delicate test for diacetic acid. Folin also pointed out that the so-called acetone odor of the diabetic's breath is in reality not due to acetone.- Weiland, quoted by L. lilum, :! says that acetone may amount to 1.07 grams and yet the (ierhardt test for diacetic acid be negative, while at other times the presence of 0.1 gram of acetone in the twenty-four hours is sufficient to make the Gerhardt test positive. The different tests for acetone are in reality tests for diacetic acid. Legal's test is as follows: A few crystals of sodium nitro- prusside are dissolved in 5 c.c. of urine, which is then rendered alkaline with sodium hydrate. A few drops of glacial acetic acid are then slowly added and a distinct purple color appears, which, if the test-tube is shaken, is best seen in the foam. CM (3-o.ri/biifiinr Arid (CII,( < II(O11),('II,C'OOII). There is no simple qualitative test for /3-oxybutyric acid. (b) Quantitative Tests.- The determination of the extent of the acidosis is of prime importance in the treatment of any severe case of diabetes. 1 sympathize with any physician who must treat a THE EXAMINATION OF THE VRIXE 187 severe case of diabetes without a knowledge of the degree of acid poisoning present. Fortunately, comparatively simple methods are at hand which are quite satisfactory in routine treatment; b"+ even most of these simple methods are too complicated for a physician with a large practice who has. only a few eases of diabetes in the course of a year. (1) Reaction of Urine. The most easily performed of the urinary tests are concerned with the reaction of the urine. Although the urines of normal individuals are frequently neutral or alkaline, this is seldom the case with urines of diabetics, for the protein-fat diet of the diabetic favors an acid reaction; but it should be stated emphatically that an acid reaction of the urine does not necessarily imply acidosis. It has been proposed to estimate the degree of acidosis by deter- mining the quantity of alkali which it is necessary to give the patient to render the urine alkaline. I do not recommend this method, for I do not approve of giving alkalis to diabetic patients unnecessarily. It is mentioned to illustrate in a homely way the intensity of the diabetic acidosis. Ordinarily, 5 to 10 grams of sodium bicarbonate will render the normal urine alkaline, but in diabetic patients with an acidosis of moderate severity 20 to 30 grams are required, and in severe cases of acidosis 50 grams daily for a week or more make little impression, and even with doses of over 100 grams sodium bicarbonate in twenty-four hours the urines of patients in coma frequently remain acid. A table has been constructed by von Xoorden showing approximately the quantity of acidosis which can be assumed to be present when various quantities of sodium bicarbonate are required to render the urine alkaline. TABLE 97. THE RELATION' IJETXVEEX THE QUAXTITV OK SODIUM BICAU- HOXATE REQUIRED TO REXDER THE UKIXE ALKALINE AXD THE ACID BODIES PRESENT. NallCOs required to i urine alkaline, grams. 20 30 to 40 40 + 'nder Approximate acidosis in terms of jg-oxybutyric arid, grams. Under 15 20 to 30 30 to 40 I do not know of a simpler test for the estimation of acidosis in the treatment of diabetes, but since there are other tests which do not necessitate the giving of an alkali I again advise against its use. A more accurate method of determining the alkali tolerance has been described by ^ellards, 1 and also by Palmer and Henderson.' 2 1 Pollards:, Johns Hopkins Hosp. Bull., 1912, xxiii, p. 2S9. 2 Palmer ;ind Henderson: Arch. Int. Med., 1913, xii, p. 153. 188 URINE, BLOOD AND RESPIRATION IN DIABETES (2) Ammonia. The quantity of ammonia in the urine is a measure of the reaction of the body to counteract the acidosis produced in it. To this extent its estimation gives a more accurate idea of the acid production of the body than any other of the urinary tests at our disposal, which simply show the quantity of acid leaving the bodv, The test, however, becomes of less value so soon as AIR FROM WASHBOTTUE 3 -> SUCTION B Ki<;. 4. Folia's aeration method for determination of ammonia adapted for titration. Apparatus for use with suction. (Simon.) extraneous alkali is administered, because under such conditions the ingested alkali is used by the body in preference to ammonia. The normal amount of ammonia in the urine varies between 0.5 to 1 gram, and the ratio between the ammonia nitrogen to the total nitrogen in the urine is fairly constant at 1 to 25 (4 per cent.). In severe diabetes the ammonia may gradually increase, and in Case THE EXAMINATION OF THE URINE 189 No. 344 it amounted to 8 grams in one day. The nitrogen upon this same day was 19.2 grams, giving an ammonia-nitrogen nitrogen ratio of 34.3 per cent. On another day this ratio reached 44.4 per cent., but the absolute quantity of ammonia was only 4.4 grams and the nitrogen 8.7. In fact, these high ammonia-nitrogen nitro- gen ratios are ordinarily only obtained when the total quantity of nitrogen in the urine is small. The two tests which we have employed for the determination of ammonia follow. Of the urine to be tested 1 or 2 c.c. are pipetted, by means of an Ostwald pipette, into tube .1 (Fig. 4). Two or three drops of kerosene and a few drops of potassium oxalate, and potassium carbo- nate solution (which contains 15 per cent, each), are added and the stopper quickly inserted. By means of a suction pump the ammonia is drawn over into tube B, which contains 10 c.c. of j\ hydrochloric acid. Tube A is connected with a wash bottle containing 10 per cent, sulphuric acid, so that when the air is drawn through the urine it will be completely ammonia-free. For the first minute aeration should be slow, then the air current may be as rapid as the apparatus will stand. Aeration should be complete in from twenty to thirty minutes and the excess acid is titrated with T N sodium hydroxide. The amount of acid used by the ammonia times 0.0002 gives the amount of ammonia-nitrogen while the amount of acid used times 0.00024 gives the amount of ammonia in the quantity of urine (1.0 or 2.0 c.c.) used for the test and from this the amount of nitrogen or ammonia in twenty-four hours may be easily computed. The glass tube which passes nearly to the bottom of tube B is sealed at the lower end, but contains a number of small holes, which allow only fine bubbles to pass into the receiving acid and thus aids in the complete absorption of the ammonia. This method is based on that of Folin and Macallum. 1 However, they Nesslerize and determine the nitrogen by means of the Duboscq colorimeter. We have found titration reliable and simple. Honchese-Malfatti Method fur the Determination of Ammonia. (o) To 25 c.c. of urine in a 200 c.c. Frlenmeyer flask, add about 25 c.c. of distilled water, about 10 grams (1 or 2 teaspoonfuls) of powdered potassium oxalate, and a few drops of indicator (phenol- phthalein). Shake a few times to dissolve the oxalate, then titrate with tenth-normal sodium hydroxide until the first faint pink color is permanent. (b) Take 5 c.c. of commercial formalin solution in a test-tube, add a few drops of phenolphthalein indicator, and then titrate with tenth-normal sodium hydroxide until a faint pink is obtained. 1 Folin and Macallum: Jour. Biol. Chem., 1912, xi, p. 423. 100 I'lllXK, BLOOD AM) IfKM'IKATlOX L\ DIAHKTKX (<} Add this neutralized formalin to the urine, which lias just been titrated, and titrate again with tenth-normal sodium hydrox- ide until the previous pink is attain obtained. (Calculation: The number of cubic centimeters of tenth-normal alkali used in titration (<) multiplied by 0.0017 gives the number of grains of ammonia in '2~i c.c. of urine.) Xo account need be taken of the amount of sodium hydroxide used in titrutions (a) and (/>). The method depends upon the fact that formalin combines with free XI I 3 and forms hexamethylenetetramin. The ammonia is liberated from its salts by means of XAOII. (o) p-o.ry1nrti/ric Acid. The tests for /3-oxybutyric acid are all complicated, because they depend upon the extraction of the acid. Estimation of the fi-oxy butyric acid based upon the difference between the quantity of sugar as determined by Fehling's, fermen- tation and polarization tests are inaccurate. They simply suggest the presence of /3-oxybiityric acid. llhiclc'x Method. One hundred cubic centimeters of urine are measured with a pipette into an evaporating dish and made dis- tinctly alkaline by the addition of .sodium bicarbonate. The urine is then evaporated to a thick syrupy liquid (4 or f> c.c.), using the gentle heat of a water-bath or the low heat of an electric stove. After cooling the residue is made distinctly acid with strong hydro- chloric acid, and is then formed into a thick paste, and later into a porous meal by the gradual addition of plaster of Paris. This porous meal is placed in an extraction apparatus (a Soxhlet or a modification of it) and extracted with about 00 c.c. of ether for three hours. The ether extract is then transferred to an evaporating dish, where the ether is allowed to evaporate spontaneously. The residue is treated with ."> c.c. of water and 0.4 gram of bone-black added to decolorize, then filtered, and washed until perfectly clear and made up to a known volume, usually LT) c.c. '[ he (8-oxybutyric acid is determined with a polariscope, using the following formula: Shaffer'* Method for j3-o.i't/l>i/c Acid, ^Icctotic mid Di&cctic Add Combined.- The advantage of this method is that it not only enables the j3-o.\ybutyric acid to be determined, but the acetone and diacetic acid as well. 'The test is well described by Sahli: 1 " Introduce into a ")()() c.c. volumetric ilask :_'."> to LMO c.c. of the urine 1 Sahli: Diagnostic Methods, I'.tl 1, W. H. Saunders. Company, Philadelphia, trans- luted by Nathaniel Bowditch Potter. THE EXAMINATION OF THE URINE 191 to be examined. The amount of urine used will depend upon the i'.niount of (8-oxybutyric acid which is contained in it. In the ease of urines giving a strong ferric chloride reaction, 25 to 50 c.c. of urine will suffice. It is desirable to use such a volume of urine as will give 25 to 50 ing. of acetone. To the urine is added an excess of basic lead acetate and 10 c.c. of ammonium hydroxide. It is necessary to be sure that an excess of the lead salt has been added. The mixture is made up to the 500 c.c. mark, shaken thoroughly and filtered through a dry filter-paper; 200 c.c. of the filtrate are distilled after the addition of 300 to 400 c.c. of water and 15 c.c. of concentrated sulphuric acid and a little talcum to prevent bump- ing. About 200 to 250 c.c. of distillate are collected. This distillate contains, in addition to acetone from the preformed acetone and aceto-acetic acid, certain volatile acids. If it be made slightly alkaline and a second distillation made, the distillate will contain all the acetone from acetone and aceto-acetic acid, and this may be estimated by the iodin method. "The residue from the urine containing the sulphuric acid con- tains all the /3-oxybutyric acid, and must now be converted by this oxidation into acetone. The residue is diluted and 0.5 gin. of potassium dichromate added. It is then distilled from a flask provided with a dropping funnel and water added gradually to make up for the amount distilled over. Instead of water, one may use a 0.5 per cent, solution of potassium dichromate. After about 500 c.c. of distillate have been collected the distillate is redistilled after the addition of 20 c.c. of 3 per cent, hydrogen dioxide and sufficient potassium hydroxide to make the solution alkaline. Collect 300 c.c. of distillate and estimate the amount of acetone with iodine and thiosulphate ; 1 mg. of acetone is equivalent to 1 .704 nig. of (S-oxybutyric acid. 1 " (4) Acetone and Diuretic Acid (The Huppert-Messinger Quanti- tative Method).- "This method determines the sum of acetone' and the diacetic acid, which is transformed to acetone by the distillation. The solutions necessary are-: 1. Acetic acid, 50 per cent. 2. f^- iodine solution. .'). -j^- sodium thiosulphate solution. 4. A thin starch solution (1 gram of starch dissolved in 500 c.r. of boiling water). "To make up these solutions 24. S grams of crystallized sodium thiosulphate (XaaSgOa + 5H 2 0) are carefully weighed, dissolved in distilled water, and the solution made up accurately to one liter. Next 25 grams of potassium iodide are dissolved in a little 192 i'ltIXE, BLOOD AM) RESI'Ih'ATIOX IN DIABETES water, 12.7 grams of iodine added, and the solution made up to about ( JOO to !)")() e.e. To standardize tin's solution 20 c.e. of the thiosulphate solution are carefully measured into a small flask, a fe\v dro])s of the starch solution added, and then the iodine solution added from a burette with glass stop-cock until the blue color just appears. This titration is repeated several times until the amount necessary for the end-reaction is accurately determined, then to the iodine solution is added the necessary amount of water, so that 20 c.c. of this solution will exactly equal 20 c.c. of the thiosulphate solution. This dilution should be confirmed by an additional titration. "Both of these fluids are to be kept in dark glass bottles with ground-glass stopjx'rs. The iodine solution must be restandardized frequently. One cubic centimeter of the thiosulphate solution equals 0.0127 gram of iodine. The formula of the reaction is 21 + 2Xa 2 S 2 () ;i -f 2Xal = Xa 2 S.,O 6 . The first free iodine in excess will form the blue starch iodine compound. "The urine if alkaline is first just acidified with acetic acid. To 500 c.c. of acid urine (if rich in acetone, c. . 177. THE EXAMINATION OF THE URINE 105 technic is simplified to boiling the mixture for an hour and a half under a reflux condenser, and weighing the precipitate which forms. Neither the size of sample nor mode of procedure have required variation for different urines; the same process yields accurate results for the smallest significant amounts of acetone bodies and likewise for the largest that are encountered. The precipitate is crystalline and beautifully adapted to drying and accurate weigh- ing; but when facilities for weighing are absent the precipitate can be redissolved in 5 per cent, hydrochloric acid and the mercury titrated by a suitable method, such as that of Personnel Solutions llequ ired. Twenty Per Cent. Copper Sulphate. Two hundred gms. of C'uSO 4 .5 aqua dissolved in water and made up to 1 liter. Ten Per Cent. Mercuric Sulphate. Seventy-three gms. of C'.P. red mercuric oxide dissolved in 1 liter of HoSOtof 4 normal concentration. The solution of the oxide is assisted by warming on a steam bath. Fifty Volume Per Cent. Sulphuric Acid. Five hundred c.c. of sulphuric acid of 1.835 specific gravity, diluted to 1 liter with water. Concentration of HoSCX should be 17 normal. Colloidal Iron. Merck's "Dialyzed Iron" solution containing 5 per cent, of Fe 2 O.3. Ten Per Cent. Calcium Hydrate Suspension. Mix 100 gms. of Merck's fine light "Reagent" Ca(OH) 2 with 1 liter of water. Five Per Cent. Potassium Dichromate. Fifty gms. K 2 ( r 2 O 7 dis- solved in water and made up to 1 liter. Removal of Clucose and Other Interfering Substances from Urine. Place 25 c.c. of urine in a 250 c.c. measuring flask. Add 100 c.c. of water, 50 c.c. of copper sulphate solution and mix. Then add 50 c.c. of 10 per cent, calcium hydrate, shake and test with litmus. If not alkaline, add more calcium hydrate. Dilute to mark and let stand at least one-half hour for glucose to precipitate. Filter through a dry folded filter. This procedure will remove up to S per cent, of glucose. Urine containing more should be diluted enough to bring the glucose down to 8 per cent. The filtrate may be tested for glucose by boiling a little in the test-tube. A precipitate of yellow CuoO will be obtained if the removal has not been complete. White CaCos precipitate means nothing. Removal of Proteins from Blood and Llasma. Of whole blood 10 c.c. are measured into a 250 c.c. volumetric flask, half-full of water. Sixty c.c. of colloidal iron are added and mixed, then 1 c.c. of satu- rated sodium sulphate solution. The flask is filled to the mark and shaken. The contents are filtered through a dry folded filter. 1 Sutton's Volumetric Analysis, 10th edition, p. 264. 196 URINE, BLOOD AND RESPIRATION IN DIABETES For plasma the procedure is the same except that only 8 c.c. are taken in a 200 c.c. flask with 15 c.c. of colloidal iron, and 1 c.c. of saturated sodium sulphate. In the case of either whole blood or serum, 125 c.c. of filtrate, equivalent to 5 c.c. of the original sample are taken for analysis. Simultaneous Determination of Total Acetone liodiex (Acetone, Diacetic Acid and /3-oxybutyric Acid), of Urine or Blood in One Operation. Place in 500 c.c. Erlenmeyer 25 c.c. of urine filtrate plus 100 c.c. of water or 125 c.c. of blood filtrate. Add 10 c.c. of 50 per cent, sulphuric acid and 35 c.c. of the 10 per cent, mercuric sulphate. Connect the flask with a reflux condenser having a straight condensing tube of 8 or 10 mm. diameter, and heat to boiling. After boiling has begun, add 5 c.c. of the 5 per cent, dichromate through the condenser tube. Continue boiling gently one and one-half hours. The precipitate which forms consists of the mercury sulphate compound of the preformed acetone and of the acetone which has been formed by oxidation of the hydroxy- butyric acid. It is collected in a Gooch or a medium density "alundum" crucible, washed with 200 c.c. of cold water, followed by a little 95 per cent, alcohol, dried for an hour at 1 10 and weighed. Several precipitates may be collected, one above the other, without cleaning the crucible. Acetone and Diacetic Acid. These substances without the hydroxybutyric acid are determined exactly as the total acetone bodies, except that (1) no dichromate is added to oxidize the hydroxybutyric, and (2) the boiling must continue for not less than thirty-five nor more than forty-five minutes. Boiling for more than forty-five minutes splits olf a little acetone from hydroxybutyric acid even without dichromate. ^-hydroxybutyric Acid in Urine. The /3-hydroxybutyric acid alone is determined exactly as total acetone bodies except that the preformed acetone and that from the diacetic acid are first boiled off. To do this the 25 c.c. of urine filtrate plus 125 c.c. of water are treated with 2 c.c. of 50 per cent, sulphuric acid and boiled in the open flask for ten minutes. The volume of solution left in the flask is measured in a cylinder. The solution is returned to the flask, and the cylinder washed with enough water to replace part of that boiled off and bring the volume of the solution to 127 c.c. Then 8 c.c. of the 50 per cent, acid and 35 c.c. of the 10 per cent, of mercuric sulphate are added. The flask is connected under the condenser and the determination is continued as above. fi-hydroyybutyric Add in Mood.- The following procedure enables one to determine separately in a single sample of blood both the acetone plus diacetic acid and the /3-acid. The acetone and diacetic acid are precipitated as above described, and the filtrate THE EXAMINATION OF THE URINE 197 poured as completely as possible through the Gooeh or alundum crucible into a dry receiving flask. Of this filtrate !(>() c.c. are measured into another Erlenmeyer flask and 10 c.c. of water are added. The mixture is heated to boiling under a reflux condenser, 5 c.c. of dichromate solution are added, and the determination continued as described for "total acetone bodies." In case only the /3-acid is desired, or enough blood is taken for double portions, the slightly easier procedure used for urine may also be followed with blood. Factors for Calculating Acetone Bodies in Urine when 25 c.c. of Filtrate, Equivalent to 2.5 c.c. of Urine, and in Blood when 125 c.c. of Filtrate, Equivalent to 5 c.c. of Blood are Used for Determination. One mgm. of /3-hydroxybutyric acid yields 8.7 mgm. of precipitate. One mgm. of acetone yields 19.7 mgm. of precipitate. The amount of precipitate obtained from /3-hydroxybutyric acid therefore corresponds to 79 per cent, of the acetone that would be obtained if each molecule of hydroxybutyric yielded a molecule of acetone. The oxidation is complete in one and one-half hours, and the conditions are so constant that duplicates usually check within 1 per cent. TABLE 98. FACTORS BY WHICH MILLIGRAMS OF PRECIPITATE ARE MULTIPLIED IN ORDER TO GIVE RESULTS CALCULATED AS: /3-hydroxybutyric Acetone. Molecular acid. [ C c M o.c. 1() - Molecular Determination. Gins, per Mg. per Cms. per MR. per acctonc concentra- liter, 100 c.c., liter, 100 c.c., or /3-ao.id tion in urine. blood. urine. blood. per liter, blood or urine. plasma. Total acetone bodies' 0.0428 2.14 0.023S 1.19 4.11 . 000206 ( /3-hydroxybutyrie- . . < 0.0460 2.30 (2.44) . 0256 1.28 (1.37) 4.42 0.000221 0.000235 Acetone and diacetic acid . 0304 1.82 0.0203 1.015 3.50 0.000175 'The "total acetone body" factors are calculated on the assumption that the molecular ratio (acetone plus diacetic acid) : (/3-hydroxybutyric acid) is 1:2. Because the hydroxybutyric yields on oxidation only 0.79 molecule of acetone, the "total acetone body" factor is absolutely accurate only when the above ratio is 1:2. But with the range of mixtures encountered in acetonuria, when the ratio is usually between 1:2 and 1:3, with extreme Hurts of 1:1 and 1:4, the use of the above approximate factors for "total acetone bodies" will seldom involve a significant error. The actual errors in percentages of the amounts determined are as follows: ratio 1:1, error 3.7 per cent.; ratio 1:2, error zero; ratio 1:3, error 1.9 per cent.; ratio 1:4, error 3 per cent. 170 2 The factors in parenthesis are the usual factors X ,> and are for use in deter- mination of /3-hydroxybutyric acid in blood when the acetone is precipitated and the /3-acid determined in 160 c.c. of the filtrate. IDS I'h'IXK, BLOOD A\l) ItKM'IHATlOX IX DIABETES S. Nitrogen.- The determination of the nitrogen in tlie urine is valuable l)eeanse it furnishes an index to the quantity of protein which the patient is disintegrating. Incidentally, this is the* easiest way to determine the quantity of protein in the diet. Since nitrogen constitutes 10 per cent, of the protein molecule, we can multiply the quantity of nitrogen obtained in the urine by (>} to obtain the protein which it represents. \Ye shall not be far wrong if to this we add 1 or "2 grams of nitrogen (0 to 12 grams protein) to offset the nitrogen of the feces, and consider this total quantity as repre- senting the protein in the food. The determination of the nitrogen is also valuable because it is often useful to know the ammonia- nitrogen nitrogen ratio as well as the dextrose-nitrogen ratio. The Kjeldahl test for nitrogen is the one often employed, but the method used by Dr. Ragle in my laboratory based upon that of Folin requires less time and is reliable. Formerly large quantities of nitrogen were obtained in the urines of diabetic patients, but modern treatment with its restriction of protein makes these excessive quantities rare. Recently, however, I have met with three such examples: thus, Case Xo. Oo2 came to me upon a supposedly restricted diet, and the nitrogen in the urine amounted to 20.2") grams. The diminution of protein in the diet removed the sugar entirely from the urine, even though 12 grams of carbohydrate were added in the form of oatmeal. Case Xo. (')l(') was not sugar-fret 1 before her entrance to the hos- pital, but in the hospital easily became so. The urine upon March IS, 11) KS, amounted to 001f> c.o., and contained 4S."> grams sugar and 07.0 grams nitrogen. She entered the hospital May 22, 10K>, and the urine was free from sugar May 2">. Five months sub- sequently the urine was examined and the quantity of nitrogen in the twenty-four hours amounted to J'>0 grams. Case Xo. 1100 has already been discussed in connection with the volume of the urine ( see p. 1 70), and will be referred to more extensively on p. )>70. A daily analysis for nitrogen is expensive, but it is a simple matter to aliquot specimens of urine for a week and then obtain the average nitrogen excretion per day. Two or three admirable methods are available for the determina- tion of total nilrogen in urine provided a completely equipped laboratory is at hand. Folin's a ("'ration method' requires besides the usual apparatus an air current and a Duboscq colorimeter. Bene- dict has modified this method and distils the ammonia. For two years my former assistant, B. II. Ragle, now of the Massachusetts (Jeneral Hospital, has used a method which he devised, and it has been found accurate and rapid. The ordinary straight Jena glass 1 Folia and Farmer: Jour. liiol. Chciu., l'.H2, xi, p. -UK*. THE EXAMINATION OF THE URINE 199 tube 200 x 2 mm., is perfectly satisfactory for Folin's aeration method but unsatisfactory for distillation. Jn September, 1914, \ve had Eimer & Amend, of New York, blow us the simple tube of heavy Jena glass as shown in the diagrams. By a slight modifi- cation of the distillation frame one can use, instead of these special Jena glass tubes, Erlenmeyer flasks of 250 c.c. capacity made of non-sol glass, and these we now employ. Digestion and distillation are carried out as follows: 1 c.c. of the urine, diluted 1 to ">, is pipetted by means of an Ostwald pipette into one of the special tubes, according to Folin. Add 1 c.c. of FIG. n. Oxidation tube with Folio's fumo absorber. concentrated sulphuric acid, 1 gram of potassium sulphate, one drop of 10 per cent, copper sulphate, and a small pebble or glass bead. Digest over a micro-burner and allow to cool until the digestion mixture just becomes viscous, then add about 10 to 12 c.c. of water, a drop of alizarin indicator, and a pinch of talcum powder. Transfer to distilling frame and after connecting as in diagram, stopper tightly. A glass tube should perforate the one-hole stopper and reach almost to the bottom of the tube. (Figs. 5 and (>.) Start the micro-burner under the tube and then by means of a pipette, and by releasing the stop-cock at A, introduce )> or 4 c.c. of 40 per cent, sodium hydroxide. This will be sufficient to render 200 I'RINE, BLOOD AXD RESPIRATION IN DIABETES strongly alkaline, and the ammonia is now quickly forced over into ]() to 20 c.c. of - 7 N ',j hydrochloric acid. In six to eight minutes all the ammonia will have been driven over and the excess acid is titrated with 7 N ff sodium hydroxide. liable uses the -^ solutions and I c.c. of urine in his determinations in order to simplify the computation. One cubic centimeter of 7 N hydrochloric acid := 0.0002 gram nitrogen. If S c.c. of the acid are neutralized by the ammonia the calculation is (S x 0.0002 x ,"> x 1000 (c.c. in 24 amount) = 8 grams nitrogen. Fie. G. A micro-Kjcldahl movable distillation frame designed for small laboratories by J3. II. Ha^le. 0. Albumin. The test for albumin in the urine should be per- formed at frequent intervals during the care of diabetic patients. Too often the diagnosis of diabetes leads us to neglect the general treatment of the case. Case No. I]47, already cited (p. 170), is only too poignant an illustration of this tendency. 10. Casts. From the time of Kiilz the irritation of the kidneys in the first stages of diabetic coma has been observed. Over and over again I have seen typical "showers" of casts at the beginning of diabetic coma. They may occur at times when the albumin in the urine amounts to the slightest possible trace. Casts in the urine, even though occurring in showers, do not necessitate the development of fatal coma, for these were found in the urine of Case No. 705 on December 0-7, 1915. (See Table 99.) THE EXAMINATION OF THE URINE 201 TABLE 99. CASE No. 765. SEVEKE ACIDOSIS; "SHOWERS" OF CASTS WITHOUT COMA. Date. \"^' , Albumin. Diaeetii acid. Ammonia, gins. Sugar, urine, gms. Carbo- hydrate balance, gms. Soda admin- istered. Weight IbB. Alveo- lar air, C0 2 mm. HK. 1915. Dec. 0- 7 3.3 1 3.0% Shower* of yranular t asts. 7- S LSOO ' Slight + + 3.9 ; 25 -20 88 20 trace 8- 9 1200 Slight + + + 2.G 7 -7 o 88 26 j trace 9-10 1200 Very + + 2.4 2 ;^ 89 29 slight trace 10-1 1 1000 Slightest + + 2.9 G I 89 32 possible trace 11-12 1400 ; + + 14 16 88 33 12-13 900 Slightest + + 1 .5 9 -9 89 32 possible trace Death in coma, Octob cr 20, 1910. 2 11. Chlorides. The importance of a knowledge of the chlorides in the urine is due to the remarkable changes of weight of diabetic patients which are apparently related to the excretion and retention of sodium chloride. This subject is discussed under Influence of Sodium Chloride upon the Weight, page 120. Volhard's Quantitative Method. Solutions required: 1. Silver nitrate solution: 29.06 grams are dissolved in 1000 c.c. of distilled water. One cubic centimeter of this solution is equal to 0.01 gram of sodium chloride. 2. Ammonium sulphocyanate solution: Thirteen grams of the chemically pure salt are dissolved in 1000 c.c. of distilled water. This should, however, be accurately standardized by titration with the silver nitrate solution so that 1 c.c. is equal to 1 c.c. of the silver salt. 3. Indicator: A saturated solution of ammonio-ferric alum. 4. Concentrated nitric acid. 5. A 15 per cent, solution of potassium permanganate. Ten cubic centimeters of the urine to be tested are pipetted into a 100 c.c. volumetric flask. Add 20 c.c. of the standard silver nitrate solution, about 5 c.c. of nitric acid and 5 c.c. of the ammonio-ferric alum solution. If this mixture is dark add 4 or 5 drops of the potassium permanganate solution, when the mixture will become a pale yellow. 1 Ammonia in 14 (!). 2 For continuation of case see Table 145. 202 URINE, BLOOD AND RESPIRATION IN DIABETES Now add distilled water to the 100 e.c. mark and then shake and filter off the precipitated siK'er chloride. Fifty cubic centimeters of the filtrate, corresponding to 5 c.c. of the urine, are pipetted into a white porcelain dish and the excess of silver nitrate is determined by titration with the standard ammonium snlphocyanate solution. The end-point, the first constant reddish color, is quite definite. The number of cubic centimeters of ammonium sulphocyanate solution multiplied by 2 and subtracted from 20 will give the number of cubic centimeters of the silver salt precipitated by the chlorides in 10 c.c. of the urine. The method can be .shortened without detriment by using 10 c.c. instead of 20 c.c. of the standard silver nitrate solution (adding 10 c.c. more if an end-point is obtained with one drop of standard ammonium sulphocyanate). Dilution and filtration are then unnecessary. In the calculation subtract the number of c.c. of ammonium sulphocyanate used from 10, which will give the number of cubic centimeters of the silver salt precipitated by the chlorides in 10 c.c. of the urine. This number divided by 10 and multiplied by 0.01 and by the number of cubic centimeters of urii.e in the twenty-four-hour amount gives the amount of chlorides in grams. B. THE EXAMINATION OF THE BLOOD. The blood required for the various tests which follow is easily and painlessly obtained if a 22-gauge needle is employed. For con- venience, during the last few months, we have used platinum needles. These can readily be sterilized and sharpened. The total quantity of blood required seldom exceeds 20 c.e. and generally far less because it is rare that tests are necessary for acidosis. For the Benedict-Lewis sugar test it is best to have "> c.c. which can be transferred to a flask containing a few crystals of potassium oxalate; for the blood fat 15 c.c., for the Van Slyke (X)o test (> c.c., and for the "Wishart acetone test 4 c.c. For the prevention of coagulation, as a routine, at the Deaconess Hospital, we are accustomed to insert in the flask three drops of a saturated solution of sodium citrate instead of a few crystals of potassium oxalate, and to moisten the inside of the 20 c.c. syringe with a similar quantity of the citrate solution, rinsing out the syringe thoroughly therewith and expelling the same before the blood is drawn. However, it this method is adopted the blood should be shaken vigorously for three minutes (by the watch) after it has hem transferred to the flask. This would be objectionable in the case of blood drawn for the Van Slyke ('()-.> test, and in that case a few crystals of potassium oxalate should be employed. EXAMIXATION OF THE BLOOD 203 1. Blood Sugar. (a) Lewis-Benedict Method. 1 At the present time this method appears to be the one in most general use and the method which gives the most reliable results. The color obtained by heating a dextrose solution with picric acid and sodium car- bonate is employed as the basis of the method proposed by Lewis and Benedict for the determination of blood sugar. The blood protein is removed by precipitation with picric acid. The method of blood-sugar determination is as follows: Two c.c. of blood, and more if other blood tests are desired, are collected in a small flask containing one or two crystals of potassium oxalate. For the test 2 c.c. of blood should be transferred at once into a 25 c.c. volumetric flask containing 5 c.c. of water. The contents of the flask are shaken to insure thorough mixing and the consequent hemolysis of the blood. Then 15 c.c. of saturated aqueous solution of picric acid are added, as well as a drop or two of alcohol to dispel any foam, and the contents of the flask are made up to the mark with water and then shaken. After filtration 8 c.c. aliquots are measured out into large Jena test-tubes for duplicate determinations. Two c.c. of saturated picric acid solution and exactly 1 c.c. of 10 per cent, sodium carbonate are added (as well as two glass beads and two or three, or even five to ten drops of mineral oil) and the contents of the tubes are evaporated rapidly over a direct flame until precipitation occurs. About 3 c.c. of water are added, the tube is again heated to boiling to dissolve the precipitate, the contents of the tube are transferred quantitatively to a 10 c.c. volumetric flask, cooled, made up to the mark, shaken, and then filtered through cotton into the colorimeter chamber. The color is compared at once with that obtained from 0.65 ing. of anhy- drous dextrose, 5 c.c. of saturated picric acid, and 1 c.c. of 10 per cent, sodium carbonate when evaporated to precipitation over a free flame with beads and oil and diluted to 10 c.c. The calculation of the sugar present in the unknown blood sample is very simple. The original 2 c.c. of blood were diluted to 25 c.c., and of this amount 8 c.c. were taken for a determination. In other words, the aliquot contained the equivalent of -^ x 2 c.c., or 0.04 c.c. of blood. A permanent standard can be made by dissolving picramic acid in an aqueous solution of sodium carbonate. For details, see original article. At the Deaconess Hospital the Lewis-Benedict tests are performed in a manner slightly different from the Lewis-Benedict method as described above, but in general according to the modification of Myers and Bailey. - Place 0.5 to 1 gram picric acid in a 25 c.c. volumetric flask, add 1 Lewis-Benedict : Jour. Hiol. Choni., 1915, xx. p. 01. 2 Myers and Bailey: Jour. Biol. Choin., 1910, xxiv, p. 147. 204 URINE, BLOOD AND RESPIRATION IN DIABETES 5 c.c. blood, dilute to 25 c.c. with distilled water and shake until the complete disappearance of the red color. Centrifugalize, then filter the supernatant liquid, take -'> c.c. of the filtrate, measured accurately with pipette, put in 10 c.c. volumetric flask and to this add 1 c.c. of a 20 per cent, solution sodium carbonate (2!>.4 per cent, solution of monohydrated sodium carbonate). Heat on the water-bath for fifteen minutes. Allow to cool, make up to 10 c.c. with distilled water and read against the standard picramic acid in the colorimeter. (b) Bang's Method. 1 The test depends on the assumption that copper is reduced by sugar without regard for the amounts present. In other words, a sugar solution from 100 mg. of blood will reduce as much copper proportionately as a solution from 100 c.c. of blood. If the microscopic amount of copper involved in the reaction can be determined the results will be as accurate as if larger amounts of the metal are used. Accordingly an albumin- free sugar solution is made from a small amount of blood and is boiled with copper solution. The cuprous oxide which is formed is held in solution by the method which Bang has described previously and is reoxidized by iodine in the presence of potassium carbonate. Cupi-ic carbonate and potassium bicarbonate are formed while potassium iodide is set free and can be tested for with starch as an indicator. The reaction is quantitative and extremely delicate. The technic for the test is given in detail by Fitz, 2 from whom I quote freely. The copper solution is made by dissolving 100 grams of potassium bicarbonate in from 600 to 700 c.c. of distilled water. 100 grams of potassium carbonate are added, (>(> grams of potassium chloride and 100 c.c. of a 4.4 per cent, copper sulphate solution (C\iS() 4 + 51 1 2 O). When everything is dissolved the mixture is diluted to 1 liter and is ready for use after it is allowed to stand for twenty-four hours. Titration is made with two-hundredth-normal iodine solution prepared from 1 c.c. of 2 per cent, potassium iodate solution, 2 grams of potassium iodide and 5 c.c. of tenth-normal hydro- chloric acid, the mixture being diluted to 100 c.c. This solution is unstable and must be renewed every three or four days. Two or three drops of blood are placed on a small piece of thick blotting paper, 3 which previously has been boiled in dilute acetic acid and water to remove traces of albumin and sugar. The blot- ting paper should be cut in oblongs 10 to 20 mm. and weighed before and after the blood is added. In this way the weight of blood to within 1 mg. is found. In these experiments small weigh- 1 HaiiK: Dcr Blut/ucker, Wiesbaden, I'U.'i. 2 Fitz: Arch. Int. Mod., i!)M, xiv, p. IM. 5 \Yo have used filter paper (No. /is'.), 7 cm. dia.), purchased from Kimer & Amend, New York. EXAMINATION OF THE BLOOD 205 ing flasks were used to hold the blood and paper. After the blood is weighed the protein is precipitated by hydrochloric acid in potas- sium chloride. For this purpose 1360 c.c. of saturated potassium chloride are diluted with 640 c.c. of distilled water and 1.5 c.c. of 25 per cent, hydrochloric acid is added. Of this mixture 6.5 c.c. are used for the precipitation and added boiling hot; 6.5 c.c. are also added to wash the solution into the boiling flask. The albumin in the blood precipitates on the blotter, while the sugar diffuses through the solution. After half an hour's cooling the sugar solution is transferred to a 50 c.c. Jena flask with a straight neck and is washed in by 6.5 c.c. potassium chloride solu- tion. In this way all the sugar in the blood is removed from the weighing flask and is ready for titration. One cubic centimeter of the copper solution is now added and the mixture is boiled over a micro-burner. The heating is an important factor in the test. The solution must be warmed so that it begins to boil in between one minute and fifteen seconds and one minute and a half, and must be boiled for exactly two minutes, since the copper solution can be reduced by longer or more vigorous boiling and will give improper readings. At the end of this time the flask is immersed in cold water and titrated. A source of error in the determination is air oxidation of the reduced copper. During boiling this is discounted. But after cooling, air must be kept out of the solution by manipulating it under carbon dioxide gas. For this purpose Bang introduced into the neck of the flask a hooked glass tube which was connected to a carbon dioxide tank. The gas was turned on as soon as the flask was put into cold water and in this way air oxidation was prevented. Thannhauser and Pfitzer 1 used small flasks with side- arms which were easier to handle. Similar flasks were used in our experiments. After the two minutes' boiling the gas is introduced and the flask is immediately cooled. This can be done in less than three seconds. A few drops of 1 or 2 per cent, soluble starch solution are now added as an indicator. The titration is made from a 1 or 2 c.c. differential pipette reading in hundredths of a cubic centimeter, and the iodine solution is added until the light blue color of the copper solution takes on the deep blue color of the starch and iodide. Bang has tested the method carefully with known sugar solutions. The amount of iodine used by increasing amounts of sugar is so nearly constant that he has been able to make a formula for deter- mining the amount of sugar present. The amount of iodine used in titration is divided by 2.2. This result minus 0.01 will give the 1 Thannhauser and Pfitzer: Munch, rned. "Wchnschr., 1913, Ix, p. 2155. 200 ritl.\K, BLOOD AND RESPIItATIOX IX DIABETES milligrams of sugar present in the solution. In these experiments a special table was constructed, however, which showed the iodine readings for varying amounts of sugar from which the blood-sugar readings per 100 c.c. of blood were calculated. The entire test takes less than three-quarters of an hour, requires so little blood that it can be repeated on the same individual at will, and is fairly easy to perform after short practice. Comparative experiments show that the results obtained by the micro method were not absolutely accurate if the results obtained by the Bertram! method are considered correct. In a series of forty-two determinations by both methods, however, Fit/ found in only '2 cases a difference of more' than 0.0.' 5 gram per 100 c.c. of blood. This means that while the absolute error was large the relative error was slight. The curve of blood-sugar determinations made from all the examinations was closely parallel. Hence, for comparative and repeated studies in blood-sugar estimations the micro method is excellent. Furthermore, since the amount of blood required to make the test is so small and the technic so simple the method affords an excellent clinical means for the estimation of blood sugar. 1 (<) Other Methods. Epstein 2 has devised a .simple procedure for the estimation of blood sugar requiring but a few drops of blood, and of this 1 hear favorable reports. Strouse :! has called attention to the advantages of the Kowarsky method, which also requires but little blood and demands but little time. 2. Lipoids. The frontispiece is sufficient evidence of the desir- ability of examining the blood of diabetic patients for its content of lipoids. It is true, one seldom sees cases with so high (0.4) a per cent, of blood fat as Case Xo. 7S(>, and in fact, I have had no other among l!S cases for whom analyses have been made in Prof. Bloor's laboratory. 1 One reason for this is that less fat is given now than formerly and another that the blood is examined for the lipoids before instead of after a meal. Notwithstanding these changed conditions, however, the lipoids have been almost invari- ably increased in diabetic patients either as a whole or in part (as fatty acids, cholesterol, lecithin). Of the cases examined, an in- 1 Whereas this method is adinir:il>li< in the hands of skilled laboratory workers, il is not practicable for others. It, has been used for a ye:ir and a half with success by I'.. II. liable in my laboratory and by Miss Habcock at, 1 lie Nutrition Laboratory, by whom all blood-sunar analyses report eel in the first edition of this book were made s:ive seven by II. Koefod, who is also -killed in the technic. Since the early months of I'.IUi the Meyers and liailey modification of the Lewis-Benedict method has been exclusively employed 'and with far greater satisfaction. - Kp.-tein: Jour. Am. Mod. Assn., I'.UI. Ixiii, p. Hid?. 3 St rouse -.Johns Hopkins IIosp. Hull., P.M."), xxvi, p. i'1-l. 1 In :i subsequent .-'erics of KU specimen-: of blood, representing X7 of my cases, examined by Horace Gray, no value above this was obtained, save in Case Xo. "Mi, who on one occasion showed !<>.:} per cent .and on another 11.1 per cent, total lipoids. EXAMIXATIOX- OF THE BLOOD 207 crease was found in 20, a normal state in 2, and a decrease in none. The blood fat should be normal in all cases of diabetes and no case of diabetes should be considered cured otherwise. When carbohy- drate and protein are unassimilated, they appear in the urine, but when fat fails of assimilation, it collects in the blood. The normal variations of the lipoids in the blood and the changes which they undergo have been described on page 102. The method given below for the estimation of the blood lipoids is that now employed by Professor Bloor, and kindly given me by him. It is a modification of his method published in 1914, ! and of that in the first edition of this book. "The method to be described depends on a new principle the determination of the fat by precipitation in a water solution and comparison of the cloudy suspension so obtained with that of a similarly prepared standard fat solution by the use of the nephel- ometer. The determination may be completed in about three- quarters of an hour and may be carried out with from O.o c.c. to 5 c.c. of blood. Ordinarily about 2 c.c. are used. It has been found to be accurate to within 5 per cent, of the total fat. The procedure is as follows : "Extraction. Three c.c. of freshly drawn and well-mixed blood are run in a slow stream of drops into a graduated flask containing about 80 c.c. of a mixture of 3 parts alcohol and 1 part ether (both redistilled), which is kept in constant motion by rotating the flask. The solution is raised to boiling by immersion in a water-bath (with frequent shaking to prevent superheating), cooled to room tempera- ture, made up to volume with alcohol-ether, mixed and filtered. The extract if placed in tightly stoppered bottles in the dark will keep several months unchanged. ''Determination. From 5 to 20 c.c. (ordinarily 10 c.c.) of the extract, containing about 2 nig. of 'fat,' are measured with a pipette into a small beaker and saponified by evaporating just to dryness with 2 c.c. of y sodium ethylate (made by dissolving cleaned metallic sodium in absolute alcohol). After evaporation is complete 5 c.c. of alcohol-ether are added and the mixture heated slowly to boiling. A similar solution of the standard is prepared by measuring 5 c.c. of the standard fat solution (see below) into a beaker and heating to boiling as above. Fifty c.c. of distilled water are now added to each beaker and the solutions mixed by stirring, taking care that all the material in the saponifieation beaker is dissolved. To standard and test solutions are added, as nearly simultaneously as possible, 10 c.c. portions of dilute (1 to 4) hydro- chloric acid and the solutions allowed to stand five minutes, after which they are transferred to the comparison tubes of the nephel- ometer. 1 Bloor: Jour. Biol. Chem., 1914, xvii, p. 37b. 208 I'RINE, BLOOD AND RESPIRATION IN DIABETES "If bubbles appear on the walls of the tubes they should be removed by inverting the tubes two or three times. The movable jacket ou the standard tube is set at a convenient point, generally 50 nun. (Richard's nephelometer) and comparisons made by adjusting the jacket on the test solution until the images of the two solutions show equal illumination. Not less than five readings are taken, alternately from above and below, and the average taken as the correct reading. "The standard solution used is an alcohol-ether solution of pure triolein of which 5 c.c. contain about 2 ing. of fat. The alcohol and ether used for the .standard are freshly redistilled absolute alcohol and pure dry ether." '.}. Non-protein Nitrogen. During the last year the determina- tion of the non-protein nitrogen in the blood of many of my diabetic patients had been made by my assistant, Dr. W. Richard Ohler. The results of these and other tests of renal function will eventually be published elsewhere, but the facts here presented will give an indication of what has been found to date. The remarkable effi- ciency of the kidneys of patients with diabetes has always impressed me, and the onset of renal disease in cases of diabetes of long dura- tion has been far less frequent than most of us would anticipate. Proof of this is shown by the moderate changes of blood-pressure in cases of diabetes over long periods of time. (See p. 414.) On the other hand, in beginning coma renal involvement has always seemed to me to be marked. Today the question arises as to whether this may not have been in part due to the alkaline treatment at such times, which forced an excessive quantity of acid through the kid- neys. The evidence collected by Dr. Ohler will help in the study of renal involvement in diabetes, but as yet the data are not sufficient to allow decisive statements to be made. Particularly do I wish to be conservative because Dr. Reginald Fit/, 1 understand, has worked assiduously upon this problem during the last year at the Rockefeller Institute, and with his own experience with cases of renal disease in Christian's clinic at the Peter Bent Brigham Hos- pital, his conclusions must have great weight. The non-protein nitrogen of the blood varies in normal individuals between 20 and I50 mg. per 100 c.c. blood. In this series of :>0 cases and :'>(> determinations, Dr. Ohler found the non-protein nitrogen to be normal in M4 per cent., increased in (>0 per cent, and decreased in (> per cent. Should 20 to -10 mg. per 100 c.c. blood be taken as normal for the non-protein nitrogen, (IS per cent, of the observations would come within this group. Two casc-s showed a non-protein nitrogen of 1(1 and 20 ing. From Table 100 it will be seen that these are young individuals with normal blood-pressure and no evidence of disturbance of renal EXAMINATION OF THE BLOOD 209 II 14 210 1'lfIXK, BLOOD AM) RESPIRATION IX DIAMETER function, for albumin and casts arc absent from the urine and the excretion of plieuolphtlialein in one case was 5S per cent, in t\vo hours. It is noticeable that these low values were in association with the normal blood-pressure in one instance and a blood sugar of ().f)7 percent, in another, plainly showing the independence of the non-protein nitrogen in the blood and the blood sugar. The low values might well be accounted for by the fact that one of the patients had been without protein in the diet for four days and the other for one day. In the second group of cases, with the non-protein nitrogen vary- ing between 121 and )>() ing., all but two Cases Xos. 107(> and 1104 stand in agreement with the normal phenolphthalein excre- tion. Both of these cases were on a comparatively low protein diet, and this may account for the fact that the non-protein nitrogen is within normal linits while the phenolphthalein was respectively 28 and ',}() per cent. In each case the blood-pressure was increased. It is quite likely that another case Xo. 1197 might have shown a low phenolphthalein output, and it is unfortunate that such a test had not been made, for albumin, casts and a high blood-pressure we're noted in this patient. Xo particular relation is evident between the non-protein nitrogen and the evidences of acidosis, and absolutely no relation between the non-protein nitrogen and the sugar in the blood. In the group is included Case Xo. lOoo, with a non- protein nitrogen of 27.") nig., marked acidosis and evidences of renal involvement but he received only 13 grams of protein, in the preceding six days. Unfortunately no phthalein test was done. Had the protein been greater in the diet, it is quite probable that the non-protein nitrogen would have been increased in the blood. The influence of fasting on non-protein nitrogen, therefore, is by no means to be disregarded as a therapeutic means of treatment. The independence of the blood sugar of the non-protein nitrogen is again apparent. (Iroup III represents those patients whose non-protein nitrogen was on the border-line of normal namely, between o2 and K) mg. The average age of the patients in this group is but two years more than in (Jroup I, so that, age cannot be influential in causing the increase. 'That the renal involvement was con- siderable!, the presence of albumin in 10 out. of !> cases and of casts in 7 cases bear \vitness. The phenolphthalein excretion was In'low normal in I of the <> cases tested, and in 7 of the 11 cases the blood-pressure was increased. The volume of the urine was increased in ."> of the !> cases in which the twenty-four hour quantity was known for the day of the test. The blood-sugar was distinctly elevated. ( lear evidence of renal involvement is therefore manifest in these patients, and the non-protein nitrogen EXAMINATION OF THE BLOOD 211 is an expression of it. This cannot be ascribed to the diet, for that was low in protein in practically all the cases. Is the renal involve- ment primary or secondary'/ In answer to this question the dura- tion of the disease is certainly important. Whereas this was 1.5 years in Group I, it was 6.7 years in Group II and 3.4 years in Group III. The intensity of the dialxrtes was greater, as shown by the greater frequency of acidosis and the higher percentage of sugar in the blood. Therefore, so far as this group is concerned, the intensity of the diabetes appears to have affected the kidneys unfavorably. The values for ( 'ase Xo. 10SO, following an operation for removal of the prostate, are interesting in comparison with the values for the same patient shown in Group IV, which were nearly twice as high, having been obtained Ixrfore the operation. Group IV comprises cases of diabetes who showed a non-protein nitrogen of 40 nig. or over. All of these patients exhibited albumin- uria, and the phenolphthalein excretion was decrease,! in the 4 cases in which it was tried. The protein in the diet was generally low. Acidosis was present in most of the cases. Two of the higher values were for the prostatic case, Xo. 1086, after his bladder had been drained, but before the prostate was removed. The lower value, 58 mg., was obtained eleven days after the drainage of the bladder by Dr. Chute; the higher value, 70 mg., four days after the drainage. Case Xb. 1 190 was a very severe diabetic with marked inanition ; it is notable that a lower value (see Group III) was obtained soon after entrance, when he showed acidosis. For a presentation of all the data relating to this case see page 379. Case Xo. 1143, with the highest value of all, died in coma. The blood was taken immediately thereafter. Whether the latter circumstance had influenced the proportion of non-protein nitrogen in the blood I do not know. The test must be performed in other cases of coma to settle the influence of this factor. (a) Determination of Non-protein Nitrogen in Blood by Titration. The method employed is in general the same as the method used by Folin and Denis. 1 Five cubic centimeters of well-mixed blood are run into a 50 c.c. measuring flask half-filled with acetone-free methyl alcohol. The flask is now filled up to the mark with methyl alcohol and vigorously shaken. After two or more hours the contents of the flasks are filtered through hard, dry filters. To the slightly colored filtrate are then added two or three drops of a saturated alcoholic solution of zinc chloride and after standing but a few minutes the mixture is filtered 1 Folin and Denis: Jour. Biol. Chcm., 1912, xi, p. 527. 212 I'KIXE, BLOOD AND KEXP1RA TI().\ IX DIABETES through a dry filter paper. The filtrate should now be perfectly colorless. Ten cubic centimeters of this filtrate are now pipetted into a tube of the same kind as is used in urine analysis. One drop of sulphuric acid and a glass bead or pebble is added and the methyl alcohol driven off by immersing the tube in a beaker of boiling water or by gently heating over a free flame. After the alcohol is removed digestion is carried out just as in the case of urine. 1 One cubic centimeter of sulphuric add, 1 gram of potassium sulphate, and a drop or two of 10 per cent, copper sulphate are added. When digestion is complete, and the digestion mixture has cooled until viscous, 8 to lOc.c. of distilled water, a drop of ali/arin indi- cator, and a pinch of talcum are added, and the tube transferred to the distillation rack where, after neutralizing with sodium hydroxide, as in the case of urine, the ammonia is driven over into 10 c.c. of ~ 7 N - hydrochloric acid. The 10 c.c. of filtrate used arc equal to '2 c.c. of the blood. The calculation after titrating the excess acid with 7 X - sodium hydroxide is as follows: X X 0.0002 >; 100 = grams of non-protein nitrogen in 100 grams of blood, X being the number of cubic centimeters of acid neutralized by the ammonia by distillation. One cubic centimeter of the 1 - 7 X () - acid is equal to 0.0002 gram nitrogen. The calculation is made in terms of 100 c.c. of blood for the sake of convenience. 4. Chlorides. The determination of the chlorides in the blood gains importance from the fact that the water content of the body in diabetes varies so rapidly and to so great an extent. As yet few data have been published upon this point, but I insert the test recently described by McLean and Van Slyke- because of the increasing importance of the subject. The determination requires two steps: (1) Removal of proteins; ("2) titration of chlorides. Two cubic centimeters of oxalated plasma, with the aid of an accu- rately calibrated or Ostwald pipette, are run into a 20 c.c. stoppered volumetric flask which contains 10 c.c. of a 10 per cent, magnesium sulphate solution. The pipette is rinsed twice by drawing up into it the solution from the flask. Two drops of ~>0-per cent, acetic acid are added. The flask is filled to the mark with water, the contents 1 In private pract ice or in hospital wards where ii would be bothersome to carry flasks, pipettes and methyl aleohn], 10 or more c.c. of blood may be drawn from the vein into an open-mouth bottle containing about 0.1 Kram of powdered potassium oxalate to prevent, clot tins:. 2 McLean and Van Slyke: Jour. Biol. Thorn. . 101.-), xxi, p. 'Ml. EXAMIXATION OF THE BLOOD 213 mixed by inverting the flask, and then heated in a bath to 100 C. for ten minutes. By keeping the stopper loosely in place evaporation is prevented, and when cool the contents return to their original volume. A shorter period of heating than ten minutes may give lo\v values in the filtrate. The flask is then allowed to cool and the contents are poured into about 0.3 gram of Merck's " blood-charcoal reagent." After a few minutes the liquid is filtered through a dry filter and a water-clear filtrate obtained. Occasionally a small amount of charcoal passes through with the first few drops of filtrate, but this can be poured back through the same filter, and the filtrate then obtained is absolutely clear and colorless. Ten cubic centimeters of this filtrate are transferred, by means of an Ostwald pipette, to a 25 c.c. volumetric, flask, 5 c.c. of solution I 1 are added, and the whole is made up to the 25 c.c. mark with distilled water. Five cubic centimeters of solution 1 will precipitate up to 10 mg. of XaCl. In samples with high percentage of chloride only enough filtrate is taken to keep within this limit of 10 mg. Two drops of cotyl (caprylic) alcohol are added, and the vessel is stoppered and shaken gently by inverting it several times. Immediate coagulation of the protein occurs. After allowing five minutes for complete precipitation to take place the solution is filtered through a dry filter and a perfectly clear and colorless filtrate is again obtained. An aliquot part of the filtrate, preferably 20 c.c., is used for titra- tion. Thus one has in this 20 c.c. of filtrate the equivalent of 4 c.c. of solution 1 , and accordingly 4 c.c. of solution 3- must be added. The starch is first dissolved by the aid of heat in 500 c.c. of water, the citrate and nitrite then added, and the whole filtered through cotton; the filter washed with hot water, the filtrate allowed to cool and made up to 1000 c.c. The solution becomes a little cloudy upon standing, but its efficacy is unimpaired, and this slight turbidity in no way interferes with the end-point. Solution 2 (KI, 3 grams; water, 1000 c.c.) contains M/5S.5 potassium iodide, 1 c.c. of which is equivalent to 1 mg. of XaCl. The solution is standardized against the silver solution by adding 5 c.c. of the latter to 5 c.c. of solution 3 and titrating with the 1 Solution 1 is an acid M 29.2.3 solution of silver nitrate, 1 c.c. of which is equiva- lent to 2 mg. of XaCl. AgXOs 5.812 grains HXOs 2.-)0.0 c.c. Water to 1000.0 c.c. " Solution :i contains substances which respectively regulate the acidity, provide an oxidizing agent for the iodide, and serve as indicator: Sodium citrate (NasCeNjOr + 5J N 2 O) . . . 440.0 grams Sodium nitrite 20.0 grams Soluble starch 2.5 grams Water to 1000.0 c.c. 214 URINE, BLOOD AXD RESPIRATION IN DIABETES iodide solution to the blue end-point. The iodide solution is then diluted to such a decree that 10 c.c. are exactly equivalent to 5 c.c. of the silver solution. Potassium iodide is run in from a burette until the blue end-point appears. The first definite blue color is taken as the end-point, and with slight practice is unmistakable. The result may be calculated from the following formula, which applies when 20 c.c. of the filtrate from silver chloride are titrated: 12. 5 (8 c.c. Kl solution used) c.c. blood filtrate taken 5. Acetone in the Blood. It sometimes occurs that acidosis is present yet is not disclosed by the simple ferric chloride test for diacetic acid in the urine. It has long been realized that the analy- sis of the blood would give a more reliable index of the condition of the patient, but the methods for examination have been so com- plicated as to preclude their general adoption. Recently Miss Mary 15. YVishart, of the Rockefeller Institute for Medical Research, devised the following method, and I am indebted to Miss Wishart for the privilege of inserting it here. It can readily be applied by any physician. (a) Wishart Method for Detection of Acetone in the Blood. The blood is drawn into a syringe or tube containing a few crystals of potassium oxalatc, then ceutrifuged for five minutes at medium speed. The test is made on the plasma with as little delay as possible, as there is liable to be some loss of acetone on standing. For a small quantity of plasma. (0.5 c.c. or more) add solid ammonium sulphate until plasma is thoroughly saturated and pro- tein precipitated; then add two or three drops of a freshly made f> per cent, solution of sodium nitroprusside and a few drops of concentrated ammonium hydrate. If the test is positive, in from one to ten minutes a color develops which runs all the way from a pale lavendar to that of a deep permanganate hue, in this way indicating whether much or little acetone is present. This is an adaptation to the plasma, of the Rottura nitroprusside reaction as ordinarily used for urine. It is said to be sensitive to 1 part in 20,000. (h ) Van Slyke-Fitz Methods for Determination of /3-hydroxybutric Acid, Diacetic Acid, and Acetone in Urine and Blood. (See page 194.) (I. Carbon Dioxide in Blood Plasma. Whether in health or in disease, the reactions of the body remain nearly constant and the blood not only conforms to tins general law but helps to enforce it. This end is accomplished in the presence of acidosis (a condition in which there is an increase in the quantity of /3-oxybutyric acid in the body) by removing the ( '()> from the blood to make way for the stronger acid. Therefore a determination of the quantity of ('();> in the blood is an index of the degree of acidosis. Recently Van Slyke has devised a comparatively simple method by which the EXAMINATION OF THE BLOOD 215 carbon dioxide capacity of the blood plasma can be determined. I am very fortunate in being able to include a description of the method furnished me by Dr. Van Slyke and in the method which follows he has made various modifications to prevent changes in the content of C() 2 in the plasma during the performance of the test. In general, the per cent, of CO 2 in the blood plasma stands in close relation to the per cent, of CO* in the alveolar air. This was originally shown by Van Slyke and has been confirmed by Walker and Frothingham, 1 who, however, noted that the correspondence in diabetes was not as close as in other diseases, and Van Slyke' 2 stated that: "In one type of diabetes, however, especially when on a low calorie diet, a low alveolar CO 2 content may be found while the blood bicarbonate is still normal. This emphasizes the fact that the carbon dioxide content of the expired air is only an indirect measure of the blood bicarbonate and is dependent for its accuracy upon a normal respiratory apparatus and a normal respira- tory center. The close correspondence in such a variety of cases between plasma bicarbonate and alveolar CO 2 shows the tenacity with which normal respiratory regulation is usually maintained." In Table 101 this same variation is to be seen in Cases Nos. 7 (SO, 1051, 10S5 and 1103, who were patients living on an extremely low diet. Austin and Jonas 3 have also used the method extensively. In the revised description of the method given below their observa- tions about changes in the content of C() 2 in the plasma, during the test are met. The combining power of the blood plasma for CO 2 has been determined upon 23 of my patients a much smaller proportion than would have been necessary had not acidosis ceased to be a frequent factor in the cases at the hospital. The results are tabu- lated to show the outcome of the cases under the varying degrees of acidosis thus found, the comparison between the blood and alveolar CO 2 and the relation of both to the ferric chloride test for diacetic acid. Three fatal cases occurred in the group. Case Xo. 1 143 entered the hospital in diabetic coma and died the same day. The C() 2 was 10 in terms of mm. Ilg., and this is the lowest quantity we have observed. Case No. 1053 with gangrene showed a plasma content of COo of only 21 in terms of mm. Ilg. tension five days before death and this dropped to 12 mm. Ilg. tension two days before the termination of coma. For a fuller description of this case, see p. 2S5. Case No. 10S5 died of inanition after leaving the hospital, sixteen days after the test (see p. 340). 1 Walker and Frothingham: Arch. Int. Mod., 1910, xviii, p. ;$()4. 2 Van Slyke: Eighth Annual Meeting of the American Society for Clinical Inves- tigation, 1910, p. 25. 3 Austin and Jonas: Am. Jour. Med. Sc., 1917, cliii, p. 81. 2Ki rmxK, Biooj) AND RESPIRATION IN DIABETES TAHLK 101. THK COMBINING POWER OF THK BLOOD PLASMA IN DIAHETES FOIl CO-, Co.MI'ARKI) WITH THK CO- IN THE A I.VKOLAK Alii AND THK Dl.U'KTIC Adi) IN THK UltlNK. I'lasnia CO, Alveolar air Case No. Date. with ('(), subsequent corresponding. observations. Mm. [Intension Diaec-tic ^" " f mil. 'lays to coma. Mm. 11^. tension 19 10 1143 Oct. S 10 14 + + 1 1053 May 15 21 21 + + + 5 19 12 11 + + + 1 Fatal cases . 10s5 Julv 11 32 12(1 Aug. 2 11 2 in the venous blood, it is advisable when ligature is used to release it as soon as the vein is entered, and to allow a short time for the stagnant blood to be washed out of the vein before the main sample is drawn. A second and lesser error, which appears to be significant only when the whole blood stands over an hour before centrifuging is due to apparent formation of organic acids by the corpuscles, and results in lowering of the bicarbonate. The plasma once separated, however, may be kept for analysis at leisure, any loss of (X) 2 , by such reactions as 2NaIirO. i ==' Xa,(X), + H,CO, being made good before the analysis by resaturating with ('() 2 at alveolar tension as described below. The clear plasma, being pipetted oil', should, in case it is not convenient to determine its carbon dioxide capacity at once, be transferred to a paraffin-lined tube, where it will keep unchanged for a week if placed on ice. In order to determine its alkaline reserve, the plasma is satu- rated with carbon dioxide at alveolar tension. For this purpose the plasma (about o c.c. if there is plenty of material), which should be at room temperature, is placed in a separatory funnel of about )!()() c.c. capacity, and the funnel is filled with alveolar air from the lungs of the operator. The air is passed through a bottle full of glas* beads before it enters the funnel, in order to bring the moisture content down to saturation at room temperature. If one blows directly into the separatory funnel, enough moisture condenses on the wall to appreciably dilute the plasma. The funnel is closed just before the stream of breath stops, and is shaken for one minute in such a manner that the plasma is distributed as completely as possible about the walls, forming a thin layer which quickly approaches equilibrium with the carbon dioxide in the air. After the shaking has lasted a minute, a fresh portion of the alveolar air or of 5.5 per cent, carbon dioxide frm a gas tank, is run into the funnel, and the shaking completed. The determination of the carbon dioxide 1 content of the saturated plasma is performed as follows : The carbon dioxide apparatus (Fig. 7) held in a strong clamp on a ringstand, is completely filled with mercury which should fill both capillaries above the upper stop- cock. The mercury leveling bulb is placed about on a level with the lower cock. The cup at the top of the apparatus is washed out thoroughly with dilute ammonia followed by water, medicine droppers being convenient for this purpose. One cubic centimeter of the saturated plasma is now introduced into the cup, and allowed to flow down into the upper stem of the apparatus. The cup is now washed with two portions of about 0.5 c.c. each of water, care being EXAMINATION OF THE BLOOD taken that no air enters the apparatus with the liquid. One small drop of redistilled caprylic alcohol, to prevent foaming, is now admitted into the capillary connecting the cup with the upper end of the apparatus, and about 1 c.c. of 5 per cent, sulphuric acid is poured into the cup. Enough of the acid is now ^ carrying the ^ so that the apparatus is mercurv is now admitted into the apparatus, caprylic alcohol along with it, total volume of water in the exactl 2.5 c.c. A drop of bore 3 mm. bore placed in the cup and allowed to flow down to the upper stop-cock in order to seal same, and make it capable of holding an absolute vacuum. The leveling bulb (the lower cock having remained open from the beginning of operations) is lowered to a point such that the surface of the mercury in it is about 800 mm. beloAV the lower stop-cock, and the mercury in the apparatus is allowed to fall until the meniscus of the mercury has dropped to the 50 c.c. mark on the apparatus. As the latter is evacuated, bubbles of carbon dioxide are seen escaping from the water mixture in the vacuum. In order to completely extract the carbon dioxide, the apparatus is removed from the clamp and shaken by turning it upside down about a dozen times. It is then replaced, the mercury leveling bulb still being at the low level, and the water solution is allowed to flow completely into the small bulb below the lower stop-cock. The water solution is drained out of the portion of the apparatus above the stop-cock as completely as possible, but without removing any of the gas. The mercury bulb is now raised in the left hand, and the lower stop-cock is turned with the right hand so that mercury is admitted to the apparatus through the left-hand entrance of the 3-way cock without readmitting the water solution. The leveling bulb is held beside the apparatus so that the mercury level in it is even with that in the apparatus, and the gas in the latter is under atmospheric pressure. A few hundredths of a cubic centi- meter of water will float on the mercury in the apparatus, but this FIG. 7. Van Slyke's pipette for determining the combining power of the blood for carbon di- oxide. 220 URINE, BLOOD AXD HEM'IltATION IN DIABETES may be disregarded in leveling. The volume of gas above the short column of water referred to is at once read oil'. The calculation of the result into terms of volume percentage of carbon dioxide bound as carbonate by the plasma is quite compli- cated, and Van Slyke consequently computed a table which obviates the necessity of calculation. (Table 102.) TABU: 102. VAX SI.VKE'S TAKLK FOR CALCULATION ( Ibservcd vol. :is < 7t,0 20 1 3 4 5 7 s 9 0.30 1 2 3 4 5 (i 7 8 9 0.40 1 3 4 5 t; 7 S 9 0.50 1 > 3 4 5 (i 7 S 9 O.C.O H.-tr. Plasmas of normal adults yield 0. ('."> to 0.00 c.c. of gas, indicating ").'! to 77 volume per cent, of carbon dioxide 1 chemically bound by the plasma. Figures lower than ,">() per cent, in adults indicate acidosis. The normal figures for infants appear to be K) to .">,"> per cent., much lower than for adults. EXAMINATION OF THE BLOOD 221 Caution in setting up apparatus: The jaws of the clamp in which the apparatus is held should be lined with thick, soft rubber. The apparatus should be clamped very tightly because of the weight of the mercury. In order to prevent the apparatus from slipping out of the clamp, an iron rod should be so arranged as to project under the lower stop-cock, so that it will support the apparatus from this point, in case it should any time slip down from the clamp. The temperature figures at the heads of columns represent the room temperatures at which the samples of plasma are saturated with carbon dioxide and analyzed. It is assumed that both opera- tions are performed at the same temperature, which is also that of the solutions analyzed. The figures have been so calculated that, regardless of the tem- perature at which the plasma is saturated, the table gives the volume (reduced to 0, TOO mm.) of carbon dioxide that 100 c.c. of plasma are capable of binding at 20. If the figures in the table are multiplied by 0.94 they give, approximately (within 1 or 2 per cent.), the volume per cent, of car- bon dioxide bound by plasma at 37. If the figures in the table are multiplied by 0.09 they give, usually within less than 5 mm., the alveolar carbon dioxide tension of the plasma donor as determined by the Fridericia apparatus. If the figures in the table are multiplied by 0.01904, i. e., Wt. of 1 c.c. carbon dioxide at 700 mm., "100 they give the milligrams of carbon dioxide bound by 1 c.c. of plasma. The apparatus can be obtained from E. Greiner, 55 Fulton Street, New York. 7. The Hydrogen-ion Concentration of the Blood. A simple method for determining variations in the hydrogen-ion concen- tration of the blood has been devised by Levy, Rowntree and Marriott, 1 as follows: "Serum or whole blood is placed in small collodion sacs and dialyzed for five minutes against a normal saline solution. The hydrogen-ion concentration of the dialysate is determined by adding a few drops of phenolsulphonephthalein and comparing the color thus obtained with that produced on adding the indicator to a solution of kmrnn hydrogen-ion concentration. A determination can be carried out in a few minutes and requires only from 1 to 3 c.c. of blood serum. The results are, in general, comparable with 1 Levy, Rowntree and Marriott: Arch. Int. Mod.. 191"), xvi, p. .'3s9. and as abstracted in the Johns Hopkins Hospital Bulletin, 1915, xxvi, p. 114. 222 ritlXK, BLOOD AND REKFIRATIOX IN DIABETES those obtained by the use of electrometric methods." (See original article for discussion of its significance and for references to the subject.) S. Titratable Alkalinity of Blood. Any method for the detection of acidosis which depends upon the examination of the urine is misleading and often inapplicable. Realizing this fact, Sellards 1 undertook to supply a simple clincial method for determining the alkalinity of the blood by titration. To obviate the difficulties of earlier titration methods he developed a purely qualitative technic in which all normal sera would react distinctly alkaline, whereas pathological sera would react in many gradations from slightly alkaline to neutral or even acid. The difficulties encountered were: (1) the selection of an indi- cator to which all normal sera were alkaline after the conversion of the acid carbonate (XaIl(X) :i ) of the blood to normal carbonates (XaaCOs); (12) the removal of the interfering proteins; (o) the substitution of another solvent for water to reduce the amount of hydrolysis and utilization. Sellards finally employed a solution of 0.5 per cent, strength ordi- nary diacid phenolphthalein, made up in absolute alcohol. The removal of protein and the substitution of another solvent for water were combined in one step by precipitating the serum with alcohol. The determination is carried out in three steps by obtaining (1) the reaction of the residue after incineration in platinum; (2) the reaction after dilution with water and boiling; (.'>) after precipita- tion of the protein with alcohol, the reaction of the alcoholic filtrate. The first determination is carried out simply by ashing one or two drops of serum at a dull red heat in a platinum crucible. The residue is dissolved in water and the reaction tested with phenol- phthalein. In each of the other two applications 1 c.c. of serum is added, drop by drop, to 125 c.c. of absolute ethyl alcohol. Tin 1 mixture is shaken thoroughly and filtered into an evaporating dish. Three or four drops of phenolphthalein are added to the filtrate, and one of these evaporated slowly to dryness on a steam bath. This alcoholic filtrate, upon concentration, rapidly takes on a deep red color. This change takes place in all gradations; from the normal, in which color is produced early and retained for five to eight hours with constant heating, to the pathological, where the color is lost after a few minutes of heating, or may not appear at all where the depletion of alkali is marked, as in extreme acidosis in diabetes. The author concludes that in all normal individuals the unheated 1 Bollards: Johns Hopkins llo.-p. Bull., 1'Jll, xxv, p. 101. EXAMINATION OF TUK RESPIRATION 22:} alcoholic filtrate of the serum is neutral, or slightly acid, to phenol- phthalein, and becomes alkaline upon heating. The reactions are listed in the following stages and merge grad- ually into one another: 1. Persistence of red color in either alcoholic solution or dry residue for several hours. 2. Absence of color in alcoholic solution or transient appearance of color in the residue, but with good coloration in aqueous solution. 3. Total absence of color in alcoholic solution, in the residue on evaporation or after the addition of water, but with distinct color in the aqueous solution before removal of protein. 4. Total absence of color in all stages, including the aqueous dilution, without the removal of protein. (This latter test is made with a 1 to ]() dilution of the whole serum in water.) 9. Total Solids. Myers and Fine 1 show how the total solids of the blood may be very simply and accurately determined. They use a weighing bottle of the ordinary style, with stopper having a glass loop to which a block of filter paper may be fastened with a small wire hook. The block of filter paper is suspended and the bottle dried and weighed. From a small pipette 0.3 to 0.0 gram of the well-mixed blood is allowed to flow on the filter paper rapidly, the stopper quickly inserted to prevent loss of moisture, and the bottle weighed. The stopper is now tilted and the bottle placed in the drying oven at 105 (\ overnight. When convenient the bottle is cooled and again weighed. From the loss of moisture the total solids may readily be calculated. At the eighth meeting of the American Society for Clinical Inves- tigation, 1910 (see report, p. 29), Mosenthal described a method in which little tin boxes were used for the determinations. From his investigations he concluded that the body has a tendency to increase the water content of the blood as the blood sugar rises. Other factors like diuresis break in on this parallelism. C. EXAMINATION OF THE RESPIRATION. 1. The Technic. Two types of apparatus are employed to learn the exchange of carbon dioxide and oxygen in man: the respiration calorimeter and the respiration chamber, which are the same in principle, and the respiration apparatus. In the closed chamber of the calorimeter and the respiration chamber, the oxygen admitted and the carbon dioxide withdrawn can be accurately determined in periods from one-half to one hour's duration, but it is better to take the average of the results obtained in three successive periods. 1 Myers and Fine: Chemical Composition of the Blood in Health and Disease, Post-Graduate, New York, 1914-1915. 224 URINE, BLOOD AND RESPIRATION IN DIABETES In the new respiration chamber recently devised by Benedict, 1 and shown in the accompanying illustration, half-hour periods have proved very satisfactory. The calorimeter is cumbersome, expensive to construct and maintain, and the length of the experi- ment is not only disagreeable to the patient but disadvantageous in studying the results of rapid changes in the metabolism, which are desirable in a study of the utilization of foods. The respiration chamber is similar, but simpler in construction, is available for FKJ. S. General view of the complete clinical respiration apparatus installed in the respiration laboratory of the Brodbeck Memorial Cottage, X. K. Deaconess Hospital, Boston, Mass. Respiration chamber on the left. The thermometers, with reading-glass attachments, in the cover; kymograph for registering the muscular activity on the left. Universal respiration apparatus in the center. Below arc the rotary blower, the motor, and the Williams bottles; above are the spirometer, soda-lime bottles, Williams bottle, and petroleum manometer. At the extreme right are the barometer, oxygen cylinder, and Bohr meter under water. shorter periods, and serves all the purposes and has all the advan- tages of the calorimeter save that there is no direct measure of heat. However, this is not of consequence because it is possible to determine the heat given oil' by the patient accurately, but indirectly, from the oxygen inhaled or less accurately from the carbon dioxide given oil'. 1 Sec Benedict and Tompkins: Boston Mod. and Surg. Jour., 1910, clxxiv, p. 7, for a clear statement of the subject of "Respiratory Exchanges, with a Description of a Respiration Apparatus for Clinical Use." EXAMINATION OF THE RESPIRATIOX 22"} The respiration apparatus differs from the calorimeter and respira- tion chamber in that the patient is not in a closed chamber, but instead breathes through a nose- or mouth-piece. This is advan- tageous because the exchange of gases can be determined during short periods of fifteen minutes or less. It is disadvantageous, however, because, the periods being so short, errors at the begin- ning and end of the periods are magnified, and further, because of the individual breathing through a nose-piece or mouth-piece an abnormal state is introduced. Unfortunately, in each form of mouth- or nose-breathing apparatus the error of a leak falls chiefly on the oxygen, because the patient and the apparatus constitute a closed circuit, and any diminution in gas in this circuit must be offset by the addition of oxygen. A more troublesome source of error and one difficult to avoid arises from the possibility of the patient exhaling carbon dioxide, which has previously accumulated in the body, at a more rapid rate than corresponds with the oxygen inhaled. The patient is said to "pump out" carbon dioxide. This error could only occur in the relatively short periods which are employed witli the respiratory apparatus. There is also another error due to carbon dioxide which is lost by cutaneous respiration, and it has been calculated that this would lower the quotient 0.01 to 0.015. This also occurs only with the respiratory apparatus. Many pitfalls therefore lurk in the determination of the respira- tory exchange of an individual. The carbon dioxide is the more easily estimated of the two gases, and in early experiments on metabolism investigators attempted to estimate this alone. The determination of oxygen is far more difficult. In dealing with the respiratory quotient, which depends on the relation of these two determinations to each other, the problem is still more complicated and all statements regarding the respiratory quotient of individuals must be accepted with caution. The general picture of the respira- tory quotient in an individual is far more valuable as a guide to his true metabolism, if based on several experiments, than is the result of a single experiment. Similarly, it is probably safer to average the results of a series of cases than to attach great importance to figures obtained in an isolated experiment. 2. The Respiratory Quotient. The relation which the volume of carbon dioxide exhaled bears to the volume of oxygen inhaled constitutes the respiratory quotient. The respiratory quotient has been discussed at length in Section II, G, 2, to which the reader is referred. During the combustion of a food in the body a definite quantity of oxygen is consumed and carbon dioxide produced. The relation which the volumes of carbon dioxide and oxygen bear to one another constitutes the theoretical respiratory quotient of the substance. I will insert here examples showing how the respiratory 15 220 URINE, BLOOD AND RESPIRATION IN DIABETES *o h Cl CJ "-* _ -f (N X CT O c; c; ~ 1^ to ^r co 'C irr >rt ^ CC ic c i~ P-, 5 ? C v: r^ 1- y c^i X C-1 M ~'*~ u L* c ~ ~l X 1^ t- J >< C : 5 - L. * ? ;= It = ei cr. i^ i^ X X -> C t.e .*7 1 ~ ^t- -r "* -r "t 1 < < r S 2 a_. * ce *,- -,- (N CM C "H < ? ? , --^ -^ PC ^^. ^-. re ce ce : o < = ? ^ ; '. ) i i ce c^ C2 1^ ~ -' ~ZZ ^Z t^t^ X X ;c r" 5T ~ Q O T 1 1 1 c o ^ oo ^ * o -t 3- i .I 1 s L " "M C". t^ dd - M ^ c - X ^ 3 ". ce '^ C^ '^7 C-l X ~\ 1 - re ?i 1- C (M X ~ M 'ti - ?j ^ x r^ 5 X i = w - tt ^ *" *" , ei ei i ( ^ ^^ r. c re 'C (N ei x C^ O 1C -* *^ " -L. ^t i ^ r^ T'J ^ J ^ X ~ " " " ^,' ei V. C"- '^7 tC iC EC > ?1 - -' " ^^ ^ S SI : f,| T . "X 6l s -C -t -f. X s m .- u ei 'M r- -M _2 ,, "C c :.I - "t; rt ^ /. ^ "^ - f C f-~. C i ^ .5 S ~ S r-r x C *^ "~^ ^^ < K si <^ O x K 5 2 EXAMINATION OF THE RESPIRATION 227 quotient of alcohol and milk may be theoretically calculated, because the computations are based on different methods. The oxidation of alcohol (CgHsOH) requires 3 molecules of oxygen (3(O 2 ), for its complete combustion, us is indicated in the following equation : C 1 C 2 H 6 ()H =H4 +3(0 2 ) =3(H 2 0) +2(CCh). V / )hlme 2 " z) = Q.G7, respiratory H-O-H Volume 3(O 2 ) quoticnt , )f a i coh ol. The calculation of the theoretical respiratory quotient of milk is dependent upon the fact that milk is made up of known quantities of milk-sugar, fat, and protein, and that the quantities of oxygen consumed and carbon dioxide given off in the combustion of each of these substances have been determined. These values, as well as some others, are shown in Table 103. One liter milk may be considered to contain 50 grams of carbo- hydrate, 40 grams of fat, and 33 grams protein. C. (50 grains X 74(5.2) = 37,310.0 c.c. CO 2 F. (40 grams X 1431.1) = f)7,244.0 c.c. CO 2 P. (33 grains X 773.8) = 25,535.4 c.c. CO 2 120,089.4 c.c. COo C. (50 grams X 746.2) = 37,310.0 c.c. Oo F. (40 grams X 2013.2) = 80,528.0 c.c. O 2 P. (33 grams X 956.9) = 31,577.7 c.c. O 2 149,415.7 c.c. Oo 120,089.4 c.c. COo 14C) 41" " c c Oo = 0-803, respiratory quotient of milk. The calculations to determine the respiratory quotient of an individual are still simpler, and are shown in an experiment dis- cussed under the Total Metabolism in the following paragraph. 3. The Total Metabolism. The total metabolism, with its variations in diabetic patients at different stages of the disease, has been discussed in Section II, G, 1 . The calculations by which the total metabolism of an individual can be determined when the oxygen consumed and the carbon dioxide exhaled are known are illustrated in the data of the following experiment (Table 104): TABLE 104. NORMAL INDIVIDUAL (E. P. .!.). EASTING EXPERIMENT DECEMBER 23, 1914 WEIGHT, 65 KILOS. HEIGHT, 177.8 CM. CO? Oa Calories Duration. nor min. per min. Respiratory per kilo Period. min. sec. c.c. c.c. quotient. per 24. 1 15 6 152 192 0.79 20.40 2 14 59 150 194 0.77 20.51 3 15 153 196 0.78 20.77 Average = . 78 22X URINE, BLOOD AND RESPIRATION IN DIABETES "First, the calculation of the respiratory quotient of this individual should be determined, and this is done as follows: Volume CO., 152 c.c. T^-~r = O./O, respiriitorv mioticnt. \ olumc ();> 192 (..(. TABLE 105. THK ( 'ALOKIC VALUE OF OXE LITKU OF OXYOF.X FOR \'.\iuoi;s XOX-PKOTKIX RKSI>IKATORY QCOTIKXTS AND THK PKOPOKTIOXS OF THK EXKHfiY FROM ( \\KHOH VDKATK AM) I 1 ' AT M KTA HOLISM TO UK ASSIGNED TO CARBOHYDRATE AND FAT RESPECTIVELY. (WILLIAMS, KICIIK AXD LVSK.') Hespiratnry Calorics for 1 liter Oo. ( 'ar'nohydratc. Fat, quotient. Number. \A>. percent. ]>er cent . 0.70 . . . . 4.0SO 0.070SO 100.0 0.71 .... 4.000 0.07110 1.4 OS.Ii 0.72 . . . . 4.702 0.07231 -1.8 05.2 0.73 .... 4.714 0.07340 S 2 01. S 0.74 . . . . 4.727 0.()74C>0 11. (i SS.4 0.75 .... 4.730 O.C)7")74 lo.O S.l.O 0.7C) .... 4.752 0.070SS IS. 4 SI. (5 0.77 . . . . 4.7(54 0.()7S01 21. S 7S.2 0.7S . . . . 4.776 0. ('.701:5 25.2 74. S 0.70 .... 4.7SO ().(iS()24 2S.C> 71.4 O.SO .... 4. SOI ().()S13l> 32.0 C.S.O 0.81 . . . . 4.S13 O.OS247 35.4 M.Ci O.S2 .... 4.S25 O.C.S35S 3S.8 111. 2 O.S3 . . . . 4.S3S 0.()S4f)0 42.2 57. S O.S4 . . . . 4.S50 O.OX57S 45. ti 54.4 O.S5 .... 4..() 34.0 0.01 .... 4.031) O.(i0343 C.0.1 30.0 0.02 . . . . 4.04S 0.00450 72. S 27.2 0.03 .... 4. <)()() ().()0557 70.2 23. S 0.04 .... 4.073 0.00004 70. ti 20.4 0.05 .... 4. OS5 0.00771 S3.0 17.0 0.00 .... 4.007 O.OOS7S SO. 4 13.0 0.07 .... 5.010 O.OOOS5 SO.S 10.2 O.OS . . . . 5.022 0.70002 03.2 O.S 0.00 .... 5.034 0.70100 00.4 3.4 1.00 .... 5.017 0.70307 100.0 0.0 Since the body weight was CM kilos and ML' c.c. ()_> were inhaled for the whole body per minute, L J . '.).") c.c. ( 1 ( .)2 c.c. (K -=- CM kilos) were inhaled per kilo per minute, and as there are fourteen hundred and forty minutes in twenty-four hours ll'Sl.O e.c., or 1.2X4 liters oxygen (1440 X 2. ( .).")) \\-ere consumed per kilogram body weight per twenty-four hours. In order to express in terms of calories the heat which any consumption of oxygen represents, \\illiams, Kiche and Lusk have constructed a useful table (see 1 William-, Rk-he and Lu.-k: Jour. Biol. C'hoin., 1012, xii, p. .'557. EXAMINATION OF THE RESPIRATION 229 Table 105) showing the caloric value of a liter of oxygen for different non-protein respiratory quotients. The term "non-protein respira- tory quotient" is explained in a later paragraph. Consulting this table of Williams, Riche and Lusk, it will be seen that the equivalent in calories for 1 liter of () 2 for a non-protein respiratory quotient of 0.79 is 4.789. (As a matter of fact, the respiratory quotient above obtained is not the exact non-protein respiratory quotient, but this may be neglected unless the nitrogen excretion is distinctly abnormal.) 1 If we multiply 4.7X9 calories by the total number of liters of O 2 consumed per day, the number of calories per kilo per twenty-four hours is obtained, which would be in this case (4.7S9 X 4.284) 20.5 calories. It should be stated that this actually represents the metabolism for a period of fifteen minutes when the individual was under complete relaxation, and as quiet as possible. It by no means represents the average metabolism for the day. Even with this extreme basis of repose maintained dur- ing the experiment the weight of the oxygen inhaled was 550 grams and of the carbon dioxide exhaled 424 grams in the twenty-four hours. 4. The Non-protein Respiratory Quotient. If the nitrogen in the urine is known, one can calculate the amount of oxygen employed by the body for the oxidation of the protein which it represents, and correspondingly, the amount of carbon dioxide simultaneously produced. If these computed figures are substracted from the total oxygen and carbon dioxide obtained by direct experiment, the remainders represent the oxygen absorbed and carbon dioxide pro- duced by the non-protein respiratory metabolism. The relation of these to one another constitutes the non-protein respiratory quotient. In the table of Williams, liiche and Lusk the proportions of the total energy from the 1 katabolism of carbohydrate and fat assumed as produced from the two materials respectively for any known non-protein respiratory quotient between 0.70 and 1.00 are given. Thus, a respiratory quotient of 0.70 shows that 100 per cent, fat and no carbohydrate was oxidized, and a respiratory quotient of 1 .00 shows exactly the reverse. The calculations necessary for the determination of the non- protein respiratory quotient of an individual are given below. The example chosen for this purpose is that of a man undergoing a pro- longed fast at the Nutrition Laboratory. (See page 150 for the complete table.) Time. 5th day of fast 1 It will bo remembered that the respiratory quotient of protein is 0.81 which is nearly that here obtained for the total metabolism. 2 Calculated for twenty-four hours. 230 UlflXK, BLOOD AND RESPIRATION IN DIABETES The values used for the computation of the above data were as follows: ('<>:> <> 2 Xitrouen c.c. per min. grains per 21 liours 210 10.41 10.41 grams nitrogen represents the nitrogen excreted per 24 hours or 1440 minute's, and ( , -.--) 0.0072 gram is the 1 excreted \ 1440 mm. / nitrogen per minute. One gram of protein nitrogen, representing (>.() grams body pro- tein, 1 produces in its combustion 4750 c.c. ( O 2 , and consumes in its combustion 5010 c.c. (),. Therefore (4750 X 0.0072) 34.20 e.c. CO., are produced per minute, and (5910 X 0.0072) 42.55 c.c. O 2 are consumed per minute, as a result of the protein metabolism. In the experiments conducted that day, the total (X) 2 eliminated per minute was estimated at 175 c.c., and the total () 2 consumed per minute at, 240 c.c. These two amounts represented the total metabolism of protein, fat and carbohydrate for the day. If the oxygen and carbon dioxide per minute derived from nitrogen, which represents the protein metabolism, are subtracted from the total oxygen absorbed and carbon dioxide produced per minute, the remainders show the non-protein metabolism, or the metabolism which is base: I on the oxidation of fat and carbohydrate. 17."> c.c. ( '( )o per min. = total metab- 210 c.c. ()., per min. = total metal)- olism. olism. 34 = protein metab- t.'-i - protein metab- olism, olism. 141 = non-protein 107 = non-protein metabolism. metabolism. Volume 141 e.c. f '( )., = ().<2, non-protein respiratory quotient. (Ym.se- Volume 107 c.c. < ) 2 quently 4. S per cent, of the non-protein metabolism \va- due to carbohydrate and Of). 2 per cent, due to fat, according to Table 10."). . 5. Theoretical Respiratory Quotients as Calculated from the Diet. The theoretical respiratory quotient of a normal individual living upon protein and carbohydrate can be calculated as shown in the following table: TAHLK 10(5. TIIKOUKTICAI. KKSIMKATOKY QTOTIKNTS. Diet. 2 Cal. ( >_. TO, Protein, 100 grains . . . 414 <)5.0'.> 77. 3S Carbohydrates, 503 grams . 23(15 4(50.73 400.73 502.42 544.11 'In estimating the quantity of body protein burned from the nitrogen in the urine the equivalent should be employed instead of (i.2">. -'Ma2 and CO-> vary .somewhat from those given in the more recently compiled Table 106. 2 Lusk: Arch. Int. Mod., 1915, xv, p. 939. 232 URINE, BLOOD AND RESPIRATION IN DIABETES acidosis to be quantitatively estimated promptly have proven most helpful in treatment. At present three such methods are available for the detection of the (X) 2 in the alveolar air, the Plesch-Higgins, the Friderieia and the Marriott. Probably the Plesch-Higgins is the most accurate, and those familiar with it in large hospitals prefer it. However, the other two methods are quite satisfac- tory, and the techuic of either one can be learned by physician or nurse in a few minutes. In the discussion of the carbon dioxide capacity of the blood plasma, reference is made to the fact that the carbon dioxide of the blood diffuses so readily into the alveolar air of the lungs that the estimation of the latter in nearly all cases gives an index of the former. Because of the simplicity of the determination of the carbon dioxide in the alveolar air, it is today the best quantitative method which the physician has for the estimation of the acidosis of the patient. Normally, the carbon dioxide tension of the alveolar air varies between MS and 4o mm. mercury, 5.3 to 0.3 per cent. If abnormal acids are present in the blood, these displace a proportionate amount of carbon dioxide, and as the carbon dioxide tension in the alveolar air bears a direct relation to that in the blood, it is evident that the carbon dioxide in the alveolar air will vary likewise. A low carbon dioxide tension of the alveolar air therefore indicates an acidosis. If the carbon dioxide tension lies between 3S and 32 mm. mercury a slight acidosis is present, between 32 and 2X a moderate acidosis, and if it falls below 25 mm. Ilg. the acidosis is extreme. The lowest value with recovery in my group of cases has been 15 and the lowest obtained in the series was 9, and that occurred in a patient in coma. Ilornor 1 found in an analysis of 300 observations of the alveolar COo tension of my cases that when the carbon dioxide tension of the alveolar air was less than 25 mm. mercury the ammonia was 3 grains or more, when the carbon dioxide tension varied between 25 and 33 mercury the urinary ammonia varied between 3.7 grams and 1.15 grams, and when the carbon dioxide tension was 33 mm. mercury or over the ammonia was less than 2 grams. He also studied the carbohydrate balance in these cases, and learned that it was invariably negative when patients showed a tension of less than 25 mm. mercury, and in three-quarters of the cases was nega- tive when the tension was between 25 and 33 mm. mercury, but that at a higher tension the carbohydrate balance was usually posi- tive and invariably so when the carbon dioxide tension was above 30 mm. mer-nry. The ferric chloride reaction was also observed, 1 Hornor: Boston Mod. and Surg. Jour., 1910, clxxv, p. 148. EXAMINATION OF THE RESPIRATION 233 and was found to vary from negative to strongly positive, not only when the carbon dioxide tension of the alveplar air was at 30 mm., but also at 23 mm. The absence of the ferric chloride test at the lower figure speaks in favor of its value rather than otherwise, because it has been discovered that in periods when the patient is on a low caloric diet the CO-, in the alveolar air is not a reliable index of the extent of acidosis of the patient. (See p. 215.) Case No. 1120, a child with onset of diabetes at the age of six years, entered the hospital eight months later as he was recover- ing from definite diabetic coma. Under treatment he thoroughly recovered from acidosis, but upon repeated trials increase of food led to the return of both sugar and acidosis. In one of these attempts to build up his nutrition severe acidosis developed on a diet of 42 grams protein and 70 grams fat. Typical hyperpnea without cyanosis developed and repeated observations showed the CO 2 tension of the alveolar air to be 15 mm. Hg. By fasting and without the use of alkalis the patient promptly came out of this state, but he later died from inanition without a trace of acidosis, representing the third mortality from inanition which I have had. The importance which should be attached to varying degrees of acidosis as shown by the CO 2 tension of the alveolar air is shown by Table 108, which gives the results of all cases since Dec. 1, 1915, in which the tension has been found to be below 25 mm. Hg. None of the patients cited in Table 108 took alkalis, and hence the table is an index of this method of treatment. The lowest case in our series showed 9 mm. Hg. tension and with recovery 15 mm. Ilg. tension. When the patient takes sodium bicarbonate, the carbon dioxide tension is affected and it may rise to above normal. Case Xo. 513, p. 353. (a) Fridericia 1 Method. This method possesses the advantages of being simple and involving the use of apparatus which may be easily transported to the bedside. One hundred c.c. of alveolar air are collected in a closed chamber and then cooled from the temperature of the body to that of the room. The carbon dioxide in this air is then absorbed with a 20 per cent, aqueous solution of potassium hydrate, thereby creating a partial vacuum which in turn is equal- ized with water. This water is then subjected to atmospheric press- ure, when the amount of carbon dioxide replaced by water can be read in percentage of atmospheric air by reading the height in cm. to which the column of water has risen in the closed 100 c.c. chamber. This percentage may be changed to millimeters of mercury pressure 1 Fridoricia: Berliner klin. Wchnschr., 1914, li, part 2, p. 1268. 234 URINE, BLOOD AND RESPIRATION IN DIABETES TAHLE 108. RESULTS OF TREATMENT OF ACIDOSIS IN PATIENTS WITH ALVEOLAR C() 2 TENSION 25 MM. MERCURY OR LESS. Fatal. Immediate recovery nit later fatal. Heeovery. Alive May, 1917. Alveolar air No. Date. O >2 tension, mm. llfi. No. Date. Alveolar air C( >2 tension, mm. llfi. No. Date. Alveolar air C( >2 tension, mm. UK. 904 April 29, 191 (i Died 4 days later 1C' /14 2 Coma 7(15 Def. G, 1915 Jan. 24, 1910 25 Oet. 29 21 21 24 Coma 9S3 Feb. 1, 1910 2 3 23 25 939 Nov. 30, 191(1 Died 2 days later 20 Pulmonary embolism 1070 June 23, 191(1 24 July 1 10 20 21 25 25 9 12 July 12, 191(1 13 1(1 17 18 27 AUK. 3 Nov. 5 20 17 24 21 24 24 24 Coma 1004 Feb. 28, 191(5 29 Mar. 1 Died samp day ( 4.00 P.M. 1 9.30 P.M. f 0.00A.M. 1 11.00 KM. 1-1 9.00 A.M. Coma 1011 Mar. 10, 1910 29 April 3 13 10 20 28 Oct. 25 29 25 24 23 25 23 ,'24 23 ,'22 23 25 25 24 24 25 9(1(1 Jan. ">, 191(1 25/24 July Coma 1010 Mar. 17, 191ti IS Died 5 hours later 20 /20 14/14 Coma 974 Jan. 20. 191(1 25 24 24 Oft. 4 24 20 Coma 1005 Mar (1 191(1 23 /24 19 Coma April 4 July 5 Died 43 hours later 99(1 1120 Feb. 1(1, 191(1 21 25 27 Mar. 1 Oet. 13 24 24 24 Coma 1007 Jan. 5, 1910 25 1032 April 15. 1910 25 1053 May 15, 191(1 1(1 17 is 19 20 Died 14 hours later 21 22/19 18/21 15 '14 /14 12/12 1 1 Coma 1111 Sept. 1(1, 191(1 23 ,25 1 151 Oct. 30, 191(1 25 Sept. (1, 191(1 7 11 Oct. 11 14 27 2S 29 Nov. 1 2(1 21 is 29 20 15/15 24 33 Coma 11 M Nov. 10, 1910 24 119(1 Dee. S. 1910 9 10 11 12 13 21 22 /24 IS/20 20 21 24 1143 (let. 7, 191(1 Died within 21 hours 14/12 Coma 1209 Jan. 2, 1917 24 10.S5 Oct. 3d. 191(1 20 Nov. 1 5 Inanition 1214 Dee. 30, 1910 25 7SO Jan. 11, 191(1 IS/20 12 21 1 ^ 25 Mar. 31, 1917 'Tuberculosis ? 1220 Jan. 23, 1917 'A.M. 2P.M. EXAMINATION OF THE RESPIRATION 235 by multiplying by the difference between barometric pressure at the time of the test, and this varies in Boston between 770 mm., and 750 mm., and the tension of aqueous vapor at 37.5 (\, which is 48 mm. mercury. This will make a factor which lies between 722 and 702. As the reading of 700 is much the more common at C sea level, for clinical purposes the factor 715 =*= may be used satis- factorily. The patient should be in the same position and quiet for ten minutes prior to the performance of the test. After a normal inspiration the end A of the apparatus is in- serted between the lips and the patient is instructed to expire 230 I'RIXE, BLOOD AX I) ItEM'IHATIOX IX DIABETES forcibly through the apparatus Avith cocks (' and I) open, so that there is a free passage from .1 to li. The tube remains in the mouth throughout the entire expiration, and the cock (' is then closed, thus retaining between cocks (' and I) the last 100 c.c. of expired air. (As the exchange of air in the upper respiratory passage is 200 c.c., and the exchange of air from the alveoli isSOO c.c., it is plain that with any care at all a sample of alveolar and not upper respiratory air will be obtained.) The apparatus is now immersed in a glass tank of water at room temperature and allowed to remain there five minutes'. The best way to obtain water at room temperature is simply to keep the glass tank in the room with the patient for several hours before the test, though with an ordinary thermometer one can easily adjust the temperature of the water to that of the room. At the end of five minutes about 10 c.c. of 20 per cent, aqueous solution of potassium hydrate are poured into the apparatus through orifice H. A little of this potassium hydrate will leak through the hole in cock J) to chamber ('-]). Now cock I) is turned to the left so that chamber (' I) is closed and chamber B-l) is also closed. The small amount of potassium hydrate in chamber CD is shaken in the chamber for a moment. Then with apparatus in upright position cock I) is turned so that there is a continuous passage from (' and />', and the amount of potassium hydrate which will run into the chamber (' J) is allowed to do so. Now cock J) is turned to the left until B-])-E is a continuous passage, and in this way potassium hydrate is allowed to escape into the water tank. Chamber (' J) still contains 2 or )> c.c. of potassium hydrate solution, and should be thoroughly washed with this solution. F,very point in the surface of chamber ('-]) must be touched by the alkaline solution. This is accomplished by shaking very thoroughly the potassium hydrate in chamber C-J). The apparatus is again immersed in the tank of water, cock I) is turned to the left until water rises into (' I) through E-D (', and the apparatus left in the water five minutes. At the end of this time the apparatus is raised until the bottom of the meniscus of the water in chamber (' D is level with the top of the water in the tank. Xow cock 1) is turned to the right until water runs through /.' -D-B to the level of water in chamber (' 1), which is now closed. Then cock I) is turned farther to the right until (' D B is a continuous chamber. The apparatus is then again immersed to the bottom of the glass tank and the water in the arm /> I) of the apparatus should be at the same level with the water in the chamber (' D and continuous with it. If this is not so, then the amount of water in />' 1) should be changed until it reaches the height of the column of water in (' J). The reading is now taken in cm. of the height to which the column of water stands in ('-- J), and this is so gradu- EXAMINATION OF THE RESPIRATION 237 ated as to represent the percentage of CO 2 which was absorbed by alkali and replaced by water. This completes the test. The apparatus is prepared for the next test by opening cock C so that A to B is a continuous passage. The fluid in the apparatus is allowed to escape. Orifice B is put under the faucet and cold water allowed to run through the apparatus, taking care to shake sufficiently at the time, so that water touches all of the inside of the apparatus. Repeat. Then pour through orifice B about 10 c.c. of 4 per cent, solution boric acid. Rinse the apparatus very thoroughly with this acid so that there shall be no alkali remaining adherent to its sides. Wash again with cold water. Leave the apparatus so that the cocks to .1 and B are open, thereby allowing any water in the apparatus to drain out. From the above it will be seen that the apparatus necessary is, first, the Fridericia appliance, 1 a glass tank, whose depth is equal to the length of the Fridericia apparatus, a wash bottle containing 20 per cent, solution of potassium hydrate and another wash bottle containing 4 per cent, solution of boric acid. It is convenient to add an indicator, such as alizarin or litmus, to the alkaline and acid fluids. (/;) Marriott's Method. 2 A very simple method has been devised by Marriott for the determination of the CO 2 in the alveolar air. The apparatus is portable and the test requires but a few moments for execution. Two procedures are involved: the collection of the alveolar air and the analysis of the sample. "Principle of the Method of Analysis. The method depends on the fact that if a current of air containing carbon dioxide is passed through a solution of sodium carbonate or bicarbonate until the solution is saturated, the final solution will contain sodium bicar- bonate and dissolved carbon dioxide. The reaction of such a solu- tion will depend on the relative amounts of the alkaline bicarbonate and the acid carbon dioxide present. This, in turn, will depend on the tension of carbon dioxide in the air with which the mixture has been saturated and will be independent of the volume of air bloicn through, provided saturation has once been attained. High tensions of carbon dioxide change the reaction of the solution toward the acid side. Low tensions have the reverse effect; hence the reaction of such a solution is a measure of the tension of carbon dioxide in the air with which it has been saturated. "The reaction of such a solution may be determined by adding to it an indicator such as phenolsulphonephthalein, which shows over a considerable range of reaction definite color changes. A certain color indicates a certain reaction. 'Apparatus may be purchased from Emil Grciner, 55 Fulton Street, New York. 2 Marriott: Jour. Am. Med. Assn., 191G, Ixvi, p. 1594. 238 URINE, BLOOD AND RESPIRATION IN DIABETES "Solutions of a given reaction may be prepared by mixing acid and alkaline phosphates in definite proportions. Such solution may be kept unaltered for long periods of time and can be used as standards for comparison." "Collection of the Alveolar Air. With an ordinary atomizer bulb, which will deliver approximately 50 c.c. of air, force approximately (i(K) c.c. of air into the rubber bag and clamp the outlet tube with the pinch-cock. While the subject is at rest and breathing naturally and at the end of a normal expiration, place the tube in the subject's mouth and close his nose, allowing him to breathe from and into the bag four times in twenty seconds, emptying the bag with each inspiration; the observer should indicate when breathing should be in or out. More frequent breathing will not greatly alter the results. After breathing twenty seconds, at the end of an expira- tion and while the bag is inflated, clamp the tubing with the pinch- cock and use the air contained in the bag for analysis. The analysis should be made within three minutes, as carbon dioxide rapidly escapes through rubber. "In the case of comatose patients, the rubber bag should be inflated with approximately 1000 cubic centimeters of air. The comatose patient should be allowed to breathe out of and into the bag for at least thirty seconds, since it is not feasible to have him completely empty the bag of air, at each inspiration. It is necessary to use some form of mask. "A mask is also necessary for collecting alveolar air from infants and comatose patients. This may be improvised, as suggested by Marriott, by means of a nipple of a wide-mouth (Ilygeia) nursing bottle and a piece of thin rubber tissue (dental dam)." "Tcchnic ;") to 40 per cent. On the other hand, if the greatest possible care is taken to be as quiet 1 This is a larw caloric, often written with a capital ('; a small caloric deals with 1 ^nnii instead of 1 kilogram. ( 240 ) THE DIET OF NORMAL INDIVIDUALS 241 when erect as horizontal, the difference may be only 8 per cent. If the individual is in a comfortable steamer chair or propped up in a semireclining position with a back rest, the metabolism is 3 per cent, less than when lying flat in bed. 1 At the Nutrition Laboratory 89 normal men eliminated on the average 25.5 calories at rest per kilo body weight per twenty-four hours, as calculated from the oxygen or carbon dioxide, the total output per twenty-four hours being 1009 calories, and 08 normal women eliminated on the average 24.9 calories, calculated per kilogram of body weight per twenty-four hours, a total output of 1355 calories for the twenty- four hours of the day. Too often in dietetic computations it is assumed that the caloric needs of the body can be accurately esti- mated. As a matter of fact, the error in such computations is con- siderable, and it is absurd to expect to compute the needs of the individual when up and about, whether normal or diabetic, more closely than within 10 to 20 per cent, of the real value. As an illustration of the amount of work which can be performed by 1 calorie of energy, I learn from my friend, Professor Benedict, that the expenditure of 1 calorie of heat is required to rise from a sitting position in front of a door, turn the key in the door, and sit down. A single sixteen candle-power carbon lamp gives off in heat the equivalent of about 45 calories per hour. To walk one hour on a level road at the rate of 2.7 miles an hour requires 100 calories above that of the resting metabolism (Lusk). The amount of energy expended in walking on a level road can be calculated with more accuracy as follows: If the individual weighs 00 kilograms and walks 1000 meters (3281 feet) he is said to have travelled (00 X 1000) 00,000 horizontal kilogram-meters. For each horizontal kilogram-meter 0.0005 calories are required. This would represent an expenditure of (60,000 X 0.0005) 30 calories, which should be added to the resting metabolism of the individual during the time required to walk the given distance. A man weigh- ing 60 kilograms who walks four miles an hour would expend (60 (kilograms) X 4 (miles) X 1009.3 (meters in 1 mile) X 0.0005 (calories per horizontal kilogram-meter)) 193 calories. To this figure should be added 60 calories, on the basis of 1 calorie per kilogram body weight per hour, which would have been expended by the individual at rest if one desires to obtain the total expendi- ture of heat. The basis for calculations of this type is direct measurement. It should be remembered that the weight 60 kilo- grams represents the naked weight of the individual plus the weight of his clothes. If a pack weighing 5 kilograms is carried, then the calculations must be on the basis of 65 kilograms. 1 Sonderstrom, Meyer and DuBois: Arch. Int. Mod., 1910, xvii, p. 872. 16 242 DIET IX HEALTH AXD IX DIABETES If the individual ascends a height, the calculations are somewhat dill'erent. The unit is the vertical kilogram-meter. The weight of an individual in kilograms multiplied by the height ascended in meters gives the vertical kilogram-meters. The heat equivalent of the mechanical work evolved in 420.5 vertical kilogram-meters is 1 calorie. Thus, if an individual of 00 kilograms body weight walks up ten flights of stairs each o meters high, he would expend (GO X 10 X :->) ' , . . . _ 4.2 calories. As the mechanical emeiencv or the 42o.o body is only about 20 per cent., it is necessary to multiply this figure by 5 in order to determine the actual energy expended (4.2 X ">) namely, 21 calories. Here we are dealing with figures based on estimate and not on direct experimentation. To this figure must be added the calories necessary for horizontal progression, as well as the calories required during the same period of resting metabolism. Two calorics might therefore l>e added for the forward progression, and '> calorics more for the three minutes of time of resting metabolism. In other words, a man of 00 kilograms walking up ten flights of stairs each o meters (10 feet) high, in three minutes would expend the heat equivalent of 20 calories or about .'5 calories IXT flight. The basal metabolism of a fasting individual weighing 00 kilo- grams (l.'!2 pounds) has been determined by Benedict (see pp. 150 and 151) and found to vary between 1845 and 1318 calories during the thirty-one days of the fast. It has been estimated that an individual weighing 70 kilograms (154 pounds) requires, under the varying conditions set forth in Table 110, the given number of calories: TABU: 110. CALI/UIKS Ri;o.riKKi> DTKIM; TWKNTY-KOUR Horns HY AX Anri/r \Yi-:i<;m\<; 70 KILOGRAMS (lot POUNDS). ( 'oiidilioii. At rest ....... 2:>-:>() 11-14 17.->0-21()0 Liiiht work ...... 3.V10 1<)-1S 24nO-2SOO Mode-rate work .... -lO-lo 1S-20 2X00^150 Hard work ...... l.Vf.O 2027 :UoO-4200 Farmers in various parts of the I mted States have been shown to consume on an average ;>500 calorics. The total metabolism of individuals lias usually been reckoned per kilogram body weight, no standard for comparison of the metabolism according to body surface having been generally accepted. Recently I HiBois has constructed two formula.', the linear formula and the height and weight formula, by which it is possible THE DIET OF NORMAL INDIVIDUALS 243 very conveniently and accurately to make such comparisons. 1 Means 2 has confirmed their work. Below is given the table of Gephart and DuBois. According to their formula, 40 calories per square meter per hour represents the metabolism of normal men, and 37 calories that of women. The average metabolism of fat and thin subjects is the same according to surface area when the surface area is correctly measured. TABLE 111. STANDARDS OF NORMAL METABOLISM. AVERAGE CALORIES PER HOUR PER SQUARE METER OF BODY SURFACE. According to linear and Age in height-weight Subjects. years. formulas. Boys 12 to 13 49.9 Men 20 to 50 39 . 7 Women 20 to 50 36 . 9 Men 50 to 60 Women 50 to 60 Normal boys twelve or thirteen years of age have been shown by DuBois 3 to produce 25 per cent, more heat than adults when com- pared according to his linear formula of body surface. Recently Gephart 4 in a study of the dietary at St. Paul's School came to the conclusion that the boys whose average ages fell between thirteen years and six months and sixteen years and one month ate approx- imately 5000 calories daily. For these same boys, the calculated basal metabolism would amount to about 1700 calories. It is there- fore evident that children require proportionately more food per kilogram or pound body weight, and I have endeavored to allow for this in the following empirical table: TABLE 112. CALORIC NEEDS OF CHILDREN DURING TWENTY-FOUR HOURS. Age in Weight. Calories, Total years. kg. pounds. kg. pounds. calories. 2 12 26 80 36 960 6 20 44 70 31 1400 12 36 SO 50 23 1800 2. Composition of the Diet. The ordinary diet for a man at moderate work would contain about 400 grams of carbohydrate, 100 grams of protein (equivalent to 16 grams of nitrogen, approx- imately eliminated as 14 grams in the urine and 2 in the feces) and 100 grams of fat. This would amount to 2900 calories in the twenty- 1 Gephart and DuBois: Arch. Int. Med., 1910, xvii. p. 902. 2 Means: Arch. Int. Med., 1916, xxi, p. 2G3. 3 DuBois: Arch. Int. Med., 1916, xvii, p. 887. 4 Gephart: Boston Med. and Surg. Jour., 1917, clxxvi, p. 17. 244 DIET IN HEALTH AND IN DIABETES four hours, or about 40 calories per kilo for an individual weighing 70 kilograms. These figures would be proportionately reduced both for those of lower body weight and for those with lighter occupations, who would require nearer )>0 calories per kilo. As age advances the metabolic requirements are lessened, thus if 12000 calories are required at thirty years, 1800 calories will suffice at seventy and .1000 at eighty years of age. TAHLI: llo. Tin-: I'ltoroirnox OF CAUBOHYDKATK, PKOTKIN AND I-'AT IN TIIK NORMAL DIKT. Quantity, Calorics, Total Food. grams. per gram. calorics. Carbohydrate . . . . . 400 4 1600 Protein . . . . 100 4 400 Fat 100 1) 900 The figures given above are very different from the old Yoit standard in which the carbohydrate was placed at 500 grains, the protein at 125 grains and the fat at 55 grams. The more I observe the diets of non-diabetic patients, the more I have come to believe that adults eat less than has generally been supposed. Chittenden, 1 in his painstaking and scientific manner, accom- plished an immense amount of good when he showed that people ordinarily consumed much more food than physiological needs demand. lie suggests that it is more than, probable that this excess of food is in the long run detrimental to health, weakening rather than strengthening the body and defeating the very object of nutrition. ?>. Carbohydrate. From the preceding statements it will be seen that 55 per cent, of the energy of the diet of the normal '.ndividual consists of carbohydrate. These figures are only approximate, but they leave no doubt as to how large a place sugar and starch occupy in the daily ration (see p. 50). The accompanying illustration shows graphically the relative caloric value of the different foodstuffs in the total diet. The normal diet, Kic. 10. The relative calorie, value of protein, carbohydrate, and fat in a normal diet . 1 Chittenden: Physiological Economy in Nutrition, New York, 1904, p. 474. THE DIET OF NORMAL INDIVIDUALS 245 The proportion of carbohydrate in the normal diet varies in different countries, reaching its maximum in the tropics and its minimum in the arctic zones. The people in India take 484 grams carbohydrate, while the Eskimos get along very comfortably upon 52 grams. I have arranged Table 114 modifying somewhat a similar table of Lusk's. 1 It shows well the adaptability of different races to different diets. That the Eskimos live upon 52 grams of carbohydrate daily should greatly encourage diabetic patients. All who treat diabetics should be very thankful that there is a race of Eskimos through which proof is afforded that it is perfectly possible to maintain life on a diet in which carbohydrate is largely replaced by fat. 2 TABLE 114. VARIATIONS IN DIET ACCORDING TO RACE. Weight, Protein, Carbohydrate, Fat, Total Rare. kilos. gm. gm. gin. calories. Eskimo . . . Go 2S2 52 141 2004 Bengali . . . 50 52 484 27 2390 European ... 70 118 512 (>5 3055 American 3 ... 70 100 400 100 2900 The composition of the diet also varies in the same race from time to time and this has been interestingly described by Mendel. 4 Attention has already been called to the increase in the consumption of sugar in the United States during the last century. Rubner noted that the consumption of meat per capita in Germany had risen three and one-half times during a hundred years. 4. Protein. The quantity of protein in the normal diet is probably not far from 100 grams. Professor Cannon writes me that he has "reports from 40 different students who have lived carefully and made thorough reports of their conditions during four days, on the last of which unusual exercise was taken. The average excretion of nitrogen for these 46 students, on the four days in succession, was 12, 12.10, 12.88 and 12.29 grams." If we raise the nitrogen to 14 grams by allowing the difference for that eliminated in the feces, the total elimina- tion of nitrogen would be 14 grams, and this would represent the equivalent of 88 grams protein (14 X 0.25 = <88) as the normal metabolism of these students. It will be found of great advantage to accustom oneself to estimate the protein content of the diet of 1 Lusk:The Fundamental Basis of Nutrition, Yule Univ. Press, 1914, p. 31. 2 It must be acknowledged, however, that today the Danish Government supplies the Eskimo with more than this quantity of carbohydrate. Krogh: A Study of the Diet and Metabolism of Eskimos, Copenhagen, 1913. 3 Added by E. P. .1. 'Mendel: Changes in the Food Supply and Their Relation to Nutrition, Xe\v Haven, Yale University Press, 1916. 246 DIET IX HEALTH AXD IX DIABETES patients in terms of nitrogen as well as in piotein, and to control one's calculations by determining the nitrogen in the urine. Such controls, however, will not be of value if the patient is undergoing rapid changes of diet or weight. Y\ hen Professor ( 'hittendeii's 1 epoch-making studies appeared, many felt that he went to extremes, but today his statements appear very moderate. He wrote "Food requirements must with necessity vary with changing conditions . . . all the results so far obtained in this investigation with a great variety of persons point to the conclusion that the real demands of the body for protein food do not exceed .">() per cent, of the amount generally consumed. Half of the IIS grams of protein food called for daily in ordinary dietary standards is quite sufficient to meet all the real physiological needs of the body. . . ." Instead of a standard of IKS grams protein today, 100 grams is the commonly accepted American standard. Cannon's students ate only about X8 grams. All in all our only retort to the New Haven Scientist can be that we were really rather more temperate than he thought (save for those boys at St. Paul's School), and we will all agree "excess means waste, but of far greater importance is the unnecessary strain placed upon the body by this uncalled-for excess of food materials which must be gotten rid of by the expense of energy which might better be conserved for more useful purposes." The term protein is a broad one, and until recent years in dietary programs comparatively little account has been taken of its com- ponents. The old formula for hemoglobin rrosH^o.'iOassNigsFeSg+T' though by no means accurate, gives some idea of its complexity. The individual nitrogenous substances (amino- acids) which are found in protein are seventeen in number, and the different proteins vary in the percentages of each of these present. The complete proteins namely, those which are essential for constructing the different parts of the body contain the same amino-acids and such proteins are represented by milk, meat, fish, and egg proteins. In gelatin and some vegetable foods, important amino-acids are lacking and it is therefore important that such incomplete proteins be not. furnished diabetic patients, particularly children, when on a lo\v diet. Osborne and Mendel 2 have carefully studied the value which various proteins exercised upon growth and have shown that whereas upon milk or a mixed diet young rats grew normally, if a single incomplete protein gliadin was given, they appeared well, but remained dwarfs, but resumed normal growth when returned to milk or mixed diet. 1 ( 'hit tiMidon: I. <><. cit ., p. 17.~>. 2 <)sl)oriio and Mendel : .lour, of Biol. ('horn., 1913. xv, p. '.'>[ 1 ; 1!1."), xxiii, p. 439; 1911), xxiv, p. 37; 191!'.. xxv, p. 1; 191(>, xxvi, pp. 1 and 293. THE DIET OF NORMAL INDIVIDUALS 247 The quantity of protein necessary to keep in nitrogenous equilib- rium lias been found by Thomas 1 to vary according to the source. TAHI.K 11"). RKI.ATIVK QUANTITIES OK PROTEIN RKQUIKKI) TO MAINTAIN NITROGENOUS KQUILTHKITM. Protein. (Jni. Meat . 38 Milk :il Rice Indian corn Potato Bean . Bread Furthermore, a ])ossibility exists that protein may be formed in the body from the addition of nitrogen to the decomposition products of sugar (methyl-glyoxal, lactic acid, pyruvic acid) in the form of alanin. 5. Fat. The quantity of fat in the normal diet varies, partly from choice and partly from economic reasons. In general, in those cases where the carbohydrate in the diet is high, the fat is low, and vice versa. The Yoit standard placed the fat at 55 grams, but a series of 1MOO dietary studies of families, carried out among different races and in different countries, showed that the average quantity of fat eaten was about K>5 grams (4.5 ounces) per. person per day, the variation recorded being from 45 to 390 grams per person per day. 2 The more agreeable varieties of fat, such as butter, cream and oil, are expensive foods. Fat is also a concentrated food, not only because it has twice the caloric value of either carbohydrate or protein, but because it occurs more frequently in pure form. Oil, butter, and lard contain little water, whereas except for pure sugar and starch most carbohydrates and proteins are diluted five to ten times with water. The chief source of error in calculating the total caloric value of the diet and especially of the diabetic diet is in the estimation of fat. Thus for many years I have considered that on an average lean meat and fish contained 10 per cent, fat and have taught patients to reckon > grams of fat to each ounce of meat or fish. This figure is unquestionably correct for poultry and very lean meat, and is very high for most varieties of fish, such varieties as cod, haddock and flounder containing only 1 per cent., but an analysis of a mixture of ten portions of cooked meat exactly identical with similar por- tions about to be served patients at the New England Deaconess Hospital was made at the Nutrition Laboratory, and showed 14.4 1 Cited l>y Lusk, lor. cit., p. 20. 2 Holmes and Lang: Fats and Their Economical I ~se in the Home, I . >. Dept. Agriculture, 191G, Bull. No. 4G9. 24S DIET IN HEALTH AND IN DIABETES per cent. fat. It is therefore better to reckon five grams of fat to the ounce of meat when the patient is taking several varieties. Bacon is variously estimated, but I have adopted 50 per cent, as an average value for the fat in cooked bacon. Analyses from different sources show the fat of cooked bacon varies from 87 to 7!) per cent. Portions of bacon lose varying quantities of weight in the cooking, as shown in the following table: TABLK lit). Loss OF WKKJHT OF BACO.V mnuN<; COOKING. Uncooked, Cooked, Loss, grains. grams. grams. per cent. SO 46 34 43 200 100 100 50 f>0 17 33 60 60 30 23 10 37 20 62 67 Through the courtesy of Miss Katherine Blunt, I am able to insert an extensive series of analyses of bacon. These make it very evident that our caloric values for bacon must be very crude unless all the fat in the frying pan is utilized. The protein appears to vary even more than the fat in the cooked bacon. Figures like these upset our preconceived ideas about dietary values and show the necessity of carefully arranged dietary studies upon diabetic patients. TABLK 117. KFFKCT OF COOKING UPON VARIOUS KINDS OF BACO.V (IVATHEHINK BLUNT). Kind of bacon. Before cooking. After cooking little to Loss in little to cooking medium. After cooking Loss in cooking much. medium. grams. grams. grams. per cent. grams. grams. per cent. 114 771 38 33 2 30 1 85 74 Rat her lean . 105 5!) 40 44 30' 76 72 11(1 55 55 50 74 101 45 47 Very fat (special 1 Returned to pan after weighing and cooked much. 83 84 1 87 87 2 Very little cooked. Loss on Fat, Protein, Ash, Water, cooking, per cent. per cent. per cent. per cent. per cent. 61.9 30.3 7.7 . 3 82 64.5 28.1 6.2 0.3 79 71.4 21.8 2.7 1.9 65 75.2 18.5 2.G 69 80.3 12.0 2.1 2.8 58 THE DIET OF NORMAL INDIVIDUALS 249 TABLE 118. COMPOSITION OF VEIIY FAT BACON COOKED (KATHEHINE BLUNT). Much cooked Lightly cooked Eggs in some cities by law must weigh a pound and a half a dozen, an average of GO grams (2 ounces) apiece. Such eggs contain approximately G grams of protein and G grams of fat. The German tables generally figure the fat at 5 or 5-| grams. How gross our caloric reckonings are is obvious if a collection of eggs is weighed and the minimum and maximum weights noted. The weight of the heaviest egg was 72 per cent, more than that of the lightest. (See Table 119.) TABLE 119. VARIATIONS IN WEIGHTS OF ECHJS WITH THE SHELLS. Number eggs weighed. 9 Minimum, grams. 52 Maximum, grams. 63 Variation, per cent. 21 12 .... 40 62 55 11 56 63 12 12 51 69 35 12 . 48 66 38 The weight of egg shells is usually about 7 grams. These examples will serve to show that doctors, nurses and patients should not take their dietetic calculations of calories too seriously, and that carbohydrate, protein, fat and calories should not be recorded in decimals. G. Caloric Values which Every Doctor Should Know by Heart. The quantity of carbohydrate, protein and fat found in an ordinary diet must be known by a physician if he wishes to treat a case of diabetes in modern fashion. If he cannot calculate the diet he will lose the respect of his patient. The value of the different foods in the diet can easily be calculated from the diet Table 120. This is purposely simple, because a diet chart, to be useful, must be easily remembered. With these food values as a basis it is possible to give a rough estimate of the value and composition of almost any food. Various foods are also classified according to the content of carbohydrate (see page 2GO) in f), 10, 15 and 20 per cent, groups, and the lists are so arranged that those first in the group contain the least, those at the end the most. This is a practi- cal and sufficiently accurate arrangement, because except in the most exact experiments the errors in the preparation of the food are too great to warrant closer reckoning. It is practically impossible except when accurate analyses of the diet are made to reckon the 250 DIET IN HEALTH AND IN DIABETES carbohydrate for the twenty-four hours closer than within 5 to 10 grains, and we had best acknowledge that fact. It is really surpris- ing, however, how reliable the figures are if we do not push the matter to extremes. For example, the protein in the diet of one of my patients at the Xew England Deaconess Hospital, as recorded by the nurses on the chart, was calculated to contain 10 grams of nitrogen, and of similar portions of the same food at the Nutrition Laboratory by analyses 17 grams of nitrogen were found. On another occasion the protein was analyzed in 10 portions of cooked lean meat, similar to 10 other portions served the same day at the New England Deaconess Hospital. In these analyses it was found that the protein content was 30 per cent. TAHLE 120. FOOD YAI/TKS IMPORTANT i\ TI:E TREATMENT OK DIAI:ETES. 30 grams (1 <>z.). Carbohydrates, Protein, Pat, Contain approximately grams. grams. gramt Oatmeal, dry weight 20 ,") 2 Cream. 40% 1 1 12 ( 'ream, 20' ', 1 1 G Milk . l.o 1 1 Brazil nuts 2 .") 20 Oysters, six 4 G 1 Meat (uncooked, lean) . (I 3 Meat (cooked, lean) s 5 Bacon 5 15 Egg (one) . . . . (I 6 Vegetables, ii' ( group . 1 ()..') Vegetables, 10' '' ( . group 2 o . 5 Potato . . . . . . (I i Bread IS 3 Butter . .... 25 Fish, cod, haddock (cooked) 6 Broth 0.7 Small orange or \ grapefruit 10 1 gram protein, 1 4 calories. 1 " carbohydrate, 4 " 1 " fat, 9 1 " alcohol, 7 " (i. 2") grains protein contain 1 gram nitrogen. 1 kilogram = 2.2 pounds. .'50 grams (gin.) or cubic centi- meters ((!.('.) = 1 ounce. A patient at rest requires 25 cal- ories per kilogram body weight. ras. Repeatedly physicians have requested me to arrange the above table in terms of household measures. To a considerable extent this is impracticable because the diabetic diet deals with so small a quantity of carbohydrate. An earnest attempt to do this has been made, using the sensible table arranged for general food values by F. W. White,- and tin 1 100-calorie portions so much in vogue and described in the excellent book of Rose; 1 but 'These equivalents arc not exact and must not lie omployr (Sec Table 207, pane 47_'.) - Food Values in llu-.chold Measures. F. II. Thomas Co., 3 Hose: Feeding the Family. The MacMillan Co., 1<>1I>, p. K THE DIET OF DIABETIC INDIVIDUALS 251 without success. The only safe way for diabetic patients at the commencement of their training is to weigh their food. After a few days of weighing, patients can select utensils which conform to the size of the portions of their own special diets and use these exclusively. Fig. 11 shows types of diabetic crockery, suggested by Miss O'Reilly, a visiting nurse to the New England Deaconess Hospital from the Royal Victoria Hospital in Montreal, which my patients often use to advantage. FIG. 11. Diabetic crockery. The ramekin level full with Quaker Oats holds 30 grams. When packed tightly with 5 per cent, vegetables or potato, it holds 90 grams, but when filled in the ordinary manner holds 75 grams. (Ramekin and pitcher for sale by Jones, McDuffee and Stratton Co., Boston, Mass.) In order to illustrate the errors which easily arise from general statements about foods, the accompanying illustration, Fig. 12, is inserted. B. THE DIET OF DIABETIC INDIVIDUALS. 1 . Caloric Needs of the Diabetic. The diet of the diabetic patient should contain, except for brief intervals, the minimum number of calories which the normal individual would require under similar conditions. 1 am convinced that many normal individuals must live upon less than 30 calories per kilo, and repeatedly one sees diabetic patients over fifty years of age who live upon less for long periods of time. Thus Case No. 1094, aged fifty-seven years, whose diabetes began at fifty-five, showed a glycosuria of 7.2 per cent, on July 2(>, but became sugar-free in three days by gradual restriction of carbohydrate to 20 grains and protein to 10 grams. During the following fourteen days she received an average of 718 calories each twenty-four hours and her weight fell from 71 kilos to 70 kilos (157f to 155 pounds). During the following nineteen days it remained unchanged (156 pounds) and the calories per day averaged 1837. 252 DIET IX HEALTH AND 7A T DIABETES FIG. 12. Variations in the sizes of common foods. 1 Three E0 Protein in one ejztc 7 (> 5 Fat in one e^ 7 5 Three Point <>/'.<. Weight of one potato 120 00 00 Carbohydrate in potato 21 IS 12 Three Ornn/jiK. "Weight of one orange lii~>0 225 I.IO Carbohydrate in ouo oningc 20 !."> 10 Thnc (!rti]>t'fruit. Weight of one grapefruit 900 GOO 300 Carbohydrate in one grapefruit .... 40 30 20 1 See footnote, p. 2. r >3. THE DIET OF DIABETIC INDIVIDUALS 253 Case No. 1100, with onset of diabetes at the age of sixty, came to me eight months later. She was sugar-free and evidently living upon a moderate diet. During twelve days at the hospital her weight was constant at 58 kilos and the calories per day averaged 1400. Like other patients of similar age, she felt overfed. If the physician allows his patient to go untreated and excrete sugar and /3-oxybutyric acid in the urine instead of burning these in the body, he must add for each gram of sugar so lost 4 calories (actually 4.1) and for each gram of /3-oxybutyric acid 5 calories TABLE 121. ESTIMATED DIET OF CASE No. 1147 PRIOR TO TREATMENT. Carbohydrate, Protein, Fat, Food grams. grains. grams. grams. Eggs, 12 72 72 5 per o^nt. vegetables, 450 ... 15 8 Milk, 2000 96 64 64 40 per cent . cream, 240 .... 8 8 96 Butter, 90 75 Meat, 120 32 20 Bread, 100 60 10 Totals . . . 179 194 327 449 Total calories 716 776 2943 Total calories 4435 -4- 72 (kilograms) = approximately 60 calories per kilogram body weight. (actually 4.693). I use the figure 5 calories for 1 gram /3-oxy- butyric acid because, along with this acid, a certain amount of acetone and diacetic acid also escape so that the allowance of 5 calories for all three is undoubtedly too low rather than too high. Case No. 344, page 148, shows the importance of the loss of acid bodies as well as of sugar. (See Table 177.) And this is Fig. 12 shows: 1. How readily errors may occur in estimating the food values of the diet unless definite quantities of foodstuffs are prescribed. 2. The absurdity of reckoning food values to the fraction of a gram unless actual analyses of each food as served are made. Errors in eggs may compensate themselves, because the eggs average about 60 grams (and must so average in some communities) ; errors in potatoes, oranges and grapefruit must necessarily be very great. The largest of the three potatoes is actually a small potato; the potato weighing GO grams is about the size of an egg; the oranges from left to right are sold under the trade names of 126, 170 and 250 (to the box) and the grapefruit under the trade names of 28, 64 and 96 (to the box). It is partly on account of the ease with which large errors in the carbohydrate content of food may occur that it is desirable to give to patients with a low carbo- hydrate tolerance their carbohydrate in the form of 5 per cent, vegetables exclusively, for an error in weighing, reaching 120 grams (4 ounces), would amount to but a few grams of carbohydrate. The weights and food values given for the various foods in the illustration are not absolutely but they are approximately correct. 254 DIET IN HEALTH AND IN DIABETES not all, for when acidosis is extreme, Benedict and I have shown that the metabolism is increased about 15 per cent., and consequently more calories are required to meet this demand. It is seldom one can obtain definite knowledge of a patient's diet before treatment is begun, but with Case No. 1 147 the figures appeared reliable, though they show an attempt to lower the carbohydrate in the food. A lady, aged thirty-five years, with diabetes of two and a half years' duration, lost in a little more than this interval 00 pounds. On October 0, 1910, the volume of urine was estimated at 0000 c.c., and the sugar was found to be 5 per cent, or 300 grams the equivalent of a loss of 1200 calories in twenty-four hours. She reported her daily diet to have been as shown in Table 121, on page 253. After a two weeks' stay in the hospital she felt more content with a diet of 1 000 calories than when upon that at entrance, and her weight was constant during the last four days. 1 It is obvious that the saving of food hereby entailed would be considerable. It is the diet of the untreated diabetic which is expensive. 2. Carbohydrate. The total carbohydrate in the diet of diabetic patients is almost invariably restricted, and seldom exceeds 100 grains. This is a decrease to approximately 25 per cent, of the normal carlx)hydrate ration, and so radically changes the composi- tion of the normal diet as to make it self-evident that rapid changes from a normal to a diabetic diet containing even 100 grams carbo- hydrate might easily cause indigestion in normal as well as in diabetic individuals. The character of the carbohydrate has comparatively little effect upon its assimilability. Various methods have been employed to test the assimilation of the different carbohydrates and the evidence is presented in the following paragraphs. (a) The Assimilability of Starch and the Various Sugars. The assimil- ability of sugar can be tested by the oral, subcutaneous, and intravenous methods of administration, and Allen 2 has pointed out in detail the striking differences which result. The oral method is convenient, but inaccurate on account of variation in the rate of absorption. The intravenous method merely gives the saturation limit of the blood and tissues for sugar which can take place without the sugar overflowing the kidneys, and hence is affected by slight variations in renal permeability. The subcutaneous test is the best test of all of the power of the body to utilize sugar. By it errors in absorption are minimized and the liver avoided, and the actual efficiency of the tissues in storing or burning the sugar is determined. The relations of the tests are brought out interestingly by the case of levulose. ''The oral tolerance; of levulose is not much less than that of dextrose, because the liver stops nearly all the levulose. The intravenous tolerance of levulose is approximately the same as that of dextrose, for the. THE DIET OF DIABETIC INDIVIDUALS 255 value represents a more immediate saturation limit. The subcutaneous tolerance of levulose is a very small fraction of that of dextrose, because this method tests the power of the general tissue to utilize levulose, and this power is easily exceeded." In other words, the intravenous method shows the saturation limit, the oral method furnishes a test for the hepatic function, and the subcutaneous method shows the utilization by the tissues, Blumenthal found that the amount of sugar injected into a rabbit to cause glycosuria varied between 1.8 and 2.8 grams per rabbit (about 0.8 grams per kilo) in injections lasting from one to two minutes, and ho records that the saturation limits of dextrose and levulose are almost equal. For galactose it is much less; for saccharose and lactose it is very small. Comessatti found the tolerance to be between 2 and 2.5 grams of dextrose; by having the rabbits run in a treadmill the tolerance was raised about 20 per cent. Kausch, with Bcrendes, has recommended the use of intra- venous injections of dextrose for their stimulating and nutritive value. The solutions used were from 2 to 10 per cent. They gave intravenous injections of 1 or 2 liters, slowly and without ill effect, and glycosuria was absent or insignificant. One woman with puerperal sepsis and temperature above 10(i received up to 2900 c.c. daily, with excretion varying only from 0.2 gram to 3 grams. The rate of injection is an important factor. Allen, 1 from whom I have abstracted most of these data, says: "The rate of utilization is the essential thing to be tested; not the quantity of sugar which may perhaps circulate and recirculate in the blood without overflowing the kidney, but the speed with which the tissues can withdraw the sugar from the blood." Allen believed, writing prior to Woodyatt, Sansum, and Wilder, that the intravenous method may yield results interesting for comparison with those of other methods, but considering that its apparent exactness is probably fallacious, it cannot be accepted without further demonstration as a test of the power of the tissues to utilize sugar. The whole subject of the assimilation of carbohydrates by the oral, intravenous and subcutaneous methods was exhaustively discussed by Allen in 1913, and I abstract freely from his writings. The assimilability of starch is practically without limit in health, though glycosuria may occur after its administration to febrile and alcoholic patients. It is doubtful if there is any difference in the assimilability of various forms of starch. The question will be discussed more fully under oatmeal and potato cures in diabetes. The apparently better assimi- lability of one form of starch over another is very likely to be explained by variation in the rate of absorption. The assimilability of various sugars in health has been carefully studied but with quite divergent results. This is not surprising because little account has been taken of the capacity of the body to store sugar. Woodyatt, Sansum and Wilder 2 recently have urged the importance of time in determining sugar tolerance. Tolerance must be measured and expressed in grams of glucose or other sugar per kilogram of body weight per hour of time. They point out that sugars administered by the stomach take varying lengths of time to be absorbed, according to the motor power of the stomach or intestines, and even if sugars are given subcutaneously or by any other route, which demands absorption as a prelude to their entering the blood, the rates at which they enter the blood will depend ui)on the rates at which they are absorbed. They have devised a method 1 Allen: Loc. cit., p. 54, 2 Woodyatt, Sansum and Wilder: Jour. Am. Med. Assn., 1915, Ixv, p, 2067. 256 DIET IN HEALTH AND IN DIABETES by which solutions of the various sugars can be easily and painlessly injected into animals or man for hours at a time. By the use of these methods, they found that a man can utilize O.S5 gram of glucose per kilogram of body weight per hour for from six to twelve hours without producing any glycosuria, or even diuresis. In other words, a man weighing 70 kilograms when resting quietly in bed may and did receive and utilize GO grams of glucose by vein per hour without glycosuria. This would amount to 252 calories per hour. This apparently is the normal tolerance limit for glucose. In cases of exophthalmic goitre, the intravenous tolerance was found to be as low as 0.(>5 gram per kilogram per hour. For lactose it was nearly zero. If twice the normal tolerance limit for glucose that is, 1.7 grams per kilogram per hour are given, 0.17 gram glucose per kilogram per hour are excreted by the urine, but if glucose is given at a rate faster than 2 grains per kilogram per hour, 50 per cent, of all the glucose injected appears in the urine when constant conditions have been secured. In a normal individual, the greatest rate of glycosuria which can be produced by feeding any quantity of glucose by mouth is limited, and at most approximates that which is produced by an intravenous injection at the rate of 1.8 grams per kilogram per hour. Hence it would be reason- able to believe that when sugar is given by mouth in health, it is never absorbed faster than at this rate. The above investigators also found that when glucose was given at rates exceeding O.So gram per kilogram per hour and the glucose begins to accumulate in the tissues and pass out into the urine, it carries water with it. When the sugar is given at the rate of 5.4 grams per kilogram per hour, marked diuresis results thus, in a dog the urine output rises rapidly to the vicinity of 350 c.c. per hour. If each hour the dog received enough water to make good the loss, this rate of diuresis or a higher one could be main- tained for long periods of time. If enough water is not given to counter- balance the urinary loss, the volume of urine decreases and the dog suffers from thirst and death may ensue. On the other hand, if too much water is given with the hourly injections of glucose of 5.4 grams per kilogram per hour, there is danger of stopping the heart mechanically. In giving intra- venous injections of glucose there arc then two things to avoid: too great dehydration on the one hand, and heart failure from imposing too much mechanical work on the other. These can both be avoided by knowing the number of grams of glucose which enter the body hourly, and what volume of water is removed by such a rate of sugar injection. TABLE 122. THE RELATIVE ASSIMILABILITY OF VARIOUS SUGARS AS DETERMINED BY THE APPEARANCE OF SUGAR IN THE URINE. Intravenous method, Oral administration. Wootlyatt, Sansum and Wilder. Total in Per kilo, per hour, Von Noorden. grains. grains. Dextrose 150- ISO O.So Lovulose 120-150 0.15 Galartose 20-15' 0.10 Lactose 120+ approximately (1) Ulncosc. The tolerance for glucose is apparently unaffected by season or sex, and there is no constant difference for different races. Il is interest- ing that children are supposed to have a far higher tolerance 1 than adults, 1 Hoffmann: Ztschr. f. Kxp. Path. u. Therap., 101-1, xvi, p. :WS; cit. Jour. Am. Med. Assn., 1015, Ixiv, p. 253. THE DIET OF DIABETIC INDIVIDUALS 257 but Allen's experiments with young animals were to the contrary. The increased tolerance in children might be explained by their increased activity. In general, the tolerance for glucose, like that of other carbohy- drates, varies inversely with the rapidity of its absorption. Its assimilation is increased in fever. The occasional appearance of glycosuria in fever is considered by Allen not to invalidate this statement, because that is a toxic glycosuria. Actual increased temperature favors tolerance, possibly by rendering the kidneys less permeable, but chiefly by increasing utilization in the tissues. The influence of muscular work in increasing the tolerance for dextrose has been proven by experiments upon animals and is analogous to the effect of work upon the assimilation of carbohydrates in general. Glycosuria may be produced experimentally by intravenous, subcutaneous, oral and intraperitoneal administration of sugar. It is not produced by rectal injections, this being the one method by which glycosuria is impos- sible under normal conditions. Various methods have been employed to test the presence of diabetes by the administration of glucose, and Bergell describes a method by which a predisposition to diabetes can be detected. (See p. 287.) With this I have no experience. Xaunyn's 1 method is as follows : The patient has a breakfast of 80 to 100 grams of bread and a cup of coffee with milk. Two hours later he is given 100 grams dextrose. If glycosuria occurs to a degree which can be determined quantitatively, the diminished power of assimilation for dextrose is demonstrated. Mere traces in the urine are disregarded. Alimentary glycosuria of this type generally begins, according to von Noorden, 2 within three-quarters to one hour after a large dose of sugar, and lasts three hours. The total excretion is not under 2 per cent., and not over 5 per cent, of the sugar ingested. In Bright's disease the test may be obscured, and in diabetes, complicated by Bright's disease, the effect may not show itself. Table 123 compiled by Allen from the experiments of Tachau is interesting. Tachau tested the blood-sugar fasting, and one hour after 100 grams dextrose were given by mouth. TABLE 123. CHANGES IN BLOOD SUGAR AFTER ADMINISTRATION OF 100 GRAMS OF DEXTROSE. 3 Blood sugar. After 100 Kin. Fasting, dextrose, Glycosuria, Condition. per cent. per cent. per cent. Normal O.OS5-O.OS6 0.060-0.070 Diabetes 0.110-0.169 0.207-0.225 0.3-0.4 Lead poisoning (with glyco- suria) 0.097 0.149 0.1 Four other cases of lead poison- ins 0.105 0. 121-0. 256 Neg. in 3 out of 4 Various fevers O.OS5-0. 110 0.125-0.200 Chronic nephritis .... 0.091-0.104 0.111-0.216 Higgins 1 has studied the time at which the various sugars begin to be burned in the body by the change in the gaseous exchange and especially in the respiratory quotient. He found that glucose and maltose were not utilized as food as soon as the other sugars sucrose, lactose and levulose for the former begin to be burned in twenty to thirty minutes, but the latter 1 Xaunyn: Loc. cit., p. 37. - Von Noorden: Loc. cit., p. 21. 3 Cited from Allen. See Tachau: Deut. Arch. f. klin. Med.. 1911, civ, p. 437. 4 Higgins: Jour. Phys., 1916, xli, p. 25.S. 17 258 DIET IX HEALTH AXD IX DIABETES \vithin ton minutes. He further observed a distinct difference in men between the metabolism of glucose and of levulose and g'alactose. \Ve often erroneously think of an excess of sugar in the blood as being removed by the kidneys. As a matter of fact, this is not the case, for the kidneys remove only a trifle of the excess of the sugar. They by no means regulate the percentage of sugar in the blood. By far the, greater portion of the sugar is removed from the blood by the tissues, as Meltzer and Kleiner 1 have so beautifully demonstrated, and recently Woodyatt, Sansum and Wilder have confirmed. (2) Saccharose. Le ( loff- found that saecharosuria and glycosuria resulted in 100 per cent, of the cases after 100 grams cane-sugar \vere administered orally. Saccharose has lit tie place in the diet of primitive races, and Allen has pointed out a possible connection between the increased use of saccharose and the increased incidence of diabetes. (See p. o().) Tntil the increased incidence of diabetes is thoroughly proven, one must be slow to consider this relation established. (.'->) Lactose. The poor assimilation of lactose is remarkable because it is the one distinctively normal sugar. This lack of assimilation holds in diabetes as well as in health. It is perhaps due to the fact that lactose is broken up into galactose and dextrose before it is absorbed. (The question will be further discussed under the so-called Milk Cure, page 377.) It should be pointed out here that milk contains f> per cent, lactose, whereas there is in cream little more than 2.5 to 3 per cent. Buttermilk and skimmed milk contain essentially the same percentage of sugar as pure milk, but with the fat removed the milk is far less nourishing and thus if given exclusively it represents a type, of low calorie diet. Fermented milks may contain ,'i per cent, sugar. (4) M(dtoxc. It is commonly agreed that maltose is tolerated less well by diabetics than any other form of sugar. (o) Lcruluxc. The advertisement of levulose by a leading drug firm as completely assimilable in both mild and grave forms of diabetes would be amusing were it not that occasionally one sees doctors and patients who believe the statement to be true. The behavior of levulose in depancreatized dogs is quite different from that of glucose. For example, .such dogs can form glycogen from levulose when they cannot form it from dextrose. This is worthy of consideration because; of the possibility of this fact being utilized in some way in treat- ment. For a long time one of the first additions which I have made to the diabetic diet after green vegetables, has been an orange* or grapefruit.. During the last two years Benedict and I have made 1 many tests with lovu- lose iiiion diabetie- patients. These will probably be published during this year. (See 1 p. .'->7>. ) Strauss' introduced le-vulose as a test of hepatic function. Whereas normal individuals and patients with supposedly normal live'rs assimilate 100 grams levulose completely, approximately SO per cent, of patients with hepatic disorders show levulosuria. In the last few years this test, as we'll as others of a similar nature, has lost rather 1 han gained in reputation, as pointed out- by Friedman 1 and Strauss. ((')) (riiluclo.^'. All agree that galactose is poorly assimilated, and yet it is fermented with more difficulty than eitlieT de'xtrost; or levulexse. 'Seep. HI- 2 ('it. by Allen: Coin pi. rend, clc 1'Acad. tics so., 1011, clii, p. 17S5. 3 S'r:iuss: Di'iitsrli. incd. \VcliiiM-hr., I'.Mll, xxvii, p. 7~>7. 4 l-'riedmu!i and Struu~r-: Arch. Int. Mod., I'.M I. xiv. p. f>.'U . THE DIET OF DIABETIC INDIVIDUALS 259 (7) Inulin. Inulin is a polysaccharid of Icvulosc and in this respect comparable with starch, which is a polysaccharid of dextrose. It occurs in the roots of many composites such as the tubers of dahlia and artichoke. No enzymes in the alimentary tract are known to convert it into sugar, but the normal acidity of the gastric juice may effect a partial hydrolysis of inulin to levulose. The bulk of the inulin is decomposed in the gastro- intestinal tract into non-carbohydrate products.' Following the feeding of inulin to phlorizinized animals Lewis and Franker 2 found no increase in glycosuria which one would expect if any of the inulin had been converted to levuloftC. Consequently there appears to be no reason for the employ- ment of inulin in the diabetic diet except to afford bulk. Mary Schwartz 3 found no evidence that the carbohydrates of lichens, alga 1 , and related substances were assimilated by human beings. (b) The Estimation of the Carbohydrate in the Diabetic Diet. The quantity of carbohydrate in the various foods is easily calculated and far more simply than is usually thought. This is not true if one desires scientific accuracy, for in that event analyses of the food given the patient must be made. (c) Carbohydrate in Vegetables. Loss in Cooking. It would appear perplexing to determine the amount of carbohydrate in the various vegetables which the patient eats in twenty-four hours, but this is really not the case. For convenience I have classified the vegetables and fruits which enter into the diabetic diet under four headings those containing approximately 5 per cent., 10 per cent., 1") per cent., and 20 per cent, carbohydrate. (See Table 124.) It is true that there is considerable variation in each group, but the average content is not far from that represented, the error being on the lower side. This does not hold for string beans, for often trouble occurs from the beans having developed into maturity, thus greatly increasing their content in carbohydrate. Many an unexplained trace of sugar in the urine has undoubtedly occurred in this way. A deduction should be made in the percentage of carbohydrate in the vegetables of the 5 and 10 per cent, groups because a part of the carbohydrate is in the form of cellulose, and this is not assimilable. For this reason 3 per cent, and (i per cent, represent more accurately the content of available carbohydrate in the 5 per cent, and 10 per cent, vegetables. This being the case, it is convenient and fairly accurate to consider 30 grams, or 1 ounce of 5 per cent, vegetables, to contain 1 gram carbohydrate and the same 'Editorial: Jour. Am. Mod. Assn., 1914, Ixiii, p. 326. 2 Lewis and Frankel: Jour. Biol. Chem., 1014, xvii, p. 305. 3 Schwartz: Tr. Conn. Acad., Art and Sciences, 1911, xvi, p. 247. 260 DIET IN HEALTH AND IN DIABETES quantity of 10 per cent, vegetables 2 grams of carbohydrate. The vegetables in the 15 per cent, and 20 per cent, groups should be reckoned at their full value. TAIU.K 124. FOODS AHKANCKD AITKOXI.MATKI.V AI\<; TO I'KKCKNTACE OF ( 'AKHOIIYDKATKS. VECJ ETABLES* (Fres h or canned) per cent. 10 per cent. !."> J>:T cent. 2'') per cent. Lettuce Tomatoes Pumpkin ( Ireen jjeas Potatoes ( 'ucumbers Brussels Turnip Artichokes ^helled beans Spinach sprouts Kohl-rabi (Jerusalem) Baked beans Asparagus Water cress Squash Parsnips Green corn Rhubarb Sea kale Beets ( 'aimed lima Boiled rice Endive Okra Carrots beans Boiled Marrow Cauliflower Onions macaroni Sorrel Kgg plant Mushrooms Sauerkraut Cabbage Beet greens Radishes I )andelions Leeks Swiss chard String beans ( Vlery Broccoli I 'HUITS Ripe olives (20 ]>er cent, fat) Oranges Apples Plums ( irapefruit Cranberries Pears Bananas Lemons Strawberries Apricots Prunes Blackberries Blueberries Gooseberries Cherries Peaches Currants Pineapple Raspberries Watermelon ' Huckleberries Butternuts NUTS Brazil nuts Almonds Peanuts Pignolias Black walnuts Walnuts Hickory (English) Pecans Beechnuts -I" ]> grains carbohy- drate, but when boiled and the water changed, only l.S grams. Allen 2 has utilized this method of removing carbohydrate from vege- tables and thus allows patients to have bulk in their diet. He terms vegetables so prepared "thrice-cooked vegetables." " Under these conditions the vegetables may be boiled through three waters, throwing away all the water. Nearly all starch is thus removed. The most severe cases generally take these thrice-cooked vegetables gladly and without glycosuria." Patients often say that it makes little difference to them whether the vegetables are thrice washed or not. It is easy and useful to add a little salt, and if desired the vege- tables can be flavored with meat juices or meat extracts. Even when vegetables are cooked in the ordinary way consider- able carbohydrate, protein, and, what is quite unfortunate, salts are lost. Eew analyses of cooked vegetables are available, but some of those which I have found are recorded in the following table: TABLE 125. THE INFLUENCE OF COOKINC; UPON THE CONTENT OF CARBOHYDRATE IN VEGETABLES. Carbohydrates. Fresh, Cooked, Food. per cent. per cent. Asparagus 3.3 2.2 Spinach 3.2 2.6 Beans (string) 7.4 1.6 Beets ? 9.7 7.4 Carrots 9.2 6.8 Cabbage 5.0 3.7 Greens (beet) 3.2 Onions 9.9 4.9 Beets (boiled) 10.0 Parsnips 13 . 2 Peas 16.9 14.6 Potatoes IS. 4 20.9 Potato chips 46.7 Sweet potatoes . . 27 . 4 42 . 1 Through the kindness of Professor Ruth A. Wardall, of the Department of Home Economics of the State University of Iowa, 1 Von Xoorden: Loc. cit., p. 300. 2 Allen: Bost. Mod. and Surg. Jour., 1915, clxxii, p. 241. 2(Y2 DIET IX HEALTH A.\D IX DIABETES o c THE DIET OF DIABETIC INDIVIDUALS 203 working in Professor Mendel's laboratory in Yale University, I am able to insert Table 120, which shows the carbohydrate in washed vegetables. The results shown in this table are simply preliminary experi- ments, but they are of so much value that they deserve attention. Professor Wardall finds it no disadvantage to use the boiling temperature for each of the extractions. In the data recorded below, the extractions were made by starting the vegetables in cold water and bringing this to the boiling-point and maintaining it at this temperature for one minute. Hot water was added for each of the other extractions, and all were boiled one minute. If the first extrac- tion is kept at 150 F., as has sometimes been recommended, the second extraction leads the list in the power of reduction and in fermentation, the first extraction ranking after the third or fourth. In general, 100 grams of the fresh, clean, dry vegetable, weighed from the edible portion, were used for each analysis, and all calcula- tions were made on this basis. \Yith the exception of canned asparagus, the four or five extractions necessary to remove all reducing substances left the vegetable still attractive in flavor and appearance. For practical purposes these extractions will probably be found sufficient. For the first, place the vegetables in cold water and bring the same to the boiling-point, maintaining the tempera- ture for three to five minutes, but for the others begin with hot water. It is only right to Professor Wardall to again state that the figures cited in the above table represent preliminary analyses. (1) Potatoes. The variation in the percentage of carbohydrate in potatoes before and after cooking is negligible, save with potato chips, in which it more than doubles. The loss of protein is slight, but if soaked in cold water before boiling the loss of protein is 25 per cent, and of mineral matter 8X per cent. If the potatoes are not soaked, but dropped at once into boiling water the loss is much decreased and if the potatoes are boiled with the skins on the loss is very slight. Emphasis should be laid upon the comparatively small amount of carbohydrate in potato in comparison with its bulk and in comparison with the percentage of carbohydrate in bread. A considerable number of my milder cases of diabetes, by giving up bread and bread preparations entirely, have been able to eat potatoes freely. In prescribing potatoes for diabetic patients it is desirable to designate baked potatoes, for these can be eaten with the skins if pains are taken to have them carefully cleaned with a scrubbing brush in the kitchen. This is advantageous in two ways: the skins are quite an addition to the meager diet of the diabetic, and, furthermore, they counteract constipation. 2(14 DIET IX HEALTH AND IN DIABETES (2) Nuff}. Xuts containing 15 and 20 per cent, carbohydrates are probably far less objectionable than most other foods with a similar carbohydrate content. This is due to the fact that in such nuts as almonds and peanuts a larger part of the carbohydrate is in the form of pentosan, galactan, or other hemicellnloses which probably do not readily form glucose. I do not know how much of these carbohydrates is utili/ed by the individual. This would furnish an interesting problem for investigation. (3) Fruit. Fruit is most desirable for a diabetic patient if his tol- erance will allow him to take it. The taste is agreeable, it serves instead of a dessert, and so relieves the patient of the embarrass- ment of sitting idly at the table when others are eating. The best varieties of fruit for diabetic patients are grapefruit (5 per cent.), strawberries (7.4 per cent.), and oranges (11. (> per cent.). These fruits are safer than apples (14.2 per cent.) for the patient, because they contain 5 to 10 per cent, less carbohydrate and are more' satisfying. Furthermore, it is less easy thoughtlessly to eat an orange than an apple and thus break dietetic restrictions. This practical point is worthy of consideration in forbidding apples. Von Xoorden also records unfavorable results from the use of apples. Unfortunately, steaming an apple removes little of the carbohydrate which it contains. TAHI.I; 127. WEKIIITS OF (iK.u'KKitriT, ORANCKS AND BANANAS COMI-ARKD \VITH EDIHLK PORTIONS AS DKTKHMINICD MY A XTRSE. Grapefruit. Whole. Kinble portion. VVhol. Orange. . KdiWc p. >rtion, Banana Whole. K.lihlcp >rtion grains, grains, '/o grains Krams. . grams, jjnims. '.[ Small 347 115 42 Medium 200 118 40 149 92 02 Large (Ml 320 47 Law 27") 1 Hi 42 149 91 61 Florida . . . Small Medium 2f>2 2,>i 113 101 4.") 40 1 2 1 74 12o 79 61 63 Large 3os 111 40 130 M 02 Small . 200 108 42 119 74 62 California Medium 27s 121 44 100 91) 00 Largo . 2S2 129 4(1 119 73 01 14.") S9 61 Yverage 44 107 101 C>1 15s i)7 61 153 91 61 Average 111 S7 111 C1) Oranf/fft. Tho quantity of carbohydrate in a small orange is not far from 10 grains. My children counted for me the number of compartments in forty-eight, oranges, and found these to vary between nine and eleven in forty-three instances, but in three there were twelve and in two there were thirteen compartments. Con- sequently, one \\ill not l>e far wrong to consider that one compart- ment of a small orange contains 1 gram carbohydrate. The same THE DIET OF DIABETIC INDIVIDUALS 265 statement will apply to one-half of a small-sized grapefruit. It is interesting to note how constantly inconsistent the variations are in the amount of edible portions in small, medium, and large oranges as prepared for patients by a nurse. Grapefruit vary more, but it is not difficult to select one of moderate size or to take less of a large one. Table 127 shows these variations. Gross appearance and the weight of the edible portion do not correspond. The observations recorded in Table 127 prompted me to pursue the matter further, for it was evident that inaccuracies must creep into the dietetic calculations in this way. In Tables 128 and 129 are shown analyses made for me by Edward M. Frankel, Ph.D., on aqueous extracts of the fruit after cautions were taken to remove the bulk of the protein and cellulose with lead acetate. The results, therefore, represent the amounts of water-soluble carbohydrate calculated as glucose after hydrolysis. No account is taken of the insoluble material on the assumption that such carbohydrate not soluble in water is of the hemicellulose type, which has been shown not to be utilized by the human body. I am especially glad to have secured Dr. Frankel 's analyses, for I am unaware that similar facts have been recorded. Frankel has added data on the acidity of the fruit calculated as citric acid, because it has been pointed out by Greenwald 1 that this substance is converted quantitatively by the diabetic organism into glucose. TABLE 128. ANALYSER OF ORANC.ES. Made by Edward M. Frankel, Ph.D., New Haven, Conn. Sold as Size. Gross weight, grams. Edible portion, grams. % Grams of sugar as glucose. Percentage of edible portion. Total acidity as citric acid, grains. Percentage of edible portion. Florida Small 188 152 81 12.0 7.9 1.5 1.0 " Small 109 143 85 12.3 8.6 " Medium 214 172 80 12.4 7.2 1.2 0.7 " Medium 215 179 83 15.7 8.8 " Large 290 222 80 19.6 8.4 1.3 0.6 " Large 357 282 79 19.7 7.0 " Large 310 243 7S 20.4 8.4 California Small 189 149 79 10.9 7.3 Small 178 149 84 10.9 7.3 Medium 250 188 75 17.5 9.3 2.2 1.2 Medium 230 170 72 16.2 9.5 1.0 0.9 Medium 270 203 75 17.2 8.5 Medium 204 193 73 15.4 8.0 Large 287 220 77 16.5 7.5 1.4 0.6 Large 322 219 68 19.3 8.8 1.3 0.6 1 Greenwald: Jour. Biol. Chem., 1914, xvii, p. 115. 2(10 DIET IX HEALTH AXD 7.V DIABETES TABI.K 129. AXAI.YSKS OF GRAPEFRUIT. Made by Kdwanl M. Frankel, Ph.D., Now Haven, Conn. Sold as Size. Gross weight. grains. Kdiblo portion, grums. r ; drums of Percentage . ' Percentage sugar as c,f edible '"'laity jis ()f ,, (lil) ],. glucose, portion. curie acid. portion. grams. Portn Kirn Small 401 301 75 24 S.O 2.1 O.S Small 42S 31 S 74 20 7 S.4 2.3 0.7 California Medium 581 419 77 31 7.1 0.5 1.4 Medium 550 3!)4 72 23 S 0.0 0.4 1.0 Larjie 070 515 70 34 s O.s 5.0 1.1 " Large 77:i 000 7s 40 .0 7.7 0.7 1.1 Florida Small 53S 40!) 70 25 .0 0.1 1.2 1.0 " Medium 712 53!) 7(1 34 .2 0.3 5.0 (I.!) " Medium 721 5S5 SI 44 7.7 0.2 1.1 " Large 1 834 024 75 40 .3 0.5 5.4 ().!) The usefulness of the above analyses has been very great, and therefore at my request Dr. Frankel has made similar analyses of lemons. In them the percentage of sugar is strikingly low, but it should be remembered that citric acid can turn to sugar in the body. Recently, I have given the sixteenth of a lemon to patients for their spinach and salads instead of an indefinite quantity of vinegar. One patient who persisted in showing sugar used 70 c.c. of vinegar at a meal. TAHI.K 130. ANALYSES OK LKMO.NS. Sold as Gross Edible Grams < f Percentage Total Percentage weight. porti on, sugar as of edible of edible grams. grams. f ; glucose. portion. citric acid, grains. portion. 92 50 54 0.570 1 . 15 3. 10 0.20 !)() 42 17 0.51s 1.20 > ,s;.j 0. 75 87 GO 09 0.459 0.77 3 . S2 0.35 California California Messina . (5) Jidnanax. Bananas can seldom be taken by diabetic patients because the content of carbohydrate is so high, being equivalent to that in potato. In general the riper a banana, and for that matter any vegetable or fruit, the more the starch in it has changed to sugar, and also the more carbohydrate it contains. Siiue 1111- ripened fruits with their lower carbohydrate content can be made palatable by cooking, a way is afforded for diabetic patients to use them. (0) Ripe (Hires. Ripe olives make a pleasing change in the diet. They contain 4.0 per cent, carbohydrate in contrast to green olives, which contain 1 .s JUT cent. Furthermore, ripe olives are more easily digested. Five ripe or ten green olives contain 1 gram carbo- 1 Greener than the others. THE DIET OF DIABETIC INDIVIDUALS 267 hydrate and a ripe olive contains a gram and a green olive half a gram of fat. The quantity of protein in ten olives is about 1 gram. (7) Milk. The carbohydrate in milk is in the form of lactose and can be reckoned at 5 per cent, or 1.5 grams per 30 c.c., or 1 ounce. It is the same in skimmed milk and whey; but cream and koumiss contain about 3 per cent., or 1 gram carbohydrate to the ounce. Buttermilk contains essentially the same quantity of carbohydrate and protein as milk, but only a trifling amount of fat. I cannot understand why doctors so frequently give it to their patients. (8) Oatmeal. Oatmeal is two-thirds carbohydrate. In calculations one should always be guided by the dry weight, because the different preparations vary greatly in bulk and weight when cooked. (See Table 105, p. 375.) The subject of the oatmeal treatment is dis- cussed at length on page 367. (9) Bread. The carbohydrate in white wheat bread amounts to about 53 per cent. If the bread is toasted, enough water is lost to raise the percentage of carbohydrate in the toast to about 60 per cent. If the bread is made without sugar and with water instead of milk the carbohydrate content is lowered and may amount to only 45 per cent. Coarse breads if made without sweetening or milk would contain slightly less carbohydrate. It is undesirable to give bread to diabetic patients unless their tolerance is very high, because they can take so little without causing glycosuria that the bread is simply an aggravation. An error in weight of 1 ounce of a 5 per cent, vegetable amounts to 1 gram carbohydrate, of potato to 6 grams, but of bread to IS grams. Crackers and zweiback contain still less water than toast, and in consequence the percentage of carbohydrate is raised to the neighborhood of 70 per cent. Diabetic breads are discussed on page 505. 3. Protein. The quantity of protein required by diabetic patients varies with the age, weight, and activity of the case as well as with condition of the kidneys. It is a safe rule at the beginning of treat- ment to attempt to increase the protein gradually up to the same quantity as that required by a normal individual. Chittenden 1 points out that 60 grams (one-half the old protein standard) protein is quite sufficient to meet all the real physiological needs of the body under ordinary conditions of life and with most individuals not leading an active out-of-door life, even smaller amounts will suffice. Chittenden, weighing 57 kilograms, and Mendel, weighing 70 kilograms, lived respectively on 34 and 41 grams protein daily, the former for nine and the latter for seven months. However, as suggested by Allen in his recent Harvey Lecture, the protein must not be too low, because it is probable that Chittenden: Physiological Economy in Nutrition, Frederick A. Stokes Company, New York, 1904. 2()S DIET IN HEALTH AND L\ DIABETES upon a low protein ration fat is not as well tolerated, particularly if carbohydrate be restricted. Intil the Chittenden low protein diet is proved to be entirely satisfactory for healthy individuals over a long period of years it is best not to have recourse to it for long periods in the treatment of diabetes. Temporarily small quantities may be given, but safety lies not far from 1 gram protein to each kilogram body weight. A quantity of protein above the usual amount, on the other hand, is undesirable for two reasons: (1) It increases the metabolism of the diabetic patient just as in the normal individual. This is unfortunate, for the diabetic metabolism is sufficiently stimulated without the need of further stimulation by protein. (2) Furthermore, in the severer cases of diabetes it increases the quantity of sugar in the urine. Reference has already been made to the excessive quantities of protein ingested by diabetic patients when living upon an unprescribed diet (see page 253), and will be made to the similarly large quantities of protein metaboli/ed both by patients in the course of acid intoxication (see page 35(>) and rarely during the course of fasting when inanition is extreme (Case No. 1190, p. 379). It has been claimed that vegetable proteins give rise to less carbohydrate in the diabetic organism than do animal proteins. As a matter of fact, carbohydrate may be formed out of any protein. Janney 1 has studied the formation of glucose from protein in diabetes by feeding pure proteins to fasting, phlorizinized dogs. lie found that no difference existed in the sugar-producing capacity between animal and vegetable proteins because of their respective origins, but that the glucose yielded in metabolism could be demon- strated to vary directly with the amount of glucogenetic amino- acids contained in each individual protein. Thus the wheat protein gliadin which contains 43.7 per cent, of the highly glucogenetic glutamic acid yields 80 per cent, of glucose in metabolism. Indeed, gliadin yields the largest amount of glucose of all proteins hitherto examined and casein and ovalbumin the least. The quantity of glucose yielded by various proteins is shown in the table: TABLE l.'il. GMTOSK YIKLDS OF INVESTED PROTEINS (.JANNEY). f'uHcin, Ovalbumin, Serum, (ielntin, Kibrin, pern-lit. percent. per c-cnt. percent. percent. percent. percent. percent. is 54 r,r> r>"> r>:5 r>r> so r>:5 Janney's work has clearly demonstrated that, the amount of sugar from protein is dependent upon the glucose-yielding amino-acids of protein and not upon the carbohydrate content of protein, which is negligible. His experiments showed that no great variation 1 Janney: Arch. Int. Met!., 1916, xviii, p. 584. THE DIET OF DIABETIC INDIVIDUALS 269 existed in the amount of glucose produced from muscle obtained from various species of animal, including man, and that the per- centage obtainable was 58, corresponding to a glucose-nitrogen ratio of 3.4 to 1. (See page 274.) It is evident that closer attention should be paid to this glucose formation from protein, and that the carbohydrate derived from protein should be added to the carbohydrate given as such in the diet in estimating the diabetic's power to burn carbohydrate. It is suggestive that the protein of white bread yields more glucose than that of other forms of protein, and more also than that of the protein of corn or rye. Thus "the total percentile sugar yield of white wheat bread is 01 , corn bread 50, and rye 44 per cent." These data are also interesting when applied to diabetic breads and also to patent diabetic foods, for although these may contain a small quantity of carbohydrate, yet the protein in them is high and capable of furnishing a large amount of glucose. The result is that such commercial products may thus actually produce as much glucose in a diabetic as does ordinary bread. However, one must not be wholly governed by analytical conditions, as Janney has pointed out, for although vegetable and animal protein yield glucose according to their content of amino-acids, it is quite possible that the vegetable proteins will be less well digested owing to the form in which they are eaten, and thus less protein is assimilated, and in consequence less glucose formed. " It seems, in view of these new results, scarcely a wise procedure to strive to eliminate every gram of carbohydrate from the diabetic's diet, inasmuch as nearly two-thirds of the proteins fed are found to be converted into glucose in course of metabolism by the diabetic." The inaccuracy of the usual methods of analysis of protein has been shown by Janney. 1 Whether the protein has been determined by extraction or difference, large quantities, even 13 per cent, of non-protein substances have been reckoned as protein. Further- more, the carbohydrate content of muscle, although small, has been reckoned as protein, and it is not certain that the extraction of fat has been complete. By new methods of analysis he has determined that the protein in the muscles of chicken, fish (halibut), ox, rabbit, cat, dog and man varies between 16.3 and 18.8 per cent. Thus, the amount of protein calculated by the older methods is seen to exceed that determined by the new analyses from 15 to 20 per cent., in all cases. The calculation of values for proteins in meat should be revised. Quite likely I have estimated too high the protein in meat and the error has been compensated by my dis- regard of the protein in broths and gelatin. 1 Janney: Jour. Biol. Chern., 1916, xxv, p. 185. 270 DIET IN HEALTH AND IX DIABETES (a) Meat and Fish. The chemical composition of meat and fish is simplified by the fact that except in liver and shell-fish, carbo- hydrate is absent. Even in liver the quantity of carbohydrate is almost negligible when \ve consider the amount and frequency with which this article of food is eaten. The analyses of liver and shell- fish will be found in the tables. The chief difficulty in computations of the nutritive value of meat and fish is due to the varying content of fat. Thus, the edible portion of chicken may contain on the average only 2.5 per cent, of fat, whereas lean ham may contain 14 per cent, of fat, fat ham as much as 50 per cent., and smoked bacon 05 per cent., though lean smoked bacon 42 per cent. It is obvious, therefore, that with- out accurate analyses of the fat in meat, only an approximate idea can be had of its caloric value, and as such analyses can seldom be performed this seriously invalidates all our reckoning of the day's diet. In general, a mixture of cooked lean meats probably contains not far from 10 to 15 per cent, of fat. Fish differs from meat chiefly in the small quantity of fat. Fven salmon, which contains more fat than most other fish, showed in its analysis only 12.S per cent, fat, shad 9.5 per cent., and herring and mackerel 7.1 per cent. In general, other kinds of fish show per cent, or less of fat. Halibut steak, for example, contains 5.2 per cent., and cod 0.4 per cent. Preserved fish, however, is quite rich in fat: thus sardines contain 19.7 per cent. In sub- stituting fish for meat, my patients are taught to add from one- half to one teaspoonful of olive oil to the diet for each )>0 grams of fish. The quantity of protein in meat also varies considerably and usually falls as the percentage of fat rises. A value of 20 per cent, for protein in uncooked lean meat represents about the average, and this is increased to 25 per cent, or more when the meat is cooked. The quantity of protein in fish is very slightly less than that in meat. Fish is especially desirable in the early days of protein feeding following the preliminary carbohydrate-feeding days, because in fish the quantity of fat is so low. Shell-fish make agree- able additions to the diet: (1) they are desirable because they are palatable; (2) they are bulky foods and so are satisfying; (o) they furnish a separate course at a meal. Half a do/en oysters or clams are quite sufficient. The edible portion of a medium-sized oyster on the shell weighs on the average half an ounce, and half a do/en oysters would amount to 90 to 100 grams. The six would contain about (') grains protein, 1 gram fat, and 4 grains carbohydrate- tile equivalent of 50 calories. Half a do/en clams on the shell (edible portion) weigh i>5 grams and contain 0.7 gram carbohydrate, '! grams protein, and a negligible quantity of fat. THE DIET OF DIABETIC INDIVIDUALS 271 (b) Broths. Broths are so extensively used upon fasting days in the treatment of diabetes that their composition deserves notice. In the Composition of American Food Materials, Bull. 28, T. S. Dept. of Agriculture, the average of three analyses of bouillon shows it to contain protein 2.2 per cent., fat 0.1 per cent., carbohydrate 0.2 per cent. This is based on the supposition that all the nitrogen is present in the form of protein, which all understand is not actually the case. It is apparent that patients taking one quart of bouillon or broth in a day must get considerable nitrogenous material. Frequently bouillon cubes 1 are used by patients. These consist chiefly of common salt; the amount of meat extract present ranges from 8 ]XT cent, to 28 ]>er cent.; and the third important ingredient is plant or vegetable extract, which constitutes from 3 to 30 per cent. In a research with Iliche upon the metabolism of ammo-acids, Lusk 2 had occasion to investigate the nutritive value of Liebig's Extract of Beef. As a result of his metabolism experiments upon fasting dogs in the calorimeter, he reached the conclusion that "Liebig's Extract is without influence upon the metabolism in spite of the glandular activity it is known to induce." Reference is made on page 284 to the large quantity of salt in the broth prepared at the Xew England Deaconess Hospital, and the full analysis of the broth is there given. Believing the matter of still greater importance, I enlisted the help of Professor Mendel, and through him analyses of various broths in use in small and large hospitals in Boston have been made by Mr. A. II. Smith, of Xew Haven, Conn. These are given in Table 132. The table shows that the quantity of salt in the broths is quite considerable and that it varies markedly in the different broths. One quart of broth at one hospital, for example, would contain 20 grains of salt, while at another less than 1 gram. It is satisfactory to record that all broths agree in being free from carbohydrate. It becomes evident, therefore, why edema is so apt to occur during the course of modern diabetic treatment. The quantity of salt in broths should certainly not exceed ().") ]>er cent. The percentage of fat in the broth as shown by the ether extract is almost invariably slight. The broth ( '-1 , in which the percentage was so considerable, was a sample sent from the hospital kitchen to the diet kitchen; but it did not represent the actual form in which the broth was served the patients. The variation of the total nitrogen in the broths is marked and is the most important feature brought out by these analyses. Here again it should be remembered that the higher analyses represent broths submitted for examination to Mr. Smith, and that if the 272 DIET IN HEALTH AND IN DIABETES THE DIET OF DIABETIC INDIVIDUALS 273 sediment of the broths as well as the fat were removed the values for nitrogen would not only be less but also more uniform. It \\ill be seen that in broth I)-2 over 10 grams of nitrogen were present to the liter, and that the protein nitrogen in this broth amounted to (>1 per cent, of the total. In general, it is a safe statement to make that nearly three-quarters of the total nitrogen in broths is made up of protein- and amino-nitrogen. Such a large quantity of protein demands cognizance and must be allowed for in any dietetic computations. It is notable that the broths with high values for nitrogen were broths from private hospitals and were the broths usually given all the hospital patients rather than to diabetic patients. It is obvious, therefore, that broths for diabetic patients should be prepared in a different manner than the rich broths for other patients who are often overfed with design. In general, therefore, thin, clear meat broths, agreeably seasoned, lightly salted, can be considered desirable for diabetic patients. If the broths are concentrated, form a jelly when cold, are served without complete removal of the fat and the sediment, they are unsuitable and will seriously impair the value of dietetic calculations, and undoubtedly such broths have repeatedly prolonged the periods required to make patients sugar-free. (c) Increased Assimilability of Carbohydrate in Absence of Protein. The effect of a low protein diet upon the assimilation of carbo- hydrate was first strikingly brought out by Klemperer. 1 His experiments were performed with dextrose, and he showed that even dextrose would be assimilated to a considerable degree by a severe diabetic patient provided the protein in the diet was low. He gave !")() grams of dextrose to a diabetic patient while upon diets con- taining varying quantities of protein. Klemperer 's data are shown in the following table: TABLE 133. KXCRETIOX OF DEXTROSE BY A DIABETIC AFTER IXGESTIOX OF lot) CRAMS DEXTROSE WITH VARYING AMOUNTS OF PROTEIX. Excretion of dextrose, grains. ram.-. 10 31 00 37 110 5(5 Clinicians of the older school, beginning with Cantani, appre- ciated the importance of restricting the quantity of protein. The tendency has been, however, to restrict the protein far more than these authorities advocated, and it has not seldom occurred in my own experience that I have allowed my patients too little rather than too much protein. Naunyn, for example, frequently mentions 12o gram., protein (20 grams nitrogen) (.100 grains cooked 1 Klemperor: Die Therap. dor ( Jegenwart , 1911, Hi, p. 447. 18 274 DIET IX HEALTH AND I\ r DIABETES = 025 grains uncooked meat) in the dietaries of his patients, but I am inclined to believe he usually employed a somewhat smaller amount. The deciding factor is the tolerance of the diabetic for protein, and this should be determined just the same as is the toler- ance for carbohydrate. It should furthermore be determined in the presence of carbohydrate and fat as well. Protein is indispensable to a diabetic, and his tolerance for carbohydrate, fat, and alcohol must bo subservient to it. The chief source of carbohydrate in protein is to be found in the amino-acids therein contained. Event- ually a protein may be found in which the amino-acids containing an uneven number of carbon atoms, which are the ones giving rise to sugar, are especially low, but at present the differences in this regard in different proteins are not of great importance, yet Janney's analyses should be borne in mind. See Table 131. Unfortunately, if we avoid the amino-acids with uneven numbers of carbon atoms and give those with an even number of carbon atoms, we are offering our patient good material for the formation of /3-oxy butyric acid. It has been said that the more slowly a given variety of protein is absorbed the less likely it is to give rise to sugar. This quality may be the deciding factor after all. (<7) Dextrose-nitrogen Ratio. The quantity of dextrose which can be formed from the protein molecule has already been discussed on page 130. In dietetic computations I think it safest to consider the maximum quantity of sugar in the urine which can be attributed to the protein in the diet to be ('>() per cent., which is Lusk's dextrose- nitrogen ratio (I) : N) 3.05 grams dextrose : 1 gram nitrogen. If 3.05 grams dextrose are divided by 0.25 grams protein (the equivalent of 1 gram nitrogen) the quotient is 58.4 per cent. For convenience in clinical computations the value 00 per cent, may be employed. Lusk obtained these figures working with phlominized dogs. Janney, working with more recent analytical methods, finds the ratio 3.4 : 1. From the above it. is easily understood how sugar may IK; found in the urine of diabetic patients, although carbohydrates are excluded from the diet. On the other hand, Lusk points out that the quantities of sugar in the urine in excess of the ratio of 3.05 grams dextrose for 1 gram nitrogen, are an indication that the patient is taking carbohydrate and so far all my data support this view. The theoretical maximum of carbohydrate which can be formed from protein is S3 per cent. Lusk also found the same 1) :N ratio in the case of a severe diabetic, and he has termed a 1) :N ratio of 3.05 : 1 the fatal diabetic ratio, but further study of this ratio showed him that it is not necessarily of fatal omen. 1 Mikowski found that 2.05 grams i Lusk: Arch. Int. Mod., 1909, iii, p. 1. THE DIET OF DIABETIC INDIVIDUALS 275 of dextrose appeared in the urines of his depancreatized dogs for each gram of nitrogen, thus giving a D : N ratio of 2. (55 : 1. (e) The Carbohydrate Balance. The carbohydrate balance repre- sents the difference between the total quantity of carbohydrate ingested in the diet and the sugar excreted in the urine during the same period. When the quantity of carbohydrate in the diet is greater than the quantity of sugar in the urine the patient is said to have a positive carbohydrate balance. When the carbohydrate in the diet is less than the quantity of sugar in the urine the carbo- hydrate balance is said to be minus or negative. Under the latter circumstances it is evident that the sugar in the urine is derived either from sugar stored in the body or is being formed out of pro- tein. If it simply represents stored-up sugar, within a few days the negative carbohydrate balance will promptly change to zero and perhaps eventually to a positive balance. Case No. S illustrates this. TABLE 134. THE CHANGE IN THE CARBOHYDRATE BALANCE FROM NEGATIVE TO POSITIVE. CASE No. 8. AGE AT ONSET, SIXTY YEARS. NORMAL EXPECTATION OF LIFE, FOURTEEN YEARS AND TEN MONTHS. THE PATIENT LIVED FOURTEEN YEARS. Diac Date, 1S!)9. grams. grams, intake. balance. pounds. Remarks June 28 . . Gl ? ? 161 Diabetes discovered. 30 July 1 2 1900 Jan. 1 .... 4 45 +41 174 1909 Oct. 12 . . + + 19 54 +35 .. Carbuncle. 17 . . + 70 +70 21 . . 70 +70 140 1911 May 18 . . 42 05 +23 . . Pneumonia. 1912 Sept, 11 . . 10 30 +20 143 1913 April 20 .... 21 ? 140' Ilemiplesia. June 17 . .. 2.49c ? Pneumonia; died. This moderate case of diabetes first came under observation June 28, 1899, and for the first twenty-four hours during which the urine was collected the intake of carbohydrate was not known. Upon the following day no carbohydrate at all was administered, but sugar had existed for so considerable a period in the body that time was necessary for its excretion. The minus carbohydrate balance of (55 grams, therefore, was simply due to retained sugar in the body. This is plainly shown because upon the following day, when 10 grams carbohydrate were allowed, the sugar in the urine 1 March 23. Diacetie acid, grams. Sugar in urine, grams. Carbo- hydrate intake. Carbo- hydrate balance. Weight, pounds. 01 ? ? 161 65 -05 13 10 - 3 10 + 10 270 DIET IX HEALTH AXD IX D1AHETEH decreased to 13 grams, constituting a minus carbohydrate balance of 3 'rams; but a day later upon the same diet the urinary sugar completely disappeared and the carbohydrate balance was plus 10 grams. Six mouths later the tolerance for carbohydrate had risen to 41 grams. It rose somewhat during the subsequent years, persisted during a carbuncle, fell \vitli an attack of pneumonia, then again fell, and in 1!H3 the patient died of a second attack of pneumonia three months after a cerebral hemorrhage in the four- teenth year of the disease. On the other hand, a minus carbohydrate balance which is per- sistent is indicative of severe diabetes with the formation of sugar from protein. The severity of the diabetes in such a case is also shown by the fact that when a minus carbohydrate balance is permanently present, acidosis is extreme. (See Case No. 344, page 410.) It will be seen that the carbohydrate balance was minus on .lime 2S, and .Inly 4, 0, 10 and later remained at quite a uniform figure, although the carbohydrate, intake was varied to a considerable degree, but increased along with the severity of the diabetes. 1 hiring the period represented by these four days, extending considerably over a year and a half, the quantity of nitrogen was not very far from 1C) grams, and with allowance for fecal nitrogen might be taken as IS grams, representing the metabolism of approximately 112 grams protein. From this enough sugar could be formed to account for the minus carbohydrate balance if 00 grams of sugar are derived from 100 grams protein in accordance with Lusk's I) : N ratio of 3.05 : 1. In like the temporary minus balance in Case No. S, it will be seen that the minus carbohydrate balance in Case Xo. 344 was persistent and in consequence the acidosis was extreme. The remarkable change in the carbohydrate balance toward the end of life took place after the patient had developed diffuse tuberculosis and was about to die. Attention is called especially to the disap- pearance of acidosis at thi> time. It is referred to in the discussion of Allen's theories of diabetes, under treatment, and again in the section upon tuberculosis. Finally, Case Xo. 70.") illustrates the favorable change which takes place in the carbohydrate balance with modern treatment and at the same time the variability in the 1 ) : X ratio. (See page 314.) 4. Fat.- Fat in any form is absorbed by the diabetic patient quite as well as by normal individuals, except in those rare cases of pancreatic diabetes. One such case (XV). 070) was seen by me a few days before coma. In this instance diabetes occurred after partial loss of the gland from acute pancreatitis. The case is reported in detail by .Jurist. 1 1 Jurist: Am. Jour. Mod. Sr\, I'.tOO, rxxxviii, p. 180. THE DIET OF DIABETIC INDIVIDUALS 277 Fat, however, is not well absorbed by the dog made diabetic by the removal of the pancreas. This fact solves one of the problems experienced in producing acidosis in dogs. When Allen succeeded in making a dog with severe diabetes gain or even hold his weight by forced feeding of fat, increasing acidosis occurred. 1 The experiments of Ringer, 2 in Lusk's Laboratory, are against the formation of even trifling quantities of sugar from fat, and in this respect correspond with the general consensus of opinion in the literature. In a careful study of three severe cases of diabetes Allen and DuBois 3 reach the same conclusion. Although fat is absorbed quite as well by the diabetic as by the normal individual, on the other hand, it would be incorrect to say that it was as well utilized. This is due to the formation and ex- cretion of /3-oxybutyric acid in severe diabetes. Although all of this acid may not come from fat, the greater portion of it does. Undoubtedly future investigations will show whether this acid comes in great part from fat soon after it is eaten or whether it represents in part the breaking down of body fat. Reference has already been made to the difficulty of determining the quantity of fat in the diabetic diet. Eggs vary in size and may contain 5 grams or 7 grams fat. The fat in meat is most variable. One analysis of a combined sample of 10 portions of mixed lean meat served my diabetic patients showed the percentage of fat to be 15 per cent., and another such analysis showed 10 per cent. Bacon, both cooked and uncooked, shows wide differences in its content of fat. (a) The Increased Assimilability of Carbohydrate in Absence of Fat. Notwithstanding that sugar is not formed from fat, the addition of fat to a diet upon which a severe diabetic patient is sugar-free can easily be followed by the appearance of sugar in the urine, and the converse is equally true. It would seem as if the fat molecules displaced the sugar molecules from their attachments in the body and set them free for excretion. The increase of sugar following the addition of fat to the diet is well shown in Case Xo. 513, Table Xo. 1(10, p. 354, for the period extending between August 12-13 to Aug. 19-20, and in Case Xo. 1120, p. 338. (See also Table 136.) (b) The Value of Fat to the Diabetic. Fat forms the bulk of the diabetic patient's diet. Even with the most modern ideas upon treatment this statement holds. The figures inserted above (page 244) gave the proportions which the different foodstuffs" take in the diet and shows the extent to which diabetic patients must depend upon fat to offset the loss of carbohydrate. Remem- 1 Allen: Am. Jour. Med. Sc., 1917, cliii, p. 313. " Ringer: Jour. Exp. Med., 1910, xii, p. 105. 3 Allen and DuBois: Arch. Int. Med., 1916, xvii, p. 1010. 27S DIET IN HEALTH AND IN DIABETES Ixring that the diet of n healthy individual of 70 kilograms at office work contains approximately 300 grains carbohydrate, yielding 300 X 4) 1200 calories, and if nearly all this quantity is unutilized by the diabetic patient, it, can be calculated ho\v many calories in the form of fat must be given to replace it. Theoretically, these 133 grams -f, 1 "- fat should be taken in addition to the usual 100 grams of fat in the normal ration; but practically this is seldom necessary, because the diabetic patient is usually less active than tin' ordinary individual and the quantity of carbohydrates which we have allowed for the normal individual may l>e in excess. Further- more, most dialn'tic patients have a tolerance for quite a considerable quantity of carbohydrate. Finally, tliest; calculations are made for a patient weighing 70 kilograms. In reality most diabetic patients weigh far less and therefore require less food. It is surprising how readily in the past double and even treble the quantity of fat ingested by normal individuals was borne by the stomach of the diabetic patient. Jt is, however, unwise to push the administration of fat too energetically for fear of causing a dislike for it or even indigestion. I once prevented a diabetic boy from enjoying and deriving benefit from sardines, with their accom- panying oil, by allowing him to eat the first time as many as he liked. lie promptly ate a boxful and the disgust he then acquired for sardines was never overcome. The diabetic must eat fat, and the problem is to make it harmless for him to do so. Frequently, I see patients who have taken large quantities of fat with obvious benefit for long periods. Case No. X (p. 275) must have taken 150 grams of fat daily for fourteen years and died at the age of seventy-three. The quantity of carbohydrate in her diet for the greater part of the time was far below 75 grams. The fat could hardly be said to have done injury in this case, for the patient outlived most of her family. Case No. 93X (p. 4X5) soon became accustomed to fat. Case No. 5(')4, Table 135, age at onset, sixteen, cats 1 70 grams daily too much, I acknowledge. His case is interesting because four years ago he was three and a half months in getting sugar-free, and fora large part of the time showed acidosis. Last year, after four years of diabetes, he took his preliminary examination for Harvard. A brief abstract of his case is given below to show that even formerly patients some- times did surprisingly well. Case No. 5(>4 came under observation November 30, 1012, at the age of sixteen, three weeks after his onset, which occurred without previous symptoms after an important football game. Volume of urine eight quarts during the day. The marked acidosis at that time nearly five years ago- led me to make very gradual changes in the diet and the sugar in the urine decreased from presumably more than 500 grams a day before entrance to the hospital to 230 grains THE DIET OF DIABETIC INDIVIDUALS 279 on the first day after entrance, December 1 7-1 8, 1012. The patient was discharged, with 42 grains of sugar in the urine on February 1.3-14, 1913. I'm lor the close care of his physician and a trained diabetic nurse he became sugar-free IV larch 31, 1913, and has remained sugar-free, with the rarest exceptions, since. TABLE 135. CASE No. 5f>4. THE COURSE OF SEVERE DIABETES IN A BOY OF SIXTEEN. TREATMENT BEGUN IN DECEMBER, 1912. Uri no. Date. Carbo- hydrate in diet, grams. Sodium bicarbo- nate, grams. Naked weight, pounds. Volume. Diaeetie o.o. acid. Total Total XH 3 , sugar grains, (polar.), grams. 1912 Dec. 17-1S 5430 4- 4- + 230 160 16 99 18-19 5100 si. + 5.2 130 160 16 99 19-20 4710 4- + ISO 135 16 100 20-21 4710 + + 190 125 16 99 21-22 4050 + + 4- 4- 115 125 16 100 22-23 3S40 + + 119 115 16 100 23-24 4020 + + + 137 75 16 99 1913 Jan. 1- 2 4140 + + S9 50 20 101 2- 3 4440 + + + 115 50 20 100 3- 4 3420 ' 4- 4- + ... 75 50 20 101 4- 5 3120 + + 37 15 20 103 5- 6 4200 + 143 1 65 20 104 6- 7 2100 + + + 4 13 15 106 9-10 2700 4- + + 4 : 3 . S 49 40 102 Mar. 31 April 1 945 1.1 7- S 1240 si. + 113 1914 Jan. 5 2000 1915 April 26 1SOO 1.0 50 134 (dressed) 1916 Sept. 23 700 0.7 129 (dressed) Oct. 18-19 1600 + 6 Dec. 20-21 SOO 134 (dressed) 1917 Jan. 25 26 1000 + 1.0 45 May 1, 1915, blood sugar, 0.13 per cent.; Sept. 23, 1916, blood sugar, 0.13 per cent. The Eskimos live largely upon fat. Their duration of life can hardly be known with accuracy, and many of the men die as a result of their hazardous seafaring occupations. The duration of life of the Eskimo women should furnish an interesting study. How much fat should a diabetic patient eat ? Mainly, from what has been and will be recorded in the next section, this does not depend upon the capacity of his digestion. The safest answer would be as little as possible above the normal ration of 100 grams. Unquestion- ably the quantity will vary from time to time, and it may increase 2SO 1)1 KT IX HEALTH AXD 7.V DIABETES with years without detriment to the patient. Nevertheless, I am always glad to see a diet with a carbohydrate-fat ratio of 1 : 2 and dread to see one with a ratio of 1 : "> or more. Of late fats have been much restricted in the diet of diabetic patients, and often more than necessary simply because of a trace of acidosis. Repeatedly one observes severe cases of diabetes who at the start show a mild reaction for diacetic acid following the administration of a few grams of fat, and yet later these same patie,nts, with practically no increase in carbohydrate, bear far greater quantities of fat without the slightest evidence of acidosis. Thus Case No. !K>X, cited in detail in Section VI, page 4X5, illus- trates how the addition of an egg to the diet of a two-year-old child two days after being made sugar-free caused the appearance of an acidosis which was alarming. The supposed acute fulminating cases of diabetes may be acute, but I believe their fulminating character to be usually artificial and caused by the injudicious use of fat as in this little patient. Fat is most agreeably taken as cream, and cream which contains 20 per cent, butter fat is usually better borne than a richer cream. It is seldom advisable to allow more than half a pint (240 c.c.) of cream, although patients prefer to increase the quantity of cream at the expense of other forms of fat in the diet. There is no other form of food from which a diabetic patient can derive more pleasure for its caloric value and yet with less harm to himself than from cream. Half a pint of 20 per cent, cream contains approximately .10 grams of fat, and yet the quantity of carbohydrate in cream of this richness is but little over S grams, and may be estimated in clinical work as X grams or 1 gram to the ounce. Occasionally, patients bear butter better than cream, and, as a rule, fresh uusalted butter is preferred. Obviously, when cream is increased in the diet the butter must be decreased, and rice ccrxd. Thirty grams of butter contain 2.~> grains of fat, and this is a welcome addition to the diet. Oleo or butterine contains no sugar and has about the same per- centage of fat as butter and the cost is approximately one-half that of first-class butter. Lard, being nearly 100 per cent, fat can be used to advantage more than it now is in the diabetic'.-, diet. Crisco, also nearly 100 per cent, fat, is often more welcome than lard, because of its lack of flavor. Oil is an ideal diabetic food, because it is a pure fat. Oil is so desirable for a diabetic that I lieMtate to have a patient take more than 15 grains ( 1 table- spoonful) lest he weary of the same. If oil is disliked upon vege- tables it can be taken in small quantities after meals as a medicine. Hodgson's 1 prescription of equal parts of olive oil, castor oil, and 1 Hodgson: Jour. Am. Mi-d. AsriD., 1911, Ivii, p. 11S7. THE DIET OF DIABETIC INDIVIDUALS 2S1 glycerin mixed with one-tenth part gum arable, flavored with oil of wintergreen, not only tends to overcome the constipation for which it is intended but furnishes much nutriment. I can testify to its usefulness. Italian patients naturally bear olive oil unusually well. An Italian diabetic patient under my care at the Boston City Hospital with typhoid fever not only passed through the disease uneventfully upon oatmeal gruel and olive oil, but incidentally became sugar-free and developed no acidosis. Olive oil forms an excellent lunch for diabetic patients. I frequently advise its use upon retiring. It is the diabetic patient's cough medicine; it relieves the symptoms of his hyperacid stomach. Peanut, corn or cotton-seed oil may be substituted if expense is a factor. (c) The Danger of Fat to the Diabetic. Fat is the chief source of the dreaded acidosis, though to this in lesser degree the ami no- acids of the protein molecule with even numbers of carbon atoms contribute as well. Fat, therefore, at one time may save the life of the diabetic, but at another period may destroy it. The close dependence of acidosis upon a fat diet is beautifully shown by the following chart : TABLE 136. l THE DEPENDENCE OF ACIDOSIS UPON THE FAT IN THE DIET. Urine. Diet. Date. Total Vol., Diacetic Nils c.c. aeid. (Folin), Total sugar (polar.). Carbo- hydrate, Protein, grams. Fat, grams. Alcohol. Calories. grams. grains. f, ams. 1912 July 5 1200 + + 1.9 48 20 100 200 2310 6 900 + + 2.1 27 65 100 200 2500 27 2490 + 0.6 30 90 33 74 905 Aug. 8 2660 + + 2.7 80 190 75 200 3000 + Oct. 20 2040 + 0.6 45 64 75 30 12 1010 31 2460 0.3 38 45 75 30 12 949 Nov. 12 2160 . 5 56 50 75 30 12 1913 Jan. 28 4080 h + + 1 2.6 122 35 100 200 2370 Feb. 2 3630 h + + 3.0 152 66 90 200 2150 June 12 3600 + + + + 4.1 108 90 100 200 2640 Julv 27 3630 + + + + 4.4 123 200 150 180 + 3000 31 3930 + + + + 3 . 3 172 200 150 180 + 3000 There is no more potent agency in the prevention of acidosis than the withdrawal of fat from the diet. The absence of acidosis in totally depancreatized animals and in a human case of pancreatic insufficiency like that of Spriggs and Leigh' 2 is to be explained by the non -absorption of the fat given. ' Williams and Dresbach: Am. Jour. Mod. Sc., cliii, 1917, p. 65. 2 Spriggs and Leigh: Jour. Am. Med. Assn., 1915, p. 1952. 282 DIET IN HEALTH AND IN DIABETES If chemists agreed that from each of the higher fatty acids only one molecule of /3-oxy butyric acid could be formed, it would be perfectly evident that the diabetic would be in a little less danger of acid poisoning if he lived upon the fat of meat and fish rather than upon butter fat, which contains a number of the lower fatty acids. 1'nanimity upon this point, however, does not exist. If the vague impressions of clinical experience could be trusted, there would seem to be less danger of acidosis from the fat of meat and fish than from butter. Ordinarily, according to von Xoorden, the butyric acid in butter is not of significance ]>cr .sr unless the quantity of milk fat is in excess of 1 ">() grams. The remarkable power by which the individual can gradually become accustomed to a fat-protein diet is well shown by many diabetic patients. Thus, Case No. o44 took :>72 grams of fat on an oatmeal day September 1 .">-!(), 1010. The acidosis on this day was extreme, as shown by the excretion of 27.0 grams /3-oxybutyric acid. Formerly many diabetic patients took 100 grams fat in the form of cream in addition to that in bacon, butter, eggs and fat meat. Yon Xoorden's oatmeal cure called for 200 to 300 grams butter and one is led to ask if the simultaneous ingestion of the large quantity of carbohydrate did not save the patient's life? Allen has again made us all his debtors by a series of experiments upon diabetic dogs which show the insidious way in which fat is harmful in the manner in which it has been customarily employed in the treatment of diabetes. "Fat unbalanced by adequate quantities of other foods is a poison." 1 f). Alcohol. In no disease is the employment of alcohol more useful or more justifiable, but it cannot be taken with impunity, for neuritis rather easily develops in undernourished patients. Alcohol furnishes an agreeable form of food in a diet which is often disagreeable, and the quantity of nutriment which it contains is by no means negligible. Authorities agree that the patient can take large quantities of fat more readily in conjunction with alcohol. Yet, notwithstanding all the above, few of my cases have taken alcohol and still fewer take it now than in years gone by. This is true even when patients arc undergoing fasting. Most diabetic patients require no alcohol and do not take it, but severe cases are often heljx'd by the addition of l."> to '.}() c.c. to the diet. Thirty cubic centimeters of alcohol are to be found in approximately (10 c.c. of whisky, brandy, rum, or gin, or .'!00 e.c. of most of the sugar-free wines. Thirty cubic centimeters of alcohol are equivalent to (oOX~) 210 calorics. Few of my patients ever take as much alcohol as this, but half the quantity will repliiee about .1 per cent, of the total diet, i Allen: Am. .lour. Mod. Sr., 1917, cliii, p. 313. THE DIET OF DIABETIC INDIVIDUALS 283 and allows the omission of -?,- 12 grams of fat. This does not appear of great moment, but it would amount to 2100 calories in twenty days. Alcohol may he administered in various forms which are free from sugar. These are specified in the diet tables on pages 517 and 51S. The use of alcohol to combat acidosis will be discussed under the treatment of that condition, page 393. The effect of alcohol must be more thoroughly studied with the calorimeter. Allen and DuBois 1 found its ingestion was occasionally followed by respiratory quotients higher than would theoretically be expected. Fasting patients frequently complain of nausea if they are given alcohol, even though it be administered in dilute form. I have been surprised to observe how often they have requested that the alcohol be omitted rather than continued. Then too there is always the question as to the effect of alcohol on the cells of the body when given to a diabetic either with or without food. I am not convinced, looking at it from this point of view, that it is harmless. It has been used by me less and less of late, save with the severest cases, in which I have employed it for its caloric value, and I have at no time prescribed as much as 30 c.c. every three hours. Martin, and Mason, 2 in their series of cases used it with only one patient. One should not prescribe alcohol thoughtlessly for dial>etics. They may acquire the habit, though this is far less common than one would suppose. For this reason 9,"> per cent, alcohol may be given instead of alcoholic beverages. (>. Liquids. It is rarely necessary to restrict the liquids in diabetes. The diminution of the carbohydrate in the diet usually leads to a corresponding diminution in the thirst and quantity of urine. I hesitate to restrict liquids in severe diabetes for fear too little liquid will be available for the body with which to eliminate the acids which may have been formed. On the other hand, patients often upset the digestion by drinking large quantities of liquids rapidly. This is avoided by allowing only half a glass of liquid at a time, though the patient is instructed to take that as frequently as desired. Case Xo. 119(1 continually voided large quantities of urine, but usually I could find a cause such as the ingestion of 20 or more grams of salt, bouillon cubes in variable number or 21 half- grain saccharin tablets a day. Ice-water should be discouraged. 7. Sodium Chloride. Salt is of great service to the diabetic patient. If it is withdrawn from the diet the weight falls, due to excretion of water, and the skin and tissues of the patient become obviously dry. ' Allen and DuBois: Arch. Int. Mod., 1910, xvii, p. 1010. 2 Martin and Mason: Am. Jour. Mod. So., 1917, p. 50. 1)1 KT L\ HEALTH A\D J\ DIABETES In the early days of the fasting treatment patients often lost much weight because water alone was allowed. For example, I learned of one case who lost 1.'! pounds in four days in this manner. AVheii broths are freely given during fasting it is not uncommon, par- ticularly in the presence of acidosis, to see a patient gain weight, and invariably such patients feel better than those who lose. Salt is very freely used by diabetic patients. I do not remember to have ever seen a diabetic patient who took too little salt. One of my fasting cases was accustomed to shake it into his hand to eat. Patients will often salt their broths, although they contain considerable salt. An analysis of the duplicate portion of the broth which Case Xo. 7()5 drank in three days is recorded in Table K>7. TABLE 137. BKOTH TAKKN BY CASK Xo. 7(5") i\ THREE DAYS AT THE XKW K.\ c.c. XitrogiMi Fat . . . 2.:>7 0.6 0.21 0.05 5.2 0.5 The same patient used 20 grams of salt from the salt-cellar during the week which included two fasting days and five other days upon which she took )>()() grams washed vegetables and from !.")() to 1225 grams fish or lean meat. To avoid such excessive use of salt, nurses arc advised to serve no salt with broths. Case Xo. 9X2, a young man with onset of diabetes at the age of twenty, excreted 40. o grams of sodium chloride on February 1">, 19 1C), and a few days later the aliquoted urine of two days contained XO.C) grams. I p*.n inquiry 1 learned from his nurse; that in addition to the ordinary amount of salt in the food as it was prepared the young man filled a salt-cellar each morning and emptied it before night. TABLK loS. KKCKSSIVI; INTAK.I: OK SODITM ( 'HLOIUDK BY A DIABETIC- HOY. Vol., Nad. Dial-otic Nitrogen. Pillar, Protein.' Fat, ISO 107 M:{ 21 ( .) 1)7 111 220X 111) <)"> 111 21">2 THE DIET OF DIABETIC INDIVIDUALS 285 An oatmeal cure is accompanied by the vise of much salt. In the preparation of the standard 240 grams dry oatmeal a cook would employ about 10 grams salt. This may be of importance in the explanation of the edema which often accompanies the oatmeal cure. \ egetable foods are rich in potassium and, as any farmer's child who salts the cattle Sunday mornings knows, deficient in sodium. It is not strange, with our free use of vegetables, that diabetic patients, like cattle, crave salt. But there is still another reason, because when an excess of potassium is eaten, it is quickly dis- charged and along with the potassium goes sodium as well. A meat diet requires little salt. Attention may be called to the low excretion of salt in coma. In a recent case (Xo. 1053) two days before death the quantity of salt was 1.28 grams, and in the twenty-four hours preceding death amounted to but 0.44 gram. The part played by sodium bicarbonate in influencing the weight indirectly through its action upon salt is discussed under Treat- ment of Acid Intoxication. The salt metabolism of diabetic patients requires careful study. This should be undertaken in connection with a study of the sugar, fat, and salt in the blood and the weight of the patient, in addition to the salt, sugar, and nitrogen in the urine. 8. Lime Deficiency in Diabetes. Other anomalies aside from those associated with sodium chloride very likely occur in diabetes. Thus Max Kahn and Morris IT. Kahn 1 placed five diabetic patients of mild character upon Folin's standard diet for nine days and found that they constantly lost calcium. The administration of calcium intravenously into diabetic patients has been reported to cause a marked fall in the excretion of glucose and a gradual decline in the sugar in the blood. The injection of calcium during the height of anesthesia with magnesium quickly abolishes this state and in correspondence with this antagonism there is a rapid return of blood-sugar content to the normal. 2 I nderhill 3 also found that intravenous injection of sodium car- bonate into rabbits might induce a marked though transient fall in the blood sugar. After total pancreatectomy in dogs, Lebensohn 4 observed no change in the absorption of chlorides and no relation to the XI I 3 excretion. 1 Kahn and Kahn: Aroh. Int. Mod., 191G, xviii, p. 212. 2 Undcrhill: Jour. Biol. Them., July, 1916, xxv, p. 471. sTJnderhill: Ibid., p. 463. 4 Lebensohn: Jour. Biol. Chem., 1915, xxiii, Xo. 2, p. 513. SECTION V. TREATMENT. A. PROPHYLAXIS. THE prophylactic and etiological treatment of diabetes will surely play an important role in the future, and it is already plain that progress will be along two lines: (1) toward the early detection of the disease, and (2) toward the prevention of the development of the disease in those susceptible to it. The whole trend of Naunyn's teaching favors the energetic treatment of the slightest evidence of diabetes. Only by following this plan are regrets avoided. The importance of the treatment of incipient pulmonary tuberculosis is not greater than that of the treatment of incipient diabetes. The tendency to diabetes frequently remains latent for a long time, and this is well illustrated by the appearance of diabetes in the children before it occurs in the parent. In many instances the disease breaks out temporarily before the glycosuria becomes permanent. Case No. 129 showed sugar in the urine in 1 ( ,)01 at the age of three years "at a time when she appeared out of condition. Examining frequently after that I failed to find it and did not look for it again until in February, 1!)05, when she appeared like a full-fledged case of diabetes." Death occurred in coma in July, 1007. ( 'ase No. 2o") showed 1 .'.} per cent, of sugar in the urine January '>, 11)01, at the age of twenty-six years, one month after an attack of severe catarrhal jaundice. The glycosuria disappeared at once on restriction of diet and did not return after resumption of a liberal diet containing sugar. December, 1<)01, right pyelonephritis, urine sugar-free; .January o, 11)0."), sugar appeared with a moderate amount of acetone, but no diacetic acid, and the patient became sugar-free with a strict diet and, until July, !!)()('), was able to eat freely of toast, oatmeal, potato, rice or oranges without glycosuria. Death in coma May 1, HMO. '\ he patient was under the care of Theodore Janeway save for a few days in 1 ( )0 ( ,), when he underwent several metabolism experiments at the Nutrition Laboratory. Case No. 100S showed sugar in the urine on repeated occasions (280) PROPII Y LAX IS 287 at the age of forty years, was carefully treated for sixty days and later no sugar was found, but it reappeared when the patient was fifty-one, and he came under my observation three years later, with severe diabetes. 1. Early Diagnosis, The only way in which an early diagnosis of diabetes \vill ever be made is to search for it. It is possible that tests of the blood may soon give us still earlier hints of the presence of the disease, but at present the only practical method is to make frequent examinations of the twenty-four-hour quantity of urine. Bergell 1 has described a test to determine a predisposition to diabetes. The urine is diluted to a specific gravity of exactly 1012, but if it is more dilute, it is not concentrated; then 7 c.c. of a 15 per cent, solution of sodium hydroxide are shaken up once with 20 c.c. of urine. Three cubic centimeters of a solution of copper sulphate (138.78 grams CuSO 4 to 1000 c.c. water) are added and the whole vigorously shaken for ten seconds. It is then filtered through a coarse filter; the first 2 c.c. are thrown away and the fluid which filters through in the next six minutes is collected in test-tubes of uniform size. In the healthy the fluid is only very little discolored, but in the diabetic and in those who develop diabetes later there is a pronounced blue tint. He reports the reaction present in 80 per cent, of the children of diabetics. Randnitz 2 failed to confirm these results. The results of life insurance examinations (see page 1(>7) show how useful such examinations are, and it is a hopeful sign that the insurance companies are offering to examine gratis at frequent intervals the urines of their policy holders. The tendency toward more and more frequent examinations of the urine goes steadily forward. It is a good rule to advise everyone to have the urine examined upon his or her birthday, and no physician should see a patient without examining the urine at least every six months. But the responsibility for urinary examinations rests not alone upon the family physician. It is the duty of all specialists either to examine the urines of their patients or to assure themselves that such examinations have been recently made, and the time is not far distant when this course will be adopted by progressive dentists as well. One of the greatest aids to this policy has been the establish- ment of laboratories where patients and physicians can secure examinations of urine at trifling cost. Such institutions have come to stay. 2. Heredity. The hereditary and familial tendency of diabetes has already been discussed. Fortunately, many of these cases are mild. Recognizing the hereditary and familial tendency, relatives of a diabetic patient should always be kept under supervision, and 1 Bergell: Deutschc mod. Wclmschr., 1914, xl, p. 2094. 2 Randnitz: Wiener klin. Wchnschr., 1910, xxix, p. 1023. 2S8 TREATMENT urinary examinations should be- made at more frequent intervals than in other individuals. Particularly should the urines of sueh individuals be carefully examined when conditions arise which would favor the development of diabetes. One should also endeavor to prevent the outbreak of diabetes by instructing such individuals first to control the total quantity of food eaten by their body weight, and second to take carbohydrate almost exclusively in the form of starch rather than sugar, and never to indulge in unusual quantities of carbohydrate, such as candy, maple sugar, or sweet fruits. Third, it would be a great mistake to consider the diet alone of importance. Mental relaxation and physical exercise should be- promoted. Jf we are to bring about a decrease of diabetes in the community it will be with measures such as these. Every agency which promotes health and physical development tends to prevent an outbreak of flic diabetic tendency. "It is easier to keep well than to get well." (Greeley.) It would be most unfortunate for two individuals each hereditarily burdened with diabetes to wed, though by care in environmental conditions the consequents of heredity might be averted and the tendency decrease in the third generation. 1 3. Obesity. No preexistent abnormal condition has occurred more frequently among my diabetic patients than has obesity. (See Table 3!).) Obesity ail'ords a splendid opportunity for the physician in which to practice preventive medicine. The individual who allows himself or herself to become fat may easily develop diabetes. It is desirable to spread the information that those live longest who above the age of thirty-five years are ."> to 10 per cent, below the average normal weight and that so little an excess of weight as "> per cent, is fraught with danger. Patients should be cautioned against gaining weight at any period, but particularly after infectious diseases, pregnancy, the climacteric, and following changes from an active to a sedentary mode of life. Since obese patients are prone to diabetes, they should have frequent examina- tions of the urine made. Although emphasis is usually laid upon the appearance of sugar in the urine with a patient losing weight, it cannot be too strongly emphasized that it is a com- mon occurrence for sugar to appear in the urine when a patient is gaining weight. 4. Infectious Diseases. - The development of diabetes following infections diseases i> a striking phenomenon. Far less instances have occurred in my own experience than 1 had anticipated before a careful search of my own cases was made. (See pp. 112 and 71.) Even during the excessive carbohydrate feeding of typhoid fever it ' Pril.r.-mi: /tx-lir. f. klin. Mod.. 1014. Ixxxi. p. li'O. PROPHYLAXIS 2.S9 is said that glycosuria seldom develops. Cases of diabetes following infectious disease, however, are so startling when they do occur and at times, though by no means always (see the case of Schmitz, p. 4S), are apt to be so severe that the physician should always examine the urines of patients during an infectious disease, during convalescence and without fail before discharging the patient. This is doubly important because a nephritis might be disclosed even though diabetes were not. The usefulness of urinary examinations will become increasingly apparent just as soon as busy physicians get the habit of testing the urine for albumin and sugar in the patient's home with as little formality as they make a test of the blood- pressure. 5. Pregnancy. Sugar is apt to occur in the urine during preg- nancy. The subject will be discussed more in detail later on under the treatment of diabetes in pregnancy. It is mentioned here simply to emphasize the point that mild cases of glycosuria, which go untreated in pregnancy, may later on in the same or in subse- quent pregnancies become aggravated cases of diabetes. It is perhaps unnecessary to absolutely exclude the last trace of sugar from the urine in pregnancy when the sugar amounts to only 0.2 or 0.3 per cent, on a diet which contains as much as 100 grams of carbohydrate or more, but it is essential that the percentage of sugar should be kept as low as this, and upon the first indication of an increase routine diabetic treatment should be instituted. (>. Hygiene. (a) Physical Hygiene. Any agency which promotes physical or mental hygiene is a step toward the prevention of dia- betes in the predisposed, and the abatement of its severity when it has appeared. It should be recorded to the creditof I )r. A. J. Hodgson 1 that for years in dealing with his patients he has urged that they "should be kept mentally indolent and physically active." The ex- periments of Cannon,- Folin 3 and their associates upon the appear- ance of sugar in the urine of animals and of both normal and insane individuals following periods of great emotional excitement have demonstrated the truth of the former half of the motto. Therefore all individuals who have a tendency toward diabetes should be especially urged to take vacations, and the good effect of vacations should be generally pointed out. I have never forgotten the remark of Dr. Sabine, of Brookline, that in the course of his long practice he had observed that those of his patients who had taken active camping trips in the woods bore the stress of modern lite best. By this means exercise was combined with mental relaxation. That the good effects of each last for months is not hard to believe. It is only 19 200 TREATMENT natural to conclude that if tin- muscles, in which is stored one-halt' of the carbohydrate of the body, are kept in good condition by train- ing, a favorable effect must be exercised upon the general metabolism of carbohydrate. The man who, in middle life, gives up hard physical work and is promoted to an oflice chair with increased mental worry is exposed to diabetes. (b) Mental Hygiene.- The change in the mental attitude of patients during the course of treatment is a gratifying encourage- ment to the physician. Untreated diabetics after a moderate number of years usually show mental depression, and with women this often becomes pronounced. In the first ten years of my experience with diabetes I was much impressed with the tendency of such patients to cry, but even then, with the methods in vogue, it was interesting to see how depression disappeared with the decrease or disappearance of sugar in the urine. This could not be explained by the mental encouragement which a patient derived from his knowledge of the decrease in sugar excretion. Even when patients became free from sugar but developed acidosis, mental symptoms often improved, and to so great an extent that one could say that with treatment, even though it did end in coma, the patient enjoyed life far more thoroughly than when untreated, his life ended with debility or tuberculosis. Might this be asso- ciated with depression of the metabolism in the one instance and its stimulation in the other? During the last two years and a half the mental attitude of the patients has improved still more. The enthusiasm about new methods of treatment has been so great as to account partially for this, but the actual improvement in health which the patients have felt has been of more importance. (Jreeley explained to my patients how diabetes has largely been robbed of its terrors. lie urged the simple life as a great aid in treatment and told them not to try to be first in the Iberian village and be ill, but rather to be second in Rome and keep well. He told them to have a hobby, and not to make it a labor; to be cheerful and to keep their minds occupied, and as far as possible to continue the previous currents of their lives. Heavy responsibilities should be avoided as well as nervous upsets and emotional excitements. It is almost as dangerous for a diabetic to get angry as for a man \vith angina pectoris. Case No. 1 lf)7 had been sugar-free for five days, but it came back when he had an important conference with one of his superintendents. Allen's diverting exercises have promoted the mental hygiene of the diabetic patient in a most helpful manner. In Section II the effect of exercise upon diabetic patients has been discussed at length. Stimulated by Dr. Allen, I have gradually increased the exercise of all my patients, except those unduly weak or in a dan- PROPHYLAXIS 291 gerous condition upon entrance to the hospital. The effect of this increase of exercise upon the well-being of fat diabetics has been pronounced, and it is striking how many miles a semi-ill or an obese diabetic patient can learn to walk during t\vo weeks. The patients are encouraged to take their walks soon after meals and to go out- doors at least five times in the day. Not alone are the good effects of exercise shown by freedom of the urine from sugar with an increased carbohydrate tolerance, but by improved circulation and general well-being. Even fasting diabetics, as a rule, appear to do better when up and about the wards for a few hours a da}' than when abed. However, caution is necessary in suggesting this plan to severe cases of diabetes. It must be remembered that Case No. 304 went into coma following a period of active exercise in the year 1910, although the diet had been only moderately restricted; yet it should be said that she was a case of nine years' duration, and any lowering of the carbohydrates without a corresponding decrease in fat in such a patient is apt to prove serious. It is yet too early to say whether we shall be able to change "the emaciated, soft- muscled diabetic into an athlete," but there is no question but that it is worth while to try to do so. Xo case should be considered too far advanced for an attempt at muscular redevelopment. I have seen two patients, so weak from inanition that they could not stand, through .the help of skilful massage and carefully planned dietetic treatment again begin to walk. By exercise, as Allen points out, fat must be reduced and muscle tissues built up. I can agree with Dr. Allen that "it is hoped that an end may be put to the period of the pale, feeble diabetic, dressed in double underwear, hugging the radiator, and growing more neurasthenic all the time." I must acknowledge that I have not yet been able to wean my patients from the radiator, but I expect to be able to do so by means of exercise and more food. Exercise must not be carelessly undertaken. Graham 1 demon- strated that whereas the blood sugar after a meal of 100 grains glucose attains its maximum in twenty minutes and returns to normal in an hour to an hour and a half, under conditions which cause fatigue the blood sugar rises to a greater height and takes three or four hours to fall to its original level. Cannon 2 also observed an increase of sugar in the blood during fatigue but considered this a result of design on the part of the body's protective mechanism. If the patient, by means of exercise, can have 5 grams more of carbohydrate a day the added comfort will be enormous, for the addition of 5 grams of carbohydrate to a diet in a case of severe diabetes brings almost untold joy. It allows various alter- i: Jour. Physiol., 1916, 1, p. 285. 2 Cannon; Am, Jour. Phy^iol., 1914, xxxiii, p. 372. 292 TREATMENT natives half a small orange, .">() grains of strawberries, a small tablespoonful of cooked oatmeal, or a potato half the si/e of a pullet's egg. Finally, it is astonishing ho\v much exercise a diabetic in training can take. One of my severe cases living on a strict diet several years ago walked between twenty and thirty miles in one day. Inquiry elicited the following letter from Case No. 7SM, a Harvard student, \vlio frequently shows a small trace of sugar which borders upon the renal type of diabetes. The blood sugar one morning before break- fast was 0.07 per cent.: CAMHUID<;K, MASS., Doci'mber 1, 1015. "MY DKAK DOCTOR JOSLIX: 1 first noticed the eil'ect of exercise last spring. I was rowing for exercise at the time, and observed that if I went out on the river about half an hour after lunch and rowed for an hour or less, the test would not show any sugar in the urine at any time during the afternoon, even though 1 ate potatoes and a small amount of bread for lunch. But if I ate any potatoes (no bread) without so exercising, the test always showed sugar about two hours after the meal." 7. Syphilis. The type of diabetes which occurs in syphilis is usually mild. Apparently the disease is not very closely related to diabetes, but all statistics must stand in abeyance until the results of modern tests for the detection and treatment of syphilis are known. Approximately 2 per cent, of my own cases of diabetes surely had syphilis. In the first 400 cases the disease was recognized six times, in the second 400 cases seven times and in the third 400 cases there were six instances. At the Peter Bent Brigham Hospital the proportion of syphilis (positive YYassermaini tests) among dia- betics was found to be S per cent, among SO diabetics by \\alker and Ilaller,' but this was only about half the incidence of syphilis among 4000 of the general hospital population. These figures arc* in sharp contrast to the surprising figures of \\arthin and \Yilsoir for Michigan. They record that in six autopsies performed at the 1 diversity of Michigan upon diabetics since HK)7 histological changes of syphilis were present in all and in four spirochetes were demonstrated in the myocardium and in one of these- cases also in the 1 pancreatic lesions. As evidence of the connection between diabetes and syphilis these authors examined the pancreas in a series of Mi) cases of old latent syphilis and found it abnormal in all. I make- no effort to explain the- discrepancy between tin- Massa- chusetts and Michigan figures. Barach' 1 among Ml diabetics seen within two and a half years, 1 Walker and Hallcr: .lour. Am. Mr,!. Assn., lOKi, Ixll, p. 4SS. - Wurthin ami WiNon: Am. .lour. Mrd. Sc., HMD, clii, p. 157. 3 Burach: Boston Mod. and Surn. Jour., 1017, rlxxvi, p. 58. PROPHYLAXIS 293 believes there were primarily 3 cases of syphilis and he notes that another had a positive Wassermann reaction, another was the hus- band of a syphilitic and another the husband of a paretic. Several of the cases with a syphilitic history have been of long duration. This type of the disease has occurred so frequently in connection with known syphilis that one must suspect the presence of syphilis in mild diabetes. (See Table 139.) In illustration of the above 1 would cite the following case: Case Xo. 503 acquired syphilis at the age of sixteen years, and was energetically treated by the older methods. Fourteen years later, in May, 1912, when about to undergo an operation for appendicitis, sugar was found at a routine examination, and the Wassermann reaction, which had not previously been tried, was demonstrated to be positive. lie was given a thorough course of treatment with salvarsan, and the Wassermann reaction has remained negative. Under dietetic treat- ment sugar decreased and gradually a marked carbohydrate toler- ance was acquired. Xo further antisyphilitic treatment has been given. On August 19, 1915, the urine contained 0.5 per cent, of sugar and the carbohydrates in the diet amounted to 140 grams. The blood sugar amounted to 0.1 (> per cent, fasting. 1 On March 8, 1917, the urine was sugar-free and the high tolerance persisted. It would seem as if syphilitic diabetes was an ideal object for etiological treatment, but all writers agree that this is not usually the case. It is true that occasionally a case has been strikingly helped. For instance, Umber 2 recorded a case of diabetes, male, aged forty-eight years, with syphilitic infection in 1S9S. During 1909 indigestion, pain in epigastrium, thirst, transitory jaundice, loss of 18 pounds, and gastric tumor occurred. Upon a diet of 147 grams of carbohydrate, 113.4 grams of sugar were eliminated. Fatty stools, blood sugar 0.4:5 per cent., Wassermann reaction positive. Following the use of 0.4 gram of salvarsan the tumor decreased in si/e and the stools became less fatty. There was a continuous gain in weight for one year, and at the time of recording the case the patient lived without restriction of carbohydrate and no tumor was to be felt. Hevillet :s reports a cure with mercury. Walker and llaller report that energetic treatment of one case of diabetes developing six months after the initial lesion was without any influence on the diabetes. The blood Wassermann became much weaker, but never negative and the patient finally died in coma. X"o such similar case has come under my personal observation, but I strongly feel that one should constantly be on the watch for syphilis in diabetes, and whenever it is present should treat it by 1 See Table 113. - Umber: Mlmchen. med. \Vchnschr., 1911, Iviii, p. 2499. 3 Revillet: Lyon Med., 191G, cxxv, p. 374. 294 TREATMENT modern methods. Sufficient time has not yet elapsed since the better treatment of syphilis has been introduced to warrant copying the statement so , no sugar was found upon two examinations, and on May 27, 1 ( .)12, sugar was absent from the urine. This case was lost sight of for several years, but finally dis- covered by my secretary, Miss Y\ood, to be alive on January 4, 101C), but on April 2.">, IDIO, the urine- was obtained and showed f>.2 per cent, sugar and no diacetic acid. The patient remains in quite good condition and on May 2">, 11)17, the blood sugar was 0.22 per cent, and the urine showed 0.7 per cent, sugar and no diacetic acid. A similar case, reported by Miller 1 , has also been traced by him and he has kindly allowed UK- co report upon it. The 1 urine remained sugar-free for two or three 1 months after the operation, but later malignancy developed in the 1 pelvis, sugar returned and ee)ntinue > d until death. The ivmoval of the- prostate has also led to the disappearance of CLASSIFICATION 297 diabetes. Xo case of this sort has come under my personal super- vision and my experience with diabetes in association with enlarged prostate has been extensive, through the courtesy of Arthur L. Chute. B. CLASSIFICATION. 1. Classification for Treatment. A classification of diabetic cases based upon the assimilation of carbohydrate will always be found helpful in their treatment. It is customary to divide cases of diabetes into three types: mild, moderately severe, and severe. Such a classification, however, can never be arbitrary, because easels which at first appear to belong to the severest type of the disease may run a favorable course and cases showing at the outset only a small quantity of sugar may prove to be quite intractable. For this reason Xaunyn did not pretend to be able to distinguish accurately between the various types. Furthermore, up to the present time it has nearly always been considered that diabetes was a progressive disease, and that each patient, if he lived long enough, was destined to pass through the three stages. Time alone will decide whether this unfortunate conception of the malady may be given up. In illustration of the above, Case Xo. 344 belonged to the type of mild diabetes for nearly four years, but gradually the character of the disease changed, and ultimately reached its greatest severity in this patient; tuberculosis then intervened, acidosis disappeared, and he died, four years after the onset, of tuberculosis rather than of coma. Case Xo. S responded so well to treatment as to justify being classed as mild in type, and even at the end of fourteen years was only moderately severe, death ensuing without acidosis but as a result of arteriosclerotic complications. Case Xo. 552 appeared to belong to the severe type of diabetes, but after prolonged treatment improved so much as to reach the border-line of the moderately severe group. Diabetes in children is usually looked upon as severe, but a distinct fraction of these cases prove to be mild and the severity at first observed can be proven to have been due to the hand of man and not intrinsic in the disease. I believe few cases of diabetes are progressive in character. (See Diabetes of Fifteen or More Years' Duration, p. 4()S). Usually by "severe" diabetes is understood those cases where, to quote von Xoorden, "notwithstanding a prolonged rigid carbohydrate-free diet the urine contains sugar." While this definition is open to various criticisms, it is most useful. By a diet as free from carbohydrate as possible, I think that von Xoorden often means one which contains vegetables with 5 per cent, carbo- hydrate or less; in other words, a diet with about 10 grams carbo- hydrate. It should be emphasized that cases of diabetes of long 29S TREATMENT duration are frequently seen who, though upon an almost carbo- hydrate-free diet for a prolonged period, continue to excrete in the urine a very small percentage of the carbohydrate eaten; this persists for weeks, but finally disappears, yet at no time could such a patient be classed as severe. Such a case was No. 1220. The introduction of prolonged and intermittent fasting simplifies yet complicates the whole question of classification, because prac- tically all patients can be made sugar-free. This would banish the term "severe diabetes" from the category, and this is undesir- able at the present time. Therefore classification in this section on treatment will be based upon the supposition that severe cases of diabetes have a tolerance from to 10 grams carbohydrate, moder- ately severe cases have a tolerance from 10 to .10 grams of carbo- hydrate, and that the remaining cases are mild. This is a cheerful and deserved tribute to Dr. Allen. It is one of the best proofs of the improvement in the treatment of diabetes, but it makes imprac- ticable any satisfactory classification of all my earlier cases. Of the last 100 cases seen, writing one year ago, 12 might be con- sidered as severe cases, .12 moderate and the remaining >(> cases mild, but if 1 take the last 100 cases prior to March 1, 1<)17, the statistics are S severe, 47 moderately severe, and 4-1 mild. It should be recognised that even the severe cases might be so prepared by prolonged fasting for a test of their carbohydrate tolerance as to appear mild. A tolerance obtained under such conditions would b;^ apparent, not real. The presence, absence or intensity of acidosis affords an unsatis- factory basis for classification. Even the mildest case of diabetes by restriction of carbohydrate and increase in the amount of fat can be made to develop an acidosis which will be mild, moderate, or severe in degree. The Lusk dextrose-nitrogen ratio is another method of classi- fication. Lusk groups cases as severe which, when put upon a protein-fat diet, have a dextrose-nitrogen ratio (1):.\) of .">.(>.") to 1. Today he recognizes this test as unsatisfactory because of the dangers which lurk behind such a diet, but the introduction of fasting has given a sate basis for the determination of this ratio as Lusk and Janney have emphasized. The usefulness and reliability of this ratio are unquestioned when fasting lasts several days and constant conditions as to intake of liquids and salts are maintained and alkalis are omitted. Hut with my present way of thinking, I should consider it a matter for self-reproach to allow one of my cases to reach a stage that he is able to void a urine with a D:X ratio of .">.('>.") to 1. This very fact demonstrates the utility of this method of classification and its impracticability as well. \Vriting in April, 1017, it now appears that no such case of complete diabetes has DIETETIC TREATMENT 299 ever been reported if the criteria above mentioned for such a test have been followed. Unfortunately, this test is unserviceable for the general practitioner, because the facilities for determining the nitrogen in the urine are not at hand. Usually the dextrose- nitrogen ratio vanishes with fasting save in the few rare cases with complications. Clinicians should not expose patients to a protein-fat diet simply to determine their severity. Case Xo. 51.'!, already referred to, excreted while fasting, a few days preceding coma, suffering at the time from a septicemia, 35. 3 grams of nitrogen and 112 grams of dextrose, showing a dextrose nitrogen ratio of 3.17 : 1; the ratios were higher when he was upon a fat- protein diet. The respiratory quotient of a series of cases of diabetes was found by Benedict and myself to be 0.73 or less, and the respiratory quotient of a series of milder cases to be above this figure. But the respiratory quotient depends in such large measure on the diet as to render it an undesirable method by which to differentiate cases of diabetes even if it were practicable. Then, too, the rise in the quotient which has been observed with fasting diabetics confuses the picture. 2. Classification for Prognosis. It is still more difficult to classify cases of diabetes for prognosis. The past is not a guide. Hitherto a case of diabetes occurring in an individual under the age of thirty years would usually be set down as severe, even though he readily became sugar-free; exceptions occurred now and then when the quantity of sugar was a mere trace and the patient obese. Cases often appear severe when first seen, but upon further acquaintance it is found that this is due to some a lie viable circumstance, such as the presence of an infection, or, more commonly, the sudden institu- tion of a fat-protein diet with its attendant acidosis. The presence of obesity, a favorable heredity, an early diagnosis, or the history of benign diabetes of several years' duration with gain rather than loss in tolerance, the retention of body weight, are good prognostic signs, but a placid, cheerful, brave and honest disposition, inherent or acquired, is fully as important. C. DIETETIC TREATMENT. 1. General Discussion. The treatment of diabetes is so largely dietetic that it is especially important at the start to take a com- prehensive view of each case. Unless this is done before treatment is commenced, the attention of the physician may be so engrossed with the diet that he falls into schematic ways and overlooks other salient aspects relating to his patient. The etiology of the disease in each instance should be carefullv investigated and faultv habits 300 TUP: AT ME XT corrected. It will often be found that the symptoms which annoy the patient, bear little relation to diabetes, and these should be appropriately treated. This is particularly true in the diabetes of middle life and old age, and also with the varieties of diabetes which are the accompaniment of the diseases of the various organs of the body. Tins broad view of the case should be maintained throughout the course of treatment, and whenever unsatisfactory results an v obtained the whole situation should be investigated anew to determine whether some radical change in the plan of treatment should be adopted. A surgeon often overlooks grave medical complications in his cases simply because his attention is absorbed by the surgical aspect of his patient. More' than once 1 have discovered in my own practice, as well as in that of others, the existence of advanced tuberculosis which had previously escaped attention. Patients with diabetes often come to the physician in a state which is endurable. It is the function of the physician to improve upon this state-. Only too frequently treatment in the past has done the patient more harm than good, but the fault lies not in the principles of treatment, but rather in their application. The physician who undertakes to treat the patient with diabetes whose condition is comfortable resembles the surgeon who, operating for an interval appendix, assumes a responsibility far greater than when acute symptoms make- such an operation imperative. In illustration of this point 1 would cite the following instance: Case No. 47i>, aged forty-eight years, buried a child of diabetes at the age of fifteen years. During November, 1011, she began to lose weight, and though two quarts of urine were voided, no sugar was found in t\vo inornuifj specimens of urine either early in November or December. On December ">, 1011, a little sugar was discovered. January ~1\, 101L', she consulted a "specialist." The quantity of urine was then six quarts, the percentage of sugar 7, making the total quantity of sugar for the day nearly a pound. It was reported to me that she said that if she had diabetes " she would not do a thing for it." The "specialist" radically restricted /lie carbohydrnlc in tin 1 il/cf, and three days later, in consultation \vitli her family physician, I found her in coma. This is one of the cases which led me to consider the first year following the detection of the disease to be the diabetic's danger /.one. It is one of the cases which formerly would have gone down in the literature as "acute diabetes." The responsibility for the management of the diet of a diabetic patient should always rest upon one individual. As a rule that individual is the patient, but at times another member of the house- hold. Children who are above the age of ten vears should be taught DIETETIC THE ATM EXT 301 to plan their own diet. They readily learn to do this and in so doing make their elders blush. In fact it is more important for diabetic children to learn what and how much to eat than all the knowledge which their schools ail'ord, for upon this information their life depends. Perhaps it is because this personal responsibility is so deeply felt in the management of little children that the treatment of diabetes in them proceeds so uniformly and always produces results so much better than are expected. Conversely, the failure of diabetic patients to do well in the open wards of large hospitals has been due not so much to the alleged dishonesty of the patient as to the division of responsibility among several nurses. Errors in the diet or in the collection of the urine must be promptly traced to their source. While inaugurating the plan of treatment for a new diabetic patient, we physicians may well take t;> heart the criteria which serve as a guide to V\*. J. Mayo in his selection of the technic to be adopted in a given surgical case namely, the eii'ect of the method upon (1) the immediate mortality; (2) the permanence of the relief afforded, and (o) the resulting disability. Formerly we sinned most against the first rule, but now we are often found guilty of breaking the second and third. The treatment of a patient with diabetes lasts through life. Treatment must therefore be adjusted to this condition and should be so arranged that it can be continued for years without harm. Consequently, the patient must be taught the nature of his disease and how to conquer it. Clinicians with a large clientele of diabetic patients have almost invariably been in the habit of giving their patients printed matter in order to acquaint them with the character of the disease and how to control it. Allen has gone a step further by adding to didactic treatment the practical instruction of every patient in the simple urinary tests for sugar and diacetic acid. The instruction of the patient should not end with urinary tests: it should include didactic and practical instruction in the prepara- tion of his diet. In hospitals there should be daily lessons. The patient is at school to learn how to save his life. Time should be taken at the beginning of treatment to thoroughly describe to the patient the plan of procedure and the necessity of his being under close observation until the urine is sugar-free and until he under- stand- how to keep it so. The diabetic patient should be made to realize that he has a lesson to learn, and the earlier he learns it, the sooner he can be discharged. Time also will be saved and many advantages accrue if a diabetic nurse accompanies the patient to his home and supervises the diet for one or more weeks under conditions peculiar to his household. Moderate changes in diet may improve the condition of a patient so much that unless he understands the 302 TREATMENT whole situation, ho may bo contented with partial improvement and not care to make the efl'ort to remain sugar-free. At each visit a specimen of the twenty-four-hour quantity of urine should be brought, or better sent, in advance, together with a written list of the character and weighed quantities of the food eaten during the same period. So soon as carelessness in this regard is condoned, indiffer- ence to other advice follows. The advantages of having the patient bring a written diet list are twofold: (1) it is more accurate, and (2) it saves an enormous amount of the physician's time (see p. 477). The changes in weight of the patient should be recorded. The specimen of urine should be that of the twenty-four hours immedi- ately preceding the visit, but if it is not, do not neglect to immediately secure and test a specimen. In diabetes one cannot be in too close touch with the actual condition of the patient at the moment. The advantages of hospital as compared with ambulatory treat- ment are discussed on pages 49)3 to 502, and under the captions "Directions for Patients" and "Directions for Xurses" on pages 47(i and 477 is included material which in order to avoid repetition is omitted here. Formerly physicians hesitated to instruct all their patients to make urinary tests, and though I began this method years ago, I did not persist in it. The plan often proved unsatisfactory, because so frequently disappointment followed the reappearance of sugar and it was difficult to keep the patient sugar-free. Now the situation is different. Case Xo. Mo, seven years of age, was able not only to examine his own urine but that of other patients in the hospital, and ( 'ase Xo. 901 , a little girl of four years, takes for granted that the diet will be changed when she sees her own urine give a posi- tive Benedict test. Case Xo. 995, aged six years, glories in the fact that his Benedict test is better than that of some of the other patients. Patients weary of testing the urine, but only under very exceptional circumstances should they be allowed to omit making the test. A daily negative Benedict reaction gives confidence and, on the other hand, in the presence of danger, it is never wise to follow the habits of the ostrich. The dietetic treatment of diabetes is only successful when the urine of the patient is free from sugar, the blood sugar normal, and the blood lipoids as well. That the first of these conditions should hold has been universally recognized for years, but it is only since Dr. Allen introduced prolonged fasting in the treatment that it has l)een easy for the general practitioner to secure this result in prac- tically all cases. The increasing belief that the nature of diabetes is more functional thi.n organic in character has given a new impetus to treatment, and has emphasized the importance which Xaunyn laid upon sparing a diseased function, earlier recognized by Hoffman as a general rule for the treatment of functional disorders. Most of DIETETIC TREATMENT 303 this function is evident if the blood sugar is normal and the urine of the patient is sugar-free. In what follows, treatment is based on urinary examinations, and to a lesser extent on the estimation of the blood sugar. The study of the blood lipoids is in its infancy and, though of much value in a broad sense, it is seldom as yet therapeutically useful in the individual ease. Even examinations of the blood sugar are not yet sufficiently simple to be available to most physicians, and, furthermore, not enough experience has been gained to indicate arbitrarily how one should proceed when the urine has become sugar-free and the blood sugar is persistently high. The introduction of prolonged fasting and the new horizon which this opened up for fruitful scientific investigation, have changed the treatment of diabetic patients from a duty to a pleasure. The simplicity of the method shows its worth, and that it is safe for the overwhelming majority of patients is attested by the low mortality which has followed its use. But it must not be forgotten that the first year of diabetes following the discovery of the disease is the diabetic's danger zone. The first series of statistics of the Massa- chusetts General Hospital from 1824 to 1898, the second series between 1898 and 1913, the records of Griesinger, Xaunyn and my own statistics prior to 1910, demonstrate this by showing that 68 JXT cent, formerly (Table Xo. 15, p. 33) and 17 per cent, recently (Table Xo. 21, p. 36) of the fatal cases of diabetes succumbed during the first year of the disease. Yet diabetes is an out-and-out chronic disease, and the greatest mortality from it should occur not in the first year following its discovery but in later years. It is my firm belief that the first year of a diabetic's life should be his safest, and that the reason that this has not been the case is best explained by the method of treatment to which he has been subjected rather than to a lack of treatment. In line with this supposition is the fall in mortality during the first year of the disease among my cases, from 16.9 per cent, for the year ending l)eceml>er 1, 191.3, to 14.8 per cent, for the year ending December 1, 1916. The second year of the disease now becomes my diabetic danger zone, and with this change in conditions new problems arise. This improvement in treatment in my own series of cases 1 attribute first and foremost to fasting, and second to a better recognition of how to safeguard its use. I have endeavored to make fasting treatment safe as well as simple by preparatory treatment, by which I mean the omission of fat, the gradual reduction and final omission of protein and then continued reduction of carbohydrate with eventual fasting if required. This is not necessary for some cases, but it is advised in all severe, long-standing, complicated, obese, and elderly cases, and also in all those showing aeidosis. 304 TREATMKXT Sixty per cent., formerly two-thirds, of all the fatal eases of diabetes I have seen have died in diabetic coma. In all but one of the fatal cases in children death has been due to coma. My records show that of all those 1 diabetic patients who have succumbed to the disease during the first year of itscourse,comahasclaimed Xi2perccnt., formerly S(i per cent. If the first year of the disease is the patient's danger /one, it can be as truly said that the danger and almost the only danger is coma, and brushing aside all technicalities, coma must be considered from the practitioner's stand-point as synonymous with acid poisoning. Therefore, in the plan of treatment which follows it will be seen that from beginning to end the prevention and treatment of acid poisoning is constantly kept in mind. If aeidosis exists when the patient comes for treatment he must be in daily communication with his physician. In the past it has not been so very unusual for a patient apparently in fair health, upon beginning- treatment, to fall into coma (Case Xo. 473, page 300), and so, too, patients upon leaving the hospital likewise in good condition, though on a restricted diet, have also died a few days later in coma. (Case Xo. 317, page .170; Case Xo. 074, page 34-1.) 12. Author's Flan of Campaign against Diabetes. -Xever before has the plan of campaign for the treatment of diabetes appeared to me so clearly defined and never before have I entered upon it with as much hopefulness. The principles which I keep before me are: 1. The prevention of aeidosis in large measure accomplished by the omission of fat from the diet at the beginning of treatment, and the prompt treatment of aeidosis by similar means should that subsequently develop. 12. The attainment of a substantial positive carbohydrate balance for each patient, remembering that the owner of a carbohydrate tolerance and the owner of riches have much in common, for it is as true of one as of the other, that "whosoever hath, to him shall be given; but whosoever hath not, from him shall IK: taken away even that which lie hath." 3. The simplification of treatment, the better education of t!ie patient and closer cooperation with his physician. 4. The avoidance of inanition, not to be confused with loss of weight, by the protection of body protein and greater care in the employment of fasting. Fasting resembles in its action our best drugs: morphin and digitalis. These frequently give surprisingly good results even when they are carelessly employed, but they often do harm, and their true worth is only disclosed when they are prescribed with intelligence. Furthermore, if inanition or a tendency to inanition and hunger can be avoided, a smaller number of patients will yield to temptation, break treatment and in consequence die of coma. DIETETIC TREATMENT 1505 3. Summary of Dietetic Treatment. It will conduce to clearness to summarize the routine plan of treatment now employed by the writer, and subsequently to discuss its general phases and the exceptions to it which occasionally must be made. By this method the patient usually becomes sugar-free within a few days, and acidosis, if present, disappears. Alkalis are unnecessary, and if given do harm. Preparation for Fasting. In severe, long-standing, complicated, obese, and elderly cases, as well as in all cases with acidosis, or in any case if desired, without otherwise changing habits or diet, omit fat, after two days omit protein, and then halve the carbo- hydrates daily until the patient is taking only 10 grains; then fast. In other cases begin fasting at once. Fasting. Fast four days, unless earlier sugar-free. Allow water freely, tea, coffee, and thin, clear meat broths as desired. Intermittent Fasting. If glycosuria persists at the end of four days, give 1 gram protein or 0.5 gram carbohydrate per kilogram body weight for two days, and then fast again for three days unless earlier sugar-free. If glycosuria remains, repeat and then fast for one or two days as necessary. If there is still sugar, give protein as before for four days, then fast one, and then gradually increase the periods of feeding, one day each time, until fasting one day each week. I have seen no uncomplicated case fail to get sugar-free by this method. Carbohydrate Tolerance. When the twenty-four-nour urine is free from sugar give 5 to 10 grams carbohydrate (15;) to oOi) grams of 5 per cent, vegetables) and continue to add ~> to 10 grams carbo- hydrate daily up to ">() grams or more until sugar appears. Protein Tolerance. When the urine lias been sugar-free for three days, add about 20 grams protein and thereafter lo grains protein daily in the form of egg-white, fish or lean meat (chicken) until the patient is receiving 1 gram protein per kilogram body weight, or less if the carbohydrate tolerance is zero. Fat Tolerance. Add no fat until the protein reaches 1 gram per kilogram body weight (unless the protein tolerance is below this figure) and the carbohydrate tolerance has been determined, but then add 5 to 2."> gram:; daily, according to p:'evious acidosis, until the patient ceases to lose weight or receives in the total diet about oO calories per kilogram body weight. Reappearance of Sugar. The return of sugar demands fasting for twenty-four hours, or until sugar-free. Resume the former diet gradually adding fat last in order to maintain as high a carbo- hydrate tolerance as possible, sacrificing body weight for this purpose. Weekly Fast Days. Whenever the tolerance is less than 20 grams carbohydrate, fasting should be practised one day in seven; when 20 306 TREATMENT the tolerance is over 20 grains carbohydrate cut the diet in half on one day each week. The foods commonly employe 1 :! in determining the tolerance for carbohydrate and protein are f) per cent, vegetables, oranges, oatmeal, potato, fish, chicken (lean meat). (For the food values of these and other foods, see Table 120.) A common course of treatment when a new patient (Miters the hospital is to prescribe 1 "> percent, vegetables (150 grams) and a small orange at each meal, if uncertainty about the character of the case exists, but otherwise to commence fasting at once. To avoid repeti- tion, illustrative cases are inserted on later pages where the treatment is described more in detail. (See Tables 142, 14/5, 144, 145, 140, and other tables in Section VI.) TABLE 141. II.I.I-STKATIOX OF AMBULATORY TREATMENT WITHOUT FASTING OR OMISSION OF PUOTEIX. CASE Xo. 1237. ACE AT ONSET ix SEPTEMBER, o!( YEARS AND 5 MONTHS. Dietary proscriptions in ura.n It is by no means necessary to follow this same program, and recently the following simple schedule worked admirably with an ambulatory patient. (Table 111.) It had the advantage of avoiding reduction of protein for even a single day. ( 'ase No. 12l>7, age at onset thirty-nine years, presented himself for treatment a year and a half later, and without acidosis, but with '>><> grams sugar in the urine during twenty-one hours. The directions given may be summarized as follows: Take one-half pound (240 grams) 5 per cent, vegetables, one-quarter of a pound fish (120 grams), and one small orange at DIETETIC TREATMENT 307 a meal for two days; on the third day omit half and on the fourth day all the orange. When sugar-free, exchange one-quarter pound fish for 3 ounces meat, and next replace another quarter pound fish by 2 eggs; then replace 2 eggs with 2 ounces bacon and subsequently add one-half ounce butter a day for two days, to be followed every other day by the addition of an ounce of 20 per cent, cream until 3 are taken. Similarly, thereafter every other day add one-half orange until one-half is taken at a meal and from then on every other day 1 ounce potato until as much as desired is taken, or sugar appears. 4. Details of Treatment. It is recognized that the above outline of treatment is purely schematic, and that it should be adapted to the weight, age, digestion and tastes of the individual patient. The dietary charts and urinary records of various recent cases illustrating the plans of treatment above outlined are to be found on pages 308 and 309. It will now be in place to discuss in detail the above methods of procedure, which constitute the diabetic pharmacopoeia. (a) Preparation for Fasting. It is not only more rational, but it is easier to prevent acidosis than to treat it. This is the reason for the preparation for fasting. The majority of diabetic patients show little acidosis upon fasting, or if this has been present it will decrease. (Table 144, page 309.) On the other hand, it is not always easy to predict 1 what will occur, and an acidosis which has not been present may appear or an existing acidosis may grow worse. There- fore it is safer to take pains to avoid the development of acidosis in those predisposed to it, for it is a sound rule of all treatment that patients coming to the physician in an endurable state must not be made worse or have their lives jeopardized by the therapeutic procedures adopted. Individuals predisposed to acidosis are those in whom the disease is of long duration. These are the patients who, after having lived in a fairly comfortable condition for years, succumb when placed upon a fat-protein diet. The last case of this kind in my own experience was a very severe one (Case Xo. SS7, page 3o9), who had had diabetes twenty-nine years, and is described in full under Cases Unsuccessfully Treated by Fasting (page 350). Contrast this case with the following two cases. Case Xo. 979, a woman, aged forty-nine years, developed diabetes at the age of thirty-two. When I first saw her, seventeen years later, January 2(5, 191(5, she showed 7.4 per cent, of sugar and no diacetic acid. It \vill be seen from Table 142 how she became sugar- free without the development of acidosis by the elimination of fat and the restriction of protein, followed by the gradual diminution of carbohydrate. : Stillman: Am. Jour. Med. Sc., 1916, cli, p. 505. 308 TREATMENT TABLE 142. CASE No. 979 OF SEVENTEEN YEARS' DURATION, ILLUSTRATES IIo\v PREPARATORY TREATMENT MAKES FASTING UNNECESSARY. Date, 191(1. Urine. Diet in gran s. Dietary prescriptions in grains. Sugar. 3 | n .~ " "3 n i -c IVr Total 5 2 | cent. KIMS >. C 1 +J ed "c Calories "u '3 Vegetal) 5 per Ui C 2 a a *a "?, fc4 In *> 1 J;in. 25 25-20 21) 27 . . 7.4 f> 2 54 944 13()i 12S 300 1 1. 90 100 1 90 6 1475 '() 5.(> S3 142 IS 10 27 2S 1700 4.0 70 112 : 43 10 7(54 12S 300 1 1. 1)0 50 1 !)() 2.S-2!) 1200 2.2 20 72 3S 10 () 5S4 12S 300 1 2. '.)() 90 ; (i 20 30 1350 1.0 14 50 2S 11 411 127 300 1 1. 30 00 G 30 31 1350 0.0 ,S 40 2S 11 371 127 300 1 0. 30 00 6 Feb. 31- 1 1200 0.2 2 25 20 5 249 120 300 5 30 20 40 1 2 . . . 0.0 25 27 SO 2SO 120 300 5 30 40 20 1917 Feb. Hi . Tr. All)Uin in eonsiderab I le. 1 Case No. 1007, of fifteen years' duration, aged forty years at onset, was even more favorable, because in three days she easily became sugar-free, as is shown in Table 143. By the elimination of fat, followed by the rapid reduction of protein and the gradual reduction of carbohydrate, she became sugar-free without the development of acidosis, and fasting was rendered unnecessary. All complicated cases, especially those in which the complication involves the kidneys, heart (see Xo. 759, Tables 217, 2 IS) or thyroid demand preparatory treatment, for they are especially susceptible to acidosis. Case Xo. S(>9, p. 3C>2, with high blood-pressure and goitre is the last fatal case of this type which has come to my attention, and is also described in the section on I'nsuccessful Cases. In this group are also included elderly patients, because of their vulnerable kidneys. Very fat diabetics could appropriately be included, because of the tendency of even normal individuals who are very fat to develop acidosis, as shown by Folin and Denis, and so, too, patients about to undergo surgical operations. Finally, patients showing signs of acid poisoning are benefited by this preparatory treatment before the fast, but if they have been accus- tomed to a low diet I proceed with fasting at once. Such an instance was Case Xo. 102") (see p. 300), who was fasted for five days. The patient was a girl aged twenty-one years, with onset of diabetes a year before. She- had lived upon a fairly restricted diet, and despite the acidosis she impressed me so favorably that 1 decided to fast DIETETIC TREATMENT 309 !>inosiq 'Bp89OjfJ CO uoocg U5 joying O CO VoOjv C O IN - 833g 01 C-I s be }uoo Jji-i Of- 'ureajQ CO 'CM O ' 1 I S !.>I CO O C". .s- p30UrT!O i.O OO O , rH COCO CO a hi fm^oj OOO O O OO O -r - i M i C ^nujodcjQ --H-H OSOO oSuuj() O ? '" O 00 00 ,0^; ^; ;uao jod oi 'sa[qT!;oSio v OOO O OoOO O iO O i ^ I ~ H *^ ^uaa jod g 'sa^qB^oao \ S!?:? 8 C S S I -< ^H CO T-^ CO eqi '^qSiOAV -r co co N '^ . . . . o If -r x o o "- 1 c c-. i - X -M ^1 Cv O I-H 'COC-ll-H . ' --I- m 4 - S.'o : : i : ~ '. o o o o o I I .-I . it^ci': M ?) CM 01 -< a I fi g E P p 2 Q SH -mm -O ffl O5 O O Ci i O 3 Jib JT.[OOAJV uignui aa^sA'j )uoj jod of- 'UIBOJQ 03U13J() juoo jod o COO OO O ooooooooo o O O O O O "O O O O IN OO > o r ^) 310 TREATMENT her at once. Table 1 H- shows how successful this proved to be, both as regards sugar and aeidosis. The principle upon which preparatory treatment is based is the exclusion of the source of the acid poisoning. Since the chief source of acid poisoning is fat, this constituent of the diet is prohibited. If this is done the opportunity for the patient to develop acid poison- ing is greatly reduced, and for two reasons: (1) the chief source of acid bodies is removed, and no fat is then available for the for- mation of acid bodies except the fat of the body; (2) in conse- quence of the partial fast, which is thereby inaugurated, the pos- sibility of oxidation of some of the carbohydrate which the patient is eating is favored, and if this should fortunately take place, aeidosis is sure to decrease. So strongly have I been impressed by the stormy career of the diabetic patients in whose diet carbo- hydrates have been suddenly restricted and fat increased, in contrast to the placid course which those pursue from whose diet fat has been excluded and the carbohydrates left unchanged, that whenever I am asked to see a new case of diabetes I beg the physician either not to change the diet at all or to simply omit the fat until the consultation takes place. A little child, Case \o. 86"), page 3~)0, is the last case of this type which has proved fatal and is described with the I nsuccessful Cases. Today I am convinced that if I had instructed the physician to treat the patient according to the above outline, five days later when the patient was brought to the hospital, coma Avould not have ensued. I feel amply justified in making this assertion from experience with other cases of similar age seen both before and since that date. \\e must all recognize the dangers of fat to the diabetic. Knerr 1 begins the treatment of his cases by giving a teaspoonful of uncooked starch in a glass of warm water every two hours and accomplished in this fashion much the same object as above described, lie observed that both glycosuria and aeidosis gradually decreased and that by this simple means the use of alkalis and the danger of coma were avoided. The days of preparation for the fasting are also advantageous in that they allow the patient to become acquainted with the plan of treatment; fasting and a hospital he fears. Opportunity is also afforded for the doctor to thoroughly examine the general condition of the patient. This must never be lost sight of at the beginning of treatment. It is absurd to feed a patient without teeth with coarse' vegetables, or to give these to another patient who has diarrhea. Another form of carbohydrate must be chosen, but whether this is potato, oatmeal, shredded wheat, puree vegetables, or. in case * Kiii'rr: Missouri State Mod. Assn. Jour., 1'Jl'j, xii, p. 71. DIETETIC TREATMENT 311 protein is employed, a few oysters, whites of eggs, or boiled fish, is a matter of indifference. The bowels must be thoroughly opened, but I do not believe in free eatharsis. If the patient has not had a movement for several slays, give an enema, followed by some simple cathartic or mild aperient, and another enema twelve or twenty-four hours later, but do not purge the patient. Gain enough is obtained if a move- ment is produced once in twenty-four hours when it has only been taking place once in seventy-two. In other words, do not upset any patient who is in a tolerable state. Furthermore, allow the patient to continue his regular routine, avoiding excess in any direction. Remember what happens to an old man who is suddenly confined to bed, and the discomfort which follows confinement following a fracture. Do not force a temperate man to drink against his will. An advantage which the omission of fat from the diet affords is the rest which is given to the digestive tract. Former treatment, which increased the fat in the diet, was the converse of this, and frequently led to vomiting, with the result that patients on the verge of coma fell into it. In every way seek to prevent worry on the patients' part, and from the start give them to understand that they are at school rather than at hospital. Following the omission of fat from the diet comes the exclusion of protein. From protein comparatively few acid bodies are formed, but it is safe to avoid even these, and at the same time gradually approach the fasting treatment. Detailed observations upon the advantage of restricting protein at this stage would be advantageous. I may have erred in the restriction of protein. Note the easy course which Case Xo. 1237, p. 206, took in becoming sugar-free while protein was maintained at a normal level throughout. (b) Fasting. Fasting is never so rigorous as doctors or patients expect. Patients are more ready to undergo it than physicians to prescribe it. Quite as often it is as much a relief to the patient as it is discomfort. This is in part due to the gradual decrease in polydipsia and polyuria. Headache occurs less frequently than would be expected, and is usually dispelled by a cup of coffee. Xausea almost never occurs unless a patient is given alkali or alcohol. Children bear it more easily than adults. Case Xo. 799, with onset at eighty-three, shunned it and rightly, but she became sugar-free and now, two years later, is vigorous and remains sugar- free. It is always desirable to avoid fasting in the old, and this can be accomplished usually by the help of preparatory treatment. Fasting does not seem like fasting to the patients when they receive coffee, tea, cracked cocoa, cocoa shells and broths, and are given an unlimited supply of water. \Varm drinks are preferable. 312 TREATMENT It the quantity of urine, as it often does, falls to less than normal, the patients are urged to drink water freely. Clear meat broths are a great satisfaction. An analysis of the 1220 c.c. of broths taken by ( 'ase Xo. 7(1-") during three 1 days showed the total amount of calories therein contained to be negligible. Contrary to my experience with digestive cases, broths do not stimulate the appetite in fasting diabetics; they relieve it. (See p. 271.) The advantage of broths is probably due in part to this, but to a considerable extent to the patient receiving salt by which he may maintain the equilibrium of body fluid. In the case just cited the 1220 c.c. broths contained 10.7 grams salt and the patient took 12 grams more from the salt- cellar. It is passible that the salt is a more important factor in the treatment than has been supposed. There is an urgent, need for an accurate study of this phase of the subject. (Sec p. 2S.~>.) Patients should not be kept abed during fasting, neither should they be forced to be up all day. Reclining in a steamer chair requires no more exertion than rest in bed. They should be afforded diversion by visits from friends, walking short distances, easy handiwork, playing games, letter writing, and reading. In general they are glad to rest for the greater part of the first day of the fast, but upon each succeeding day I have noticed that they are desirous to increase the amount of exercise. Case Xo. 70") (see page o!4), a trained diabetic, who returned to the hospital in order to become sugar- and acid-free, at the end of three and one-half days of fasting, enjoyed, without fatigue, going to the theatre. 1 confess this was not with my advice, for I have endeavored to prevent exposure of diabetic patients from infection during fasting, and indeed, under all circumstances. However, several have developed intercurrent infections during fasting while in the hospital one a tonsillitis, elsewhere recorded (Case Xo. X13, p. 45!)) and Case Xo. ( ,)0"), another mild diabetic who acquired influenza from his neighbor, and rigid dieting was necessary before glycosuria disappeared. Case XO. ( ,)3S, a child, aged t\vo and a half years, underwent fasting treatment successfully in the presence of mild infection of the upper air passages. At one time a mild cory/a attacked each patient in a four-bed ward. Today any patient or nurse with a "cold" is isolated or goes oil' duty. Fortunately no infection developing after entraive to the hospital has proven fatal. It is surprising ho\v variable is the period required to render the urine sugar-free. Frequently a urine which contains 7 per cent. of sugar becomes free from sugar after fasting for four meals, and conversely, a urine with only '> per cent, of suti'ar may still retain traces after the patient has been deprived of food for three or four davs. In general, cases seen >oon alter onset become sugar-free DIETETIC TREA TMEX T 313 promptly, whereas the reverse is usually true for those of long duration. " In one ease we may be dealing with fatigue; in the other, exhaustion of an already weakened organ," according to Greeley. 1 However, Case No. 733, age at onset seventeen, was fasted twenty- six months later, when he showed 0.0. per cent, of sugar and became sugar-free in two days. The explanation in this instance wa* apparently the fact that the case was remarkably mild, being jf the obesity type; in fact, the patient's highest weight 11H) pounds was reached when he first came under observation, and during the preceding twenty-six months he had gained twenty-six pounds. (See page 340 for the conclusion of this case.) Children showing large amounts of sugar have also become sugar-free very promptly when the duration has been only a few weeks. Cases of long standing appear to become sugar-free more quickly with preparatory treatment than with an immediate fast. This is probably due to the avoidance of acidosis. Oct. lino Sept. 2,-) 26 27 28 >'.> 30 :!l I ' :i I :> (its ! 10 FIG. l.'i. A slight infection need not prevent fastin::. 485 and 4SS for dietary and urinary charts of thi.s ca-e. (c) Intermittent Pasting. The observation of Folin and Denis- that, an individual, normal save for obesity, went through a second period of fasting with less acidosis than the first, and the practice, 1 Greeley: Boston Mod. and Rurg. Jour., 1910, clxxv, p. 753. 2 Foliu and Denis: Loc. cit., p. 149. 314 TREATMENT observed by many clinicians of the old school, who advantageously fasted their diabetics one day a week, have given the cue to inter- mittent fasting. Very few of my patients have been subjected to a fast of more than four days. Xo patient has undergone a fast exceed- ing nine days. Such a prolonged fast is unnecessary, and even if the fast is carried out, it is doubtful if the patient would always become sugar-free. The apparent reason for the persistence of sugar in Case Xo. 010, who fasted the nine days, v\'as the- presence of a vulvar abscess. This patient showed a carbohydrate tolerance of only 2 grams for the subsequent year, but upon readmission to the hospital in May, 1910, and the institution of routine treat- ment, she became' sugar-free in one day and the tolerance rose to ")") grams carbohydrate and an equal quantity of protein, and the total calories rapidly reached over 20 per kilogram body weight. The gain persisted until autumn, when it gradually decreased, but it returned temporarily with renewal of energetic treatment and the use of levulose this spring. Inquiry among my friends shows that an infection of some kind is usually present when glycosuria persists after a fast of a few days' duration. This is not always the case, for the difficulty in rendering the urine sugar-free 1 may be simply due to the extreme severity of the disease. Alternate feeding and fasting are adopted when it is found that the glycosuria persists after a preliminary four days' fast. The method which 1 have found most successful has been to allow about 1 gram protein or one-half gram carbohydrate per kilogram body weight for two days and then fast again for three days unless earlier sugar-free. The sugar promptly increases in the urine, but if one averages the excretion of sugar in the urines of these two days with the two days at the beginning of the fast, the result is encouraging. The second fast is a day shorter than the first, and the second period of feeding a day longer, until by the fourth period of fasting the patient goes without food only one day and then is given food for four days. This schedule need not be Followed exactly, but the general plan has proved most efficacious. The best example which I have of the advantage of intermittent fasting is Case Xo. 1190, that portion of Table 100 from December IS forward described in full under Acidosis. Case Xo. 705, Table 140, illustrates this also, but the administration of levulose, as in the earlier part of Case Xo. 1190, complicates the picture. The use of levulose in these and other cases is discussed on page 37X. Case Xo. 1M , recorded on page 492, was the first case so treated. ( 'aseNo. 1207, Table 140, shows clearly my present plan of treatment. Few if any uncomplicated cases of diabetes exist which will not become sugar- free with treatment similar to that of ( 'asc Xo. 70."). During the DIETETIC TREATMENT 315 whole period, from the first clay to the thirty-first day, she was up and about, and upon the third day of fasting enjoyed "a matinee. I pon the remaining days she often walked several miles a day, kept in good spirits, and in connection with one of the instructors of TABLE 145. INTERMITTENT FASTING. CASE No. 765. ONSET AT 21.8 YEARS; TREATMENT AT 22.4 YEARS. Urine. Diet in grams. "5 u A . , t_ c Date, 191(5. a >> 3 -I c bt C rt u o 1 p **> "* bD 'fi " ti (H V ^: 1 12 ti u V a c Diaoetie acid. J2 - >rz c | ^ ~ S, C3 A-S b c . *c o o Z _3 M - d * R. Q I _o ~5. a 5 7 ^ +* 3 K c ~ =3"~* o ^ "c; -5 " J2 ~ ^ ~ ^ N o y, < . I Us state at entrance was far more serious and without the experience acquired with other cases coupled with other advances in treatment, so far as the Deaconess Hospital is concerned, he would not be alive today. TABLE 140.- -IXTERMITTEXT FASTING. CASK Xo. 1267. ONSET AT 38 YEARS. TREATMENT AT 43 YEARS. Date, 1917. Urine. Diet in gnims. Carb. bal- ance. Blood sugar, per rent. Alveolar air CO 2 , mm. Ilg. T i Di- x -: isr * I Us, Sugar us. t-'nis. Carb. Prot. Fat. Aleli. Cal. Mar. 19-20 I MM) + + + 59 _ 20-2 1 IMK) -f + + 2.2 52 50 8 232 _2 . 29 20 2122 1000 + + + 1 . 7 32 30 8 152 > 26 22-23 2100 + + 1 (i 15 8 92 -1 . 23 25 23-24 1300 + 1 . 9 5 00 5 27 24-25 1 100 + 25-20 1100 + 4 35 10 230 1 20- 27 1700 + 2 .2 9 35 10 : 230 -9 27-2,s 1000 SI. + 1 -1 0.22 2s 2'.) !)()() + 29-30 2100 ' SI. + .'50 120 30-31 1700 SI. + 3 44 32 4(54 -3 31- 1 1500 ; o Apr. 1- 2 1200 si. + o 34 ' 190 2- 3 2000 32 20 308 3- 4 2500 1 .5 32 38 2S 073 25 4- f) 2000 38 57 ; 2S SO I 0.1, s 5 n 1800 1 45 i 03 28 947 +1 20-21 2200 3 40 79 2S 1103 +3 21-22 2100 r> 47 79 28 1115 +5 22-23 2100 6 47 77 28 1101 +G 2:-! 24 2000 7 48 88 19 1155 +7 24-25 2100 8 48 88 1124 +8 25-20 2200 10 49 ss 109S + 10 20-27 2500 31 57 001 +0 0.13 Weight March 19, 1917, 117 pounds; weight April 27, 1917, 110J pounds. Intermittent fasting not only makes the treatment less arduous and decidedly pleasantrr for the patient, but, it gives less cause for anxiety. The circumstances which suggest themselves at first thought as favoring an alternation of fasting and feeding periods are: the avoidance of loss of weight and loss of body protein, and the prevention of acidosis. These topics, ho\vever, are of such broad scope that they \\ill be considered following the remarks upon the details of treatment. DIETETIC TREATMENT 317 ((/) Determination of Tolerance for Carbohydrate. When the twenty-four-hour quantity of urine is sugar-free, one can usually give a few grams of carbohydrate to the patient without the appear- ance of glycosuria. The carbohydrate is generally given in the from of 5 per cent, vegetables, choosing those which are especially bulky. A plateful of lettuce appeals much more to the patient than a small saucer of string beans. When a mixture of 5 per cent, vegetables is given, one can be quite sure that the average content of carbohydrate is not more than 3 per cent. 1 or approxi- mately 5 grams for the 150 grams prescribed, and for convenience this is reckoned as 1 gram of carbohydrate for each 30 grams Cl ounce). This small amount of food, of course, has little nutritive value, but is enough to break the fast. Upon succeeding days 5, 10 or even more grams of carbohydrate, varying with the severity of the case, are added daily until sugar returns or the approximate quantity is reached which it appears probable the patient will tolerate. It should be borne in mind that a patient fasting or on a very low diet often shows an apparent tolerance for carbohydrate far in excess of that which he would have shown if the necessary protein and fat in his diet were simultaneously ingested. A carbohydrate tolerance test, like a levulose test, often destroys the morale of the patient. He cannot understand why he can take so much at one time and so little at another. Therefore I believe it more practicable to work up to about the carbohydrate value which appears correct rather than to go to the very limit of tolerance. Following the trial with 5 per cent, vegetables, one can proceed to the 10 per cent, group, and these can be empirically reckoned as containing per cent, carbohydrate or approximately twice that of the 5 per cent, group, or 5 grams carbohydrate for 75 grams vegetables. From this point onward the addition of carbohydrate can be made according to the desire of the patient. Half a small grapefruit contains about 10 grams carbohydrate, and half a dozen oysters 4 grams, and they make welcome additions. A portion of carbohydrate can be given in the form of cream, but I do not like to allow much cream or other attractive food contain- ing carbohydrate until it is evident that the patient will be able to take it permanently. For this reason it is often best to gradually replace a large portion of the 5 per cent, vegetables with 10 per cent, vegetables in order to get additional carbohydrate, and if these are borne, continue to add carbohydrate as cream, fruit in the form of grapefruit, strawberries, or orange, until a small portion of fruit is allowed twice a day, and then to progress to peas in the 15 per cent, group. A good many patients can take potato and 1 See p. 260. 318 TREATMENT oatmeal. Ten grams <. oatmeal give considerable satisfaction and contain but (i grams carbohydrate, and the same is true of oO grams of potato. One seldom allows the patient to increase the carbo- hydrate above )> grams per kilo body weight, even if it is borne, unless he remains under observation or can be thoroughly trusted to control further additions by the Benedict test. Indeed, most patients with mild diabetes feel better restricting the quantity of carbohydrate than eating freely of it. A few cases can bear more, but until the urine has remained sugar-free for three months, it is best not to attempt an increase. On the other hand, the carbohydrate should not be kept too low. Is there not a possi- bility that if one restricts the carbohydrate for a long period of time, there may be a loss in power to utilize it simply because those functions of the body which are concerned in its digestion and assimilation have not been kept continually in training? The more carbohydrate a diabetic patient can take without glycosuria the more likelihood there is of increasing the carbohydrate tolerance still further. It looks as if it were worth while to sacrifice almost anything to secure a tolerance for carbohydrate; with this once attained, strength and weight follow as a matter of course. With children one often makes the mistake of increasing the carbohydrate too rapidly, forgetting the fact that ."> grains of carbo- hydrate to a child weighing 20 kilograms is in the same proportion as 1") grams of carbohydrate to an individual of GO kilograms. (c) Determination of Tolerance for Protein. So soon as the urine has been sugar-free for two days, about -(} grams of protein are added to the diet, and thereafter lo grams of protein daily until the patient is receiving 1 gram per kilogram body weight. The protein may be given either in the form of fish, lean meat or eggs. Thirty grains of fish or an egg of average size contain approx- imately G grains of protein, and ))() grams of lean meat contain approximately X grams. The white of an egg contains '.} grains of protein. By this arrangement a patient weighing GO kilograms would be taking, within six days from the time he became sugar- free, 1 gram of protein per kilogram body weight. This quantity is quite satisfying to all except children in fact, it astonishes me to find how few patients care to take as much as 1 .."> grams of protein per kilogram body weight. Children, however, crave and need consider- ably more, and indeed take with avidity as much as '2 or M grams protein per kilogram body weight. Fish is especially desirable in the early days of protein feeding because it contains so little fat. Cod, haddock and flounder, for example, contain less than 1 per cent. The advantage of giving and increasing protein simultaneously with the determination of the carbohvdrate tolerance is that one DIETETIC TREATMENT 319 approaches more nearly normal conditions. What the physician is after is to determine the carbohydrate tolerance while the patient is on a full diet, and not the tolerance for carbohydrate alone. One must not forget Klemperer's observation of the fasting diabetic who could consume nearly 100 grams of glucose with little or trifling glycosuria, or Benedict's and my cases, who have taken a gram and a half of levulose per kilogram body weight, fasting, and excreted less than 5 grams of glucose during the same twenty-four hours. A determination of tolerance for carbohydrate of that type is of little practical value, and indeed is misleading. On the other hand, 1 freely admit that a higher carbohydrate tolerance can be attained when the addition of protein following the preliminary fasting is deferred until the actual carbohydrate tolerance is learned in the absence of protein and fat. Naturally the method adopted will vary somewhat with each patient. There are very few patients who will not bear at the outset as much as 1 gram of protein per kilogram body weight, and I am very loath to allow the protein to remain permanently below this figure. This can be avoided by still further restricting the carbohydrate, either temporarily or permanently. It is always necessary to bear in mind that one food which the diabetic patient cannot do without is protein, and to it everything else must be subordinated. More and more I believe we will strive to spare body protein. Whether it is desirable for diabetic patients to take more than 1 gram of protein per kilogram body weight, /. e., the Chittenden standard, is still unsettled. The 120 grams of protein which the Voit standard places as normal for a man of 70 kilos would be reduced to 100 grams (actually 103) protein for an individual of 60 kilos body weight. A protein intake of 100 grams is a trifle higher than what Cannon found for a group of students at the Harvard Medical School, but is the protein figure now usually recognized as the American standard. It would seem, therefore, that as much as this quantity of protein should be allowed the diabetic whose diet is so much curtailed, and this would raise up to 1.5 grams the amount of protein allowed per kilogram body weight. But I would rather reduce the protein from 100 to 75 grams if thereby I could raise the carbohydrate from 15 to a total of 40 or even o() grains. In our radical rearrangement of the diet, we must take care not to give the patient too little food. A normal individual has his instinct and taste to control the quantity of his diet, and it is a serious matter to attempt to replace such controls of nature with arbitrary rules. Low quantities of protein suffice for nephritic patients for periods extending over many years. If nephritic complications are present with the diabetic, one is therefore justified in prescribing a low protein diet. Except for patients of this character, I do not recall 320 TREATMENT any diabetic patient who has been obliged to live permanently ou a protein intake of less than 1 gram per kilogram body weight. (/) Determination of Tolerance for Fat. A satisfactory test fo: the tolerance of fat by a diabetic patient as yet is not available. I am in hopes that the work of Professor Bloor and Dr. Gray here and that at various other laboratories will soon give us some indi- cation. At the annual meeting of the American Medical Association this year Bloor, (tray and I plan to present a summary of work in this direction, based upon my patients. For the time being we are compelled to fall b::ck upon indirect methods, of which there are two: signs of acidosis and glycosuria. So long as these exist tin fat must be kept low. \Miile testing the protein tolerance a small quantity of fat is included if, in addition to whites of eggs and lean fish, meat is given. Formerly I thought this advantageous, and such small quantities of fat certainly do no harm in the milder cases. In fact, the same rule holds for the testing of the carbohydrate and protein tolerance in the presence of fat as has been said f, p. .'514, who after persistent periods of intermittent fasting became sugar- and acid-free, and yet the immediate addition of .'50 grams fat per day failed to cause the reappearance of acidosis. I consider the cr.se exceptional in this regard, and can only explain it by the fact that for over a year, except, for temporary periods, the diet had been low in carbohydrate and that she had become accustomed to considerably larger quantities of fat than would have been the case with an individual whose diet was suddenly changed. ( 'ases Nos. liM.'! and \'2'.V.\ should be cited, who, though having a carbohydrate tolerance of less than 10 grains, took respectively 00 and 7.~> grams levulose and 00 and 7.~i grams olive oil in the absence of protein wit IK nit t he appearance of acidosis or of more than 1 gram sugar in the urine. Cases which have DIETETIC TREAT ME XT 321 shown little acidosis may easily he allowed an increase of 25 grams fat daily, and for such cases this is desirable because it rapidly brings the total caloric value of the diet up to a normal figure. Naturally, patients in whose treatment a loss of weight is desired would ho given smaller quantities of fat. How inadequate our measures of fat tolerance are is occasionally shown by a patient like Case Xo. 1228, whose acidosis disappeared as fat was increased. (g) The Caloric Needs of the Patient. The total number of calories which a diabetic requires varies not only with each case, hut varies with each case each day. Schematic rules do not hold. One must remember that an individual trained to he quiet and lying down can get along with only 20 calories per kilogram body weight reckoned per twenty-four hours, at the Nutrition Laboratory whereas the average of a large group of normal men and women, not especially trained for the test, consumed 25 calories per kilogram body weight reckoned also per twenty-four hours. 1 If this variation exists while at rest, how much more it must exist during the various activities of different individuals. Furthermore, one must remember that the number of calories consumed per hour varies enormously. During sleep, Benedict's fasting man burned 1 calorie per kilogram body weight per hour the day before the beginning of the experiment, while during the daytime and awake although still at rest, the number of calories rose to 1.05. During the fast while asleep the metabolism on the same basis dropped to O.S5 calorie, and when awake to 0.05 calorie. In the previous section it has been pointed out that an individual weighing GO kilos walking at the rate of four miles per hour would require an additional 19.'} calories for that hour over the resting metab- olism. Habits of individuals vary widely. Some are quiet "and some are active. All these considerations should be clearly borne in mind by doctors and patients in order not to allow them- selves to be held too rigidly by any caloric fetish. Patients coming for treatment with severe acidosis consume from 10 to 20 per cent, more calories per kilogram body weight than patients after they have become sugar-free and free from acid. One of the most interesting cases of this group is that of Case No. 7(>5, p. 314. This topic has been discussed in detail in the section on Metabolism, and a forthcoming monograph by Benedict and myself will appear as a Carnegie Publication. After the diabetic has become sugar- and acid-free, he apparently gets along, as Naunyn long ago pointed out, with a smaller amount of food than an ordinary individual. 1 The error should not be made of concluding that an energy loss of 25 calories, per kilcgram body weight in the fasting state can be replaced by a diet with the corresponding number of calories. In reality a considerably greater number of calories, would be required, due to the specific dynamic action of the food ingested. 21 322 TREATMENT This may be so. From a study of dietary charts in diabetes, it appears probable, but 1 do not believe the question yet settled. Before this can be done, studies should be made upon cases of diabetes of the severest type who have become sugar- and acid-free and remained so for a period of weeks. The caloric values of their diets should then be quantitatively determined. Ordinary calcula- tions, ba^ed on dietary tables, will not suffice. Actual analyses of the diets of a group of such patients for a period of several weeks should be made. This is by no means an impracticable matter, and I think the accomplishment of it would be of the greatest help to diabetic patients and would settle many mooted points. (1) Lout Caloric*. When Dr. Benedict and 1 recently compared the actual metabolism of diabetic patients, as determined by the Car- negie Respiration Chamber, with the calculated calories of the diet, it was perfectly plain that errors in the dietetic calculations must exist, and that the patients must be receiving far more food than was supposed. A search was immediately instituted for these lost calories. Very fe\v calories could escape notice in the form of carbohydrate, because the patients did not receive carbohydrate. Only a few calories could escape notice in the form of protein, because our calculations have shown that the nitrogen excretion and the calculated nitrogen content of the food were consistent. r l he only essential error I found in the protein calculation was in the case of one patient who was taking 10 grains in the form of gelatin each day instead of the customary '2 or .'> grams, which patients not upon a fasting diet frequently receive without record in the dietetic calculations. The protein in broths is not so easily disregarded, particularly in the case of young children, for whom a few grams of protein count much. By exclusion, therefore, the lost calories must exist in the form of fat. In previous dietetic deter- minations, the quantity of fat in an egg had been reckoned at o grams, but probably (> grains is a more correct figure and the one we now adopt, and if the eggs are large the figure would rise still higher. It seems probable that upon a diet containing three eggs a patient received :>() calories more than previously supposed. Similarly, with fat in the meat, actual analyses showed that even a mixture of lean meat, as served by nurses, sometimes contains approximately !"> per cent, of fat, and as most of the patients were receiving from 100 to l.">0 grains of meat, -!."> to 70 calorics would be gained in t his manner. These errors have been in a measure obviated for this last year by the correction of the values assigned to foods on the dietary charts. The fads are mentioned simply to show that patients very likely receive more calories than lias been supposed, and illustrate the necessity of accurate analyses of the total diets of diabetic patients before any statements have been made about the total quantity of calories which they require. For the present DIETETIC TREATMENT 323 the physician must to a large extent be guided by the appearance and weight of his patient and the condition of the urine while attempting to regulate the calories allowed. Should the calories be raised above a minimum portion in severe cases of diabetes, glycosuria will return. Therefore great care must be taken to prevent overeating. This is not the only danger. If the carbohydrate tolerance is lo\v and fat is too rapidly increased, acidosis and death follow. Hence the need for supervision after the patient has left the hospital. More deaths in diabetes must be attributed to this cause than to any other. (/?) Reappearance of Sugar. The return of sugar demands fasting for twenty-four hours, or until sugar-free. This rule should be inflexibly followed, especially with children. In hospitals it simpli- fies the treatment enormously. So soon as it is understood that the reappearance of sugar means a fast until glycosuria disappears from the twenty-four-hour quantity of urine, there is little tendency to break over the diet. Furthermore, most patients are thrifty enough to see the disadvantage of paying their board with no return. The rule must be rigidly enforced with children, because with them disobedience means death. When a patient has been made sugar- free by a preliminary fast, absence of food for twenty-four hours will almost invariably be sufficient to free the urine at once if the sugar returns. This will not be the case unless the presence of glucose is promptly detected, and hence the necessity for the patient t > examine his twenty-four-hour urine daily. Following this accessory fasting day, the previous diet of the patient may be gradually resumed, making every endeavor to regain the former tolerance for carbohydrate by slowly increasing the quantity of fat. (ireat care should be exercised, more indeed than I have often taken, not to break down the tolerance a second time. No heavier burden in this regard should be laid upon the patient than he can easily bear. Months rather than weeks should intervene before the final amounts of carbohydrate, protein and fat reached the second time, equal the quantity of carbohydrate taken when sugar reappeared. I have always been much impressed by the success of Drs. Janeway and Mosenthal in the treatment of one of their patients, because the patient had been taught to keep the carbohydrate so low that sugar did not reappear, though he was away from their supervision for a period of months. (/) Weekly Fast Days. The advantage which the older clinician derived from the use of one day's fast in seven in the treatment of his diabetic patients should ever be borne in mind. Case Xo. 1002, now under observation, who contracted diabetes twenty-six years ago, possibly in connection with gall-stones, tells me that at that period her physician, Dr. Randall, of Topsh'eld, Mass.. often told her to go without food save broths for several days in 324 TREATMENT succession, and that she would follow this advice. Her severe symptoms of diabetes subsided at the end of four years. Recently the quantity of sugar has been slight. Her tolerance on June 1, 1010, reached 11(> grains carbohydrate. One striking characteristic of Dr. Allen's helpful suggestions in the treatment of diabetes has been that he has apparently sought out every method which anyone has found useful in the treatment of diabetes, tested its worth, and endeavored to adapt it to modern conditions. The advantage of this restricted diet day each week is partly inherent in the fast or restricted diet, but to a considerable extent it is due to the attention of the patient being sharply called t;> his disease one day in seven, and the recollection which it awakens in his mind of his condition before treatment began and the difficulties which originally accom- panied becoming sugar-free. Any patient with a tolerance for less than 20 grams carbohydrate should fast one day in seven; when the tolerance is above 20 grains carbohydrate, upon the weekly fast day halve the calories. Some exceptions to the above rules may be mentioned; for example, elderly patients bear fasting poorly, and when they remain sugar-free upon a rigid diet containing only 10 grams of carbohydrate, it is my impression that it is better to simply restrict the calories of the diet one-half on one day each week rather than to institute an absolute fast. Children become fretful upon a fast day, but if they are allowed a few green vegetables in addition to broths, they get along very well indeed. Yon Xoorden pointed out that the good effects of a fast day continued many days beyond the actual fast. Several of my severest cases, whose metabolism has been jointly studied with the stall' of the Nutrition Laboratory, were able to take upon their weekly fasting day (>() to 7f> grams levulose and equal quantities of olive oil with the result that the sugar in the blood on the following morning was the same or lower than before the above were given, and the total quantity of sugar in the urine during the corresponding twenty-four hours less than 1 gram. Consider what this would mean to diabetic patients if it would hold true in every case. The 7SO and ( .)7."> calories thereby saved each week would amount to a gain of between 40,000 and .">(), 000 calories in the course of one year. (See p. 37X.) "). The Management of Mild Cases of Diabetes. The mild case of diabetes is the case which demands the most energetic treatment, but hitherto has received the least. These cases are analogous to the cases of incipient tuberculosis. As in tuberculosis, a "cure" may not be effected, but the disease is held in check. Emphasis should be placed on freedom from glycosuria. Xaunyn's dictum that "many a severe case was originally mild, but neg- lected," should not be forgotten, DIETETIC TREATMENT 325 These patients should be taught to take long vacations, secure an abundance of sleep, avoid excess in mental and physical labor, shun obesity, and provide for daily exercise. It is best to be frank with such patients, and warn them of the danger of neglecting treatment. It is not necessary for such patients to practise fasting. Formerly the reduction of carbohydrates to 100 grams would often suffice to stop the glycosuria. In our enthusiasm for new methods it should not be forgotten that even in the past good results were obtained with the majority of diabetics, and that gradual restriction of carbohydrate was the means employed. Incidentally, this is good proof that most diabetics are not severe. The simple omission of fat will lead to a great reduction in the sugar excreted, which will be accentuated if the protein is limited to a gram and a half per kilogram body weight. Combined with the above the exclusion of actual sugar from the dietary and the substitution of potato (20 per cent, carbohydrate) for bread (GO per cent, carbohydrate) will frequently secure a sugar-free urine. It is onerous for a patient to eat 100 grams starch in the form of potato, but enjoyable to consume it as bread. Later, fat is added as needed to maintain a suitable weight. This method of treatment is very successful with my good-natured and usually obedient, fat doctor-patients. The milder cases of diabetes in a few weeks attain a tolerance of more than 100 grams carbohydrates. Such cases should limit the quantity of carbohydrates in their diet for years, even though no signs of sugar appear. It is interesting to note that many of these patients like Case No. 1287 feel better and have less digestive trouble if the quantity of carbohydrates is held at about 125 grams to 150 grams even though they tolerate more. Such indi- viduals live apparently in perfect health, and there is always satisfaction in the belief, and I think justification for it, that treat- ment has prevented the progress of the disease. Mention of these cases is made in the discussion of the Cases of Fifteen or More Years' Duration, page 4(58. G. The Management of Severe Cases of Diabetes. It would be wrong to give the impression that the treatment of severe diabetes is simple and free from anxiety. It is true that it is much simpler and causes infinitely less worry to the physician than heretofore, but these patients are in most unstable equilibrium and a little upset of trivial character may lead to much danger. The physician who treats severe diabetes successfully must constantly be in close touch with his patient. Forewarned, forearmed. I would rather have one glance at a severe case of diabetes twice or possibly three times a day than to be furnished the laboratory report of any test for acidosis. Of course, both observation and report should go hand in hand and invariably I try to provide for both. 320 THE ATM EXT Severe eases of diabetes cause the most worry at the beginning of treatment. A gastrointestinal upset, a careless alteration of diet, anxiety, excitement, overexertion, a mild intercurrent disease or an infection may favor the outbreak of coma. Yet Case No. 10.1 went through a severe attack of mumps at the height of his severe diabetes successfully; Cases Nos. S, It), M, 4(i, 35S, 435, 895 and 1274 survived pneumonia, Table 174, and Case No. SI 3 withstood tonsil- litis. Nevertheless it remains true that in the presence of an infec- tion the tolerance for carbohydrate is apt to suffer and unless precautionary measures are taken coma may supervene. Case No. 705, described in detail under Intermittent Fasting (page 413), entered the hospital on December (>, 1915, with 3..'] grams of ammonia in the first twelve hours. Following drainage of an alveolar abscess, which had made its way deeply into the jaw, the acidosis decreased more rapidly than would have been expected with ordinary treatment. A somewhat similar experience has been observed with Case No. 951, who had pyorrhca. The severity of the diabetes of Case No. 1239 greatly decreased when an abscess of his lung was drained following the excision of portions of four ribs. (See p. 443.) Therefore, all such contingencies should be avoided, or at least anticipated. It is because 1 have seen so many cases of this kind in the past die, as 1 now know needlessly, that I lay so much stress upon the preparatory treatment of such patients before fasting, for such a plan of procedure will tend to lessen the dangers to which these cases are exposed in the presence of any emergency. For these severe cases a trained diabetic nurse permanently in charge of the patient, is of the greatest assistance to the patient and family. My nurses tell me that it is seldom wise for them to remain longer than three' months with any given patient. They prefer to trade cases with their friends and this plan works out happily for all concerned. If a nurse is not available, success in treatment depends upon the thorough education of the patient. The wise live long; the ignorant early succumb. Once having thoroughly mastered the essentials of treatment, these patients need little supervision. They manage themselves quite well and frequently give useful suggestions for their own treatment and that, of others; 1 can truthfully write that their chance for improvement is dis- tinctly greater today than one year ago. Case No. 032, p. 3*2, has faithfully carried out treatment and his course is an encouragement to all. Case No. 119(1, described in detail on page .'579, is as good an illustration of the treatment of a severe cast' as Dr. Ilornor and I have had. lie was later put under the charge of Dr. I>righam, who has clinched the improvement gained. The following sentences upon the behavior of severe cases of FOLLOW-UP METHODS 327 diabetes were written by me before the introduction of fasting treatment, but the essential features are true today. " I have never seen a diabetic patient do well who had wholly given up diet, and 1 confess that I have had considerable opportunity to observe such cases. The family, the friends, the patient, and the doctor will all get along more comfortably if the diet throughout the whole course of the disease is kept at a point where, in the first place, sugar remains absent from the urine (but failing this, at a point so strict that any greater restriction might be productive of coma). 1 The attitude of the patients themselves under the two conditions is very different. In the former, we have the weak, complaining melancholic, tearful patient, Avith complications of the nerves, eyes, or skin almost invariably with us. With the strict, dietetic treatment, these complications, no matter what their original cause, are far less numerous, and I have frequently been impressed by the freedom from despondency of those who live faithfully by it. The urines of the two groups of patients may show no essential differ- ence in the total quantity of sugar, and the urine of the strictly dieted case may show more acid; yet the general well-being of the strictly dieted patient is far better than that of his careless brother." D. FOLLOW-UP METHODS. Almost any physician can get his diabetic patients sugar- free, but to keep these patients sugar-free and in good health is an art. A doctor's ambition should be not to see how large a number of diabetic patients he can have, but how large a number he can keep well. The education of the patient is the basis for success along this line, and this must begin at the very start. Upon finding that the mortality among my patients during the last twelve months was largely due to ignorance, a nurse was secured to help in remedying this defect. She began work May 1, with five months' handicap on this statistical year, and I anticipate much from her efforts. Why should not all our patients have the advantage of social service, such as Dr. Brigham's diabetics in the Out-patient Department of the Massachusetts General Hospital enjoy? Hospital treatment should be beneficial, if for this reason alone. The patient comes in close contact with the doctor, his questions can and should be answered, and he has an opportunity to learn how to live. The presence of other diabetics in the hospital is also salutary, and it does no harm for him to see cases far more severe than his own, because this emphasizes the importance of early treatment. So far as is known today it is possible for patients to keep sugar-free 1 Of course this idea is now recognized as incorrect. 32S TREATMENT and maintain strength by means of fasting days and a diet with or without carbohydrate, but with a moderate quantity of protein and fat. These are the principles which must be instilled into the patient. Patients must be taught from the first that the preliminary treatment represents only a beginning of treatment; that treatment lasts for life; that whenever sugar or acidosis returns or unusual loss of weight occurs and they are unable to control these phenomena, they must report to a physician. The physician, too, must do every- tii'ng in his power to keep in touch with his patients at regular intervals. Of course, this is quite difficult because he is laid open to the suspicion of seeking practice; but it is better to run the chance of a misunderstanding on this score in one case than to let ten other cases perish. One of the (thief advantages of modern treatment is that it is less costly than former methods. Frequently all visits and all urinary examinations are avoided, and this fact itself makes it much easier for the physician to deal with his patients. Doctors who have patients temporarily under their care in hos- pitals o\ve it to their patients to keep the family physician in touch with the method of treatment employed. Formerly I sent a copy of the chart at the end of the hospital stay; beginning April 1, I hope to send weekly reports, written out, if possible, by the patients themselves. For this purpose the type of chart on page .'$()(> will be better than I have hitherto used. Once I sent a patient with exophthalmic goitre to the Mayo Clinic and received four unsolicited reports upon the progress of her successful treatment. Only too often by former methods, and 1 am forced to say by methods employed by me so recently as last year, a patient would go down hill soon after leaving the hospital, because of imperfect training as to how to live outside. It is unquestionably true that when glycosuria is abolished and strength diminished by long- fasting, and then glycosuria and acidosis allowed to return through improper diet, the last stage of that patient may be- worse than the first. Two methods are open to guard against this eventuality: (1) the education of the patient in the treatment of diabetes, and ('2) the education of his physician. E. SUMMARIES OF THE TREATMENT EMPLOYED IN THREE SUCCESSIVE GROUPS OF CASES. The treatment of diabetes is changing so rapidly that it is diffi- cult to grasp its essential features from a study of individual cases. Consequently, in the following tables 1 have endeavored to summarize the treatment of three groups of cases of diabetes, each consisting of '50 cases, seen in chronological periods during the last two years. (Tables 1-17, 1 IS and 11!).) For the sake of SUMMARIES OF THE TREATMENT EMPLOYED 329 comparison, further summaries of eaeli of the three tables are given in Table 150. It will be noted that in the tables the date of entrance to the hospital, the duration of the stay, the weight at entrance, the lowest weight reached, and the weight at exit are given. In order to simplify, yet show the general plan of diet, its composition during the first and last weeks of hospital stay is given and, for comparison between the patients, calculated per kilogram body weight. The duration of the diabetes gives some idea of the character of the cases, and this is recorded to December 1, 1910, and by noting the date of entrance it is an easy matter to learn the age of the patient on entering the hospital. Finally, the preliminary method of treat- ment is indicated, namely, whether it commenced with fasting or with preparatory treatment, that is, diminution of fat and protein prior to the fast, though often by this plan the patients became sugar-free and actual fasting was not used at all. In the last three columns the question of acidosis is considered. In the early cases but few showed its absence throughout, either developing it in the hospital or entering with it. A gratifying decrease in the number of cases who developed acidosis is evident. The criteria for the selection of cases in the table were (1) that the patients should have been over fifteen years of age, and (2) should have remained at least two weeks in the hospital. All cases recorded in these tables left the hospital alive except case No. 1)04, who died with a carbuncle. Eleven others succumbed within eight months of discharge. These cases are indicated. The duration of the fatal cases is recorded to date of death instead of to December 1, 1910. It is hoped that the introduction of this method of presenting results of treatment will prove a valuable innovation in our management of diabetes. Undertaken with little conception as to what the average results would be, it shows quite clearly the errors in the past and indicates where the weak places in modern diabetic treatment lie. 1. Loss of Body Weight. Considerable obscurity exists about the changes in body weight which take place in fasting. If one looks at the table on p. 150 for the healthy man who fasted for thirty-one days at the Nutrition Laboratory, he is struck by the uninterrupted loss of weight each successive day. No such consistency is noted in the weights of diabetics while fasting. Undoubtedly this is in some way connected with the salt metabolism, and very likely is related to the acidosis as well. (See pp. 129 and 130.) At any rate, a gain in weight of a few pounds, so far as the patient's symptoms are concerned, is a favorable indication, and a loss of weight of over a few pounds is distinctly unfavorable. It is wholesome education for the diabetic patient during fasting to see that gain and loss of weight are in no way related to his previous diet. The question of ~.'.^ht 330 TREATMENT upjcuj SUMMARIES OF THE TREATMENT EMPLOYED 331 U + : : : H- :++ : : UldJOJJ L" S ^ ~ O M rtOiNOOOOOOOOOOXOOOOOOOi -f ^) "/: c~. o i O < t- 1- cc ^ S3 g ! Q' 3 | C"a o ~J:~ " I II c; . ilil ?i R fo ?3 ri ~5 re o ''- '^ i- -3 '-: 1~ ir -j: --D x f- x x x OOOMOOOOOOOOOOOOOOOOO1>OSBOOO 332 TREATMENT : + + : : ! - >~ v. . < ~ T ' t^ ~ x c; ~. i^ 'M -1- ~\ -r :^ f. >~ '" > c re -H o O i.~ -r i- x '-: - t^ -r -^ ^ 'O re 1.7 [^ c~. -f ^ - c X w -* ^2 1.7 10 1.~ i SUMMARIES OF THE TREATMENT EMPLOYED TABLE 150. SUMMARY OF TABLES 147, 14S AND 149, SHOWING THE TREATMENT OF THREE GROUPS OF THIRTY CASES EACH OF J)IAHKTER A.I SUCCESSIVE I'EHIODS. | 3^nK^r CO o 71 3 o ii US 1 ri ; "S d " ^2 ^ - -f Cl Treatment begun with _o _. a 'S . X) s as "2 ^ a s c .5 t -o X ^ c -t 2 3 s ^ | .= ^5 2 s CO "a "3 "c EH 1 1 IoqoD IV *H * CO "^ g X o 'UIOJOJJ s s ^-0^3 J1 5 co | .- -5 sH -, c , -s r" C-'l i US CO loqoOIV C, - - W 3 - 2 nt.noaj s c-'i C-l -oqjBj 2 s Weight in kilograms. sso-j N X! o F3 5 3S , S o. tt o-! s CO 'OOUCJJUjJ g jr. s p;}i(Jsoq in =.'Ci:p -, - joqiunu aSujOAy ~' -x o "S 1- ~ -J !i4i >?a"r dnoiQ - - a 334 TREATMEXT changes in weight therefore need cause no anxiety. On the other hand, they obscure the picture, and it is difficult to tell what a patient's actual weight is as a result of fasting until he has been under treatment for several weeks, for of course these rapid changes in weight merely represent variations in the water balance of the body. To lose even 200 grains of body fat 1SOO calories must be set free. Edema occasionally develops, but so far I am able to assure my patients that I have seen no diabetic patient with edema develop coma, and have seen no serious result from edema. This can usually be dispelled with theocin or diuretin. A loss of weight in an emaciated diabetic should only be tolerated as a therapeutic necessity. Once lost in such an individual, it is difficult, but by no means impossible, to regain. (See Case Xo. 110(1, p. 379.) The mere word "fasting" awakens ideas of marked loss of weight during the course of modern diabetic treatment. This is by no means true, and my statistics enable me to disprove it. In Tables 147 to loO, the individual and average losses of weight of 00 patients above the age of fifteen years during a stay of over two weeks at the New England Deaconess or Corey Hill Hospitals can b.' seen. The cases are arranged chronologically according to date of entrance to the hospitals. In the first and second groups of 3D patients each, the losses amount to 3.07 kilos and 2.S7 kilos respec- tively, but it will be noticed in the last group of 30 cases treated th.it the loss was but !.<> kilograms per patient. These statistics were compiled for the first time during March, 1017, and they appear to me encouraging. With more care in the arrangement of the diet it would seem possible to prevent a considerable proportion of losses of weight which now occur. While in many cases such losses arc beneficial, they are not desirable in all, and each case deserves indi- vidual attention. Eat diabetic patients can most easily acquire a tolerance for carbohydrate by loss of weight. There is no necessity for this loss, being rapid. It should go hand in hand with gain in muscle. During the stay of the first group of patients treated, the loss of weight per day per patient was approximately 100 grams, in the second group Il2.~> grams, and in the last group f>7 grams per day per patient. If this loss of weight should arbitrarily be reckoned as one-half fat, it would be equivalent to 450, "><>:'> and l2.">7 calories respectively. Since the average weight of the patients was (iO, .10 and ,")(') kilograms, this would represent the equivalent of S, 0, and 5 calories per kilogram body weight per d.iy. Added interest attaches to these calculations when they are compared to the caloric intake' for the three groups of cases. As a matter of fact, if we arbitrarily consider the loss of body weight as due one-half to fat and add the equivalent of the same in calories to the caloric intake, it will be seen SUMMARIES OF THE TREATMENT EMPLOYED 335 that the calories furnished the patient (luring the three periods were not far apart (20, 30 and 28) good proof, though indirect, that the caloric expenditure under the varying conditions was also the same. These figures will be more carefully analyzed in the mono- graph with Benedict, and in that the data furnished by the indirect calorimetry of many of these patients will be available as well. 2. Loss of Body Protein. My desire to protect body protein has evidently attained some fulfilment, as shown by the records of the protein intake of the three groups of patients whose dietary data have been summarized. In the early group the protein per day per patient for the first week was but 17 grams, or less than f gram per kilogram body weight and during the last week it was 10 per cent, below 1 gram per kilogram body weight. In the two later series the figures change, and in the first week the protein almost reached -i- gram per kilogram body weight, and in the last week is fully 1 gram. A gram of protein per kilogram body weight appears to me to be the correct figure but I shall not feel satisfied until the patient receives this quantity on the average during his whole stay in the hospital. I can see no advantage in a loss of body protein, and feel that some means must be found which will improve the tolerance of diabetic patients without curtailing it. In the last few months I have used fish frequently in the early treatment of patients, because of the easy way it affords of giving protein apart from fat. I shall watch with interest the effect of treatment of cases along the general plan of the treatment of Case Xo. 1237 (see p. 30(5.) That case appears successful, but should be controlled by blood-sugar determinations and by many other cases. 3. Total Calories. The gain in calories of the third group of patients during the first week as compared with the first group has been of little importance, but there has been a steady gain in calories during the last week in both the second and third groups. At the present time the patients certainly go home in better condition and with lower blood sugars than two years ago. Still the melancholy fact remains that the patients leave the hospital with 23 calories instead of a permanent, self-sustaining diet. A gain in 3 or 4 calories per kilo would mean everything to diabetic patients. The shortage in calories now is in great part due to the weekly "fasting" day or "half" day, and the 21 and 23 calories respectively per kilo which the last group of patients daily received on leaving the hospital were chiefly taken by them during six days. If one could only obviate the fast day or half-day, the diabetic problem would be a good deal nearer solution. It is for this reason that physicians, nurses and patients at the New England Deaconess Hospital have watched with unusual interest the experiments with an exclusive carbohydrate (levulose) and fat diet on one day each week. The 3. l i, 370 cases of diabetes were seen, and )>S, or 10. !> per cent., died. There were also :>!) other deaths in this period occurring among the patients seen in the preceding eighteen years. The total deaths amount to 77 in all. The causes of death are shown in Table Ifd. The significant feature in the present report, is that during the year l!)l(i the percentage of deaths due to coma has dropped to 1 1 per cent. It will be recalled that a year ago, (11 percent, of all my deaths from diabetes were due to coma. 'This is a notable fact. It constitutes one of the best measures of the improvement which has taken place in treatment and shows that radical changes in the management of diabetes must have occurred. The second feature 1 of note is the- deerea.se in the percentage of deaths during the first year of the disease. \\ hereas at the Massachusetts (icneral I lospital up to IS! IS, ()7 per cent, of all patients who died of diabetes suc- cumbed during the first year of the disease, and in the subsequent period 10 per cent, so succumbed, and of my fatal cases up to a year THE FATAL CASES OF 1916 337 ago, 17 per cent, were fatal during the first year, I can now say that this mortality for all my fatal cases has fallen to 15 per cent, and for the fatalities of 1910, 44 per cent. In fact, of the 34 cases dying of coma only three deaths occured during the first year of the disease, and the possibility of preventing two of these deaths will be obvious to any candid observer. Reference to these will be made under discussion of the coma cases. The third feature is of a discouraging nature, for a new cause of death appears in my records namely, inanition, and to this 3 cases succumbed. Inanition, like coma, may be rightly considered at present as evidence of unsuccessful treat- ment, but even if it is said that these 3 cases were saved from coma to die of inanition and therefore should be counted as coma, these cases may be added to the coma cases and still the mortality of acidosis and inanition combined is but 48 per cent., in contrast to 04 per cent, from coma for the last eighteen years. TABLE 151. THE FATAL CASES OF DIABETES DUIIIXCJ 1916. l 1. Deaths without coma 43 I. Miscellaneous 9 Cirrhosis 1 Inanition 3 Diabetes 2 Unknown 3 II. Cancer 7 III. Tuberculosis .3 Pulmonary .... 1 Meiiingeal 1 Peritoneal 1 IV. Cardiorenal and vascular 14 Cardiac G Chronic nephritis 4 Cerebral hemorrhage 3 Enlarged prostate 1 V. Infections 10 Pneumonia . 4 Influenza 1 Appendicitis 2 Acute pancreatitis 1 Abscess lung 1 Gangrene 1 2. Deaths with coma , 34 Total 77 it is true that the cause of death was unknown in 3 cases and was attributed to diabetes a doubtful quantity in 2 cases, but these groups were represented in the former series, and even if all of these are added to the 37 deaths from coma and inanition com- bined, the mortality due to all these causes rises to but .14 per cent., a 1 Thirty-eight of these oases were seen during 1916 and thirty-nine were among patients seen in the preceding eighteen years. See Table 44 for the combined mortality of all cases. 22 338 TREATMENT figure sufficiently below 04 per cent, to he unimpeachable evidence of what has been claimed above. 1. Deaths without Coma. (//) Miscellaneous. Comment is un- necessary upon the 3 cases classed as unknown, and the 2 cases with a death certificate diabetes. On page 07 the infrequency of cirrhosis, from which there was one death this year, is discussed. The deaths from inanition demand the closest scrutiny. (1) Intutition. Three deaths from inanition have occurred in my series of cases since the inauguration of the new methods of treat- ment which had their start with Dr. Allen's articles upon fasting. No death was attributed to this cause among the many which occurred in the preceding seventeen years. The word "inanition" is used instead of the term "diabetic inanition," because the inani- tion does not appear to be distinctive of diabetes alone. Jt might occur in connection with other diseases, and, in fact, the patients do not appear like diabetics, the pain of cancer and the anxiety of fever are also lacking. But in each instance when I pushed the diet up above a few calories, sugar and acidosis appeared and experience with other diabetics made it evident that unless the diet was curtailed coma would ensue. Reference to the first of these cases (p. 08) has been made in the first section, but 1 append here brief summaries of two of the cases and their charts while at the hospital (Tables 1 ")2 and 153). My present feeling is that these deaths, too, are preventable, and I hope that some method of treatment, perhaps similar to that, described under the experiments with levulose, will enable such individuals to regain their tolerance. I cannot make myself believe that the tolerance' for carbohydrate is wholly lost when the urine is free from both sugar and acid, and the respiratory quotient shows that some material in the body other than protein and fat is being burned, even though the patient is fasting. (See pages 380 and 381 .) A little boy, ( 'ase Xo. 1120, aged seven years, with diabetes of nine months' duration, beginning with 7 per cent, of sugar, entered the New England Deaconess Hospital in September, 1 ( ,MO. lie had been thoughtfully treated in Vermont, but the urine had not remained sugar-free for more than two weeks at a time. One week before entrance his case became serious; nausea and deep respiration existed, and a strongly positive ferric chloride reaction appeared in the urine. lie became drowsy and dull, but not entirely unconscious, two days before entrance. He was fasted for the three days before entrance, and given one teaspoonful of sodium bicar- bonate three times a day. I pon arrival at the hospital his condition had improved, but the carbon dioxide in the alveolar air was 21 mm. mercury, and it fell to IN mm. the following morning. Then 1 was a slight trace of albumin in the urine and a few finely granular casts. With fasting, the omission of alkalis, and small quantities of carbohydrate and protein he became sugar- and acid-free, and later tolerated carbohydrate 3 grams, protein 20 grams, and fat THE FATAL CAXES OF 10 16 339 24 grams. Efforts to increase the diet without the return of sugar or acid failed; aeidosis and anasarea appeared. October 23, 1916, I increased the protein and fat of the diet, and aeidosis became extreme, deep breathing ensued, and the carbon dioxide in the alveolar air fell to 15 mm. mercury on repeated occasions; but again with fasting and without alkalis the aeidosis disappeared. From this time on he gradually failed, anasarea increased, and at length petechiye developed and he died without coma from exhaustion. TABLE 152. CASE Xo. 1120. DIABETES IN A CHILD OF SEVEN YEARS. DEATH FROM INANITION. Urine. Diet in grains. ^ 3 O Date, . -T = - a -[ r~ N % 1910 " Diacetic '5jf a > a C ~ 5 5- 1 ~ *"" H ^ : acid. o ~ ~3 ' b ~ ~ 'C ^ S+3 ^3 t. "o - ~ .ti M 53 | c i.' J. -if ^ s <~ Sept. 0- 7 . . . . -f- + + 7- 8 .... + 8 4 5 3 o 32 401 21 18 8- 9 .... + + 14 10 14 112 39' 9-10 .... ..17 15 15 120 39 26 10-1] . . .. + + 6 10 5 60 391 11-12 + 3 2 20 39 1 2 29 12-13 .... + + 5 3 32 39 13- 1 1 . . . . r. 3 20 6 j 146 38^ 26 14-15 .... 3 24 24 324 381 15-10'. ... + + 11 4 29 44 528 37! Oct. 10-11 . . 7.8 ++ 3.3 1.4 12 21 42 402 40J 11-12 . .8.8 ++ 3.9 1.9 lf> 10 21 3S : 466 40? 22 12-13 . .2.2 + 5.1 1.4 39? 0^32 13-14 .... | 4.5 1.1 3 o : o 39! 14-15 .... 4.8 4 40 20 15-10 .... 2.70.711 20 SO 42'. 10-17 .... 2.7 O.S 19 20 80 40i 17-18 .... 2.30.1 3 \ 41. ', 18-19 . 0.5 3.5 1 18 72 41 19-20 . . . . 4.1 0.2 23 15 227 40i . LO 20 21 .... 5.10.2 1 27 31 391 40! 2 1 22 .... . . . . 22 23 .. 20.. 6 20 45 .>17 421 42! 23-21. .2.7 4.9 . . 33 50 , 582 431 24-25 . . 2.7 ,4.9 . 5 33 50 582 43? 25-20'. .2.7 + 4.9 0.8 6 38 59 . 683 43? 20-27 . . 8.0 ' + + + 7.2 1.3 IS 42 70 798 27-28 . . 8.0 7.2 . . 25 20 17 12 256 j . . : 15 28-29 . .8.0 7.2 . . 7 24 29-30 . .5.3 6.4 4 25 30-31 . .5.3 6.4 48 1 20 Xov. 31- 1 . . 5.3 0.4 1 19 IS 242 49? 26 1- 2 . .2.4 ? 6.6 0.6 2 25 24 324 50 33 2- 3 . .2.4 ? 6.6 5 33 30 422 51i 3- 4 . .2.4 6.6 ! 5U 0.18 26 4- 5 . .2.7 8.8 5 33 30 500 51 1 24 5- (i . .2.7 8.8 5 41 33 477 50 25 0- 7 . .2.7 S.S 5 41 35 495 51? 13-U .... 6 51 34 534 49 14-15 .... 2 7 54 39 595 51 15 16 . .1.0 7.0 II 50? 10-17 . .1.0 7.0 o ! 51 34 534 50? 17 IS . l.o 7.0 1 6 51 34 531 53 l-i-19 .... ... S 6 41 35 499 52! 19 20 .... : 6.1 . . 1 r - ] l 0.28 20-21 , .j . . 0.1 1 27 23 319 21-22 . .' .. fi.i 1 1 35 28 390 II 51? 22-23 .... 1 1 35 28 390 23- 24 s . 1 12.5 10 39 33 493 24-25 . .10 12.5 15 43 37 505 25-20 . . 2.2 4.3 0.3 .. 30 12 22 1 1 Note influence of fat upon aeidosis. 2 Note terminal increase of urinary nitrogen without glycosuria. 340 TREAT ME XT Case Xo. 10S5, a frail woman of thirty-four years, with a history of diabetes of seven months' duration, entered the hospital in July, 1910, with a story of great difficulty in keeping sugar-free despite fasting and a lo\v diet. Tuberculosis was excluded by repeated physical examinations and by the fluoroscope. Her weight had fallen from 150 to 91 pounds, and in her stay of four months at the hospital dropped 30 pounds more. The urine remained free from sugar and acid nearly all the time except, when carbohydrate '2 grains, protein 40 grams, fat 90 grams and alcohol 15 grams were given for five successive days, when severe acidosis ap- peared and the sugar in the urine steadily increased. Levulose was given on three occasions in 50-, 50- and 75-gram doses, with the elimination of 1, 5 and 15 grams sugar respectively. Ten days after the second test the patient was just able to be taken home. She was sugar- and acid-free, but tolerated only 3 grams carbo- hydrate, 42 grams protein, 30 grams fat, and 30 grams alcohol (()()() calories). The blood sugar was 0. 10 per cent. A few days later, while remaining in the same condition, she died without pain or coma. TABLE 153. CASE No. 10cS"). DIAHETES IN A WOMAN OF THIRTY-FOIII Y i; AKS. DEATH FROM INANITION. Urine. Diet in II 1 ( ( 87 88 s I 87 0.28 7 ( 84 7 ( 84 41.2 0.35 8 XI 3 ( Si? > ( 82 X S3 1 1 I 82 :i ( 82 0.39 ( T'.t () s fifi s tlii 0. '2'2 tiS * : 71 ;| r 71 .-> (i!) .", I i! 1 :{ ;j 71 7 71 Such deaths 1 cannot believe necessary, \\ith the evidence of the high respiratory quotients obtained during fasting (p. 120), the specimen. THE FATAL CASES OF 1916 341 dextrose-nitrogen ratio suggesting that human diabetes is incomplete (p. 132), and in the presence of a normal quantity of sugar in the blood, a way should be found to make such deaths avoidable. (b) Cancer. Seven deaths occurred from cancer. None of these in any way could be brought into relation with the diabetes. (c) Tuberculosis. Many feared that tuberculosis would prove in- creasingly frequent with the undernutrition methods adopted these last two years. It is therefore significant that only three such deaths occurred : one due to tuberculosis of the lungs, another due to men- ingitis accompanying tuberculosis of the lungs, and another to peri- toneal tuberculosis. The rarity of tuberculosis in the presence of undernutrition will add largely to the conviction which has already been growing among those who treat consumptives that hyperfeed- ing should not of itself be an essential feature in the treatment of pulmonary tuberculosis (d) Cardiorenal and Vascular. It is really surprising that in dia- betes under the age of sixty one so seldom meets with evidence of valvular disease or failure of the heart. All recognize the fre- quency of arteriosclerosis in old diabetics, but the rarity of heart disease in young diabetics may not be simply a coincidence. Among my fatal cases of diabetes under the age of fifty, so far as I am aware, not one has died of a valvular disease of the heart. Of the 6 deaths from heart disease, 3 were due to angina pectoris. These were Case Xo. 355, whose lower extremities had been amputated on account of gangrene four years before (see p. 423); Case Xo. 499, who died at the age of sixty-two years, duration of diabetes seven years; Case Xo. 759, described in detail on p. 489. Case Xo. 1019 died of acute cardiac dilatation two days after simple drainage of a distended bladder, due to an enlarged prostate. As regards the four deaths from nephritis, I can repeat what is said elsewhere, that diabetic patients with the type of nephritis which occurs in advanced life really appear to do fully as well or even better than other cases with the nephritis alone, and Dr. Ilornor and I cannot but feel that the carefully controlled diabetic diet with its low caloric value and with these individuals low protein and salt, must contribute to this end. Three patients died of cerebral hemorrhage. (e) Infections. Infections claim a considerable mortality. Infec- tions undoubtedly lower carbohydrate tolerance and thus increase the severity of the disease. They often lead to a fatal issue in two ways: (1) by favoring the onset of coma, which will be discussed later, and (2) because local infections often deter the physician and surgeon from energetic treatment. Thus, of the tAvo deaths from appendicitis, one was a fulminating case, but Case Xo. 753 had an old history of attacks of appendicitis, and though he did not 342 TREATMENT die of a fulminating attack until five months after leaving the hospital, it is quite possible that had diabetes been absent the appendicitis would have earlier received adequate attention. Dia- betes must not be. considered a contra-indication to an operation for appendicitis. In this patient the success with which the teeth were extracted (p. 4 .'51 ) shows that the appendix could have been quite as easily removed. This was done in Case No. 29, twelve years ago, and the patient survives today. One patient died of abscess of the limy three months after leaving the hospital. Contrast this case with Case No. 12:>9, p. 44o, who was successfully operated upon for the same condition. Jt is not surprising that out of the 75 cases '.} succumbed to pneumonia at the ages of seventy-five, sixty-five and seventy-two years respectively. The death from acute pancreatitis is of course questionable, because there was no autopsy; but a good surgeon made this diagnosis when he saw the patient, who was a doctor of sixty-five years. 1 had not seen the patient for four years. Influenza caused one death. I consider that we were most fortunate at the hospital to have avoided deaths from influenza and bronchial infections, because several contact infections devel- oped during the year. Realizing the seriousness of such infections, vigorous isolation is now practised and with good results. Gangrene of the lower extremities was responsible for the death of one patient. He died with a temperature of 1()('> in another hospital, three days after operation, at which intraspinal novocain and local novocain were employed as anesthetics. lie was un- conscious but 'two and a quarter hours, and death may be attrib- uted to a continuation of the infection which had existed for four weeks before admission to the hospital. The diabetes was of ten years' duration, and I had not seen the patient for the last nine years. The man was reported to have kept sugar-free during these nine years, and in fact immediately after operation the urine was still free from sugar, but contained much albumin. Two days after operation, however, sugar was present in large quantity. 2. Deaths from Coma. More than once I have expressed the conviction that deaths from coma in diabetes are usually avoidable. The decrease in the percentage of all deaths due to coma from C>4 per cent, for the IS years, ending December 1, 1915, to 44 per cent, for the year 191C> alone and to (10 per cent, for the 19 year* ending December 1, 19 1C), together with an analysis of the present series of cases, supports this claim. In Table 154 I have analyzed these deaths from coma (1) according to the duration of the disease prior to the onset of coma, (2) to the best of my ability according to the condition predisposing to the development of coma. To my mind the conclusion is justified that although the THE FATAL CASES OF 1916 343 first year of the disease has been for years the clanger zone of the diabetic, modem treatment will now make it what it ought to be, the safest year of the life of the diabetic. For the present what was said relating to the first year as a danger zone should now be applied to the second year of the disease, and with this assurance, that just as the dangers in the first year of the disease have been already overcome, the way is clear by which to overcome also the dangers attending the second year of the diabetes. In compiling Table 154 I wrote down the cause for the death of each individual patient by coma absolutely without thought of the duration of the disease. The statistics are unbiased and the table shows well the opportunities for better treatment and also the possibility for the prevention of most of the now regrettable and unnecessary deaths from coma. TABLE 154. CAUSES UNDERLYING DEATHS FROM DIABETIC COMA DURING THE YEAR 1916. Duration of diabetes. Total cases. Dia- betes un-. treated. Xeglect of pre- paratory treat- ment. Obesity; high fat; low carbo- hydrate. Patient abandoned treatment. Imper- fect super- vision. Increas- ing sev- erity of diabetes. Ether anes- thesia. Car- buncle. 2- 6 mos. 3 1 1 7-12 " 1 1 1 -2 yrs. 10 1 2 7 2-3 " 5 1 2 2 3- 4 " 1 1 4- 5 " 4 2 1 1 5- 6 ' 3 1 2 6- 7 ' 2 1 1 7- 8 ' 2 1 1 12 ' i ; ' 2 i 1 1 1 Totals 34 Per cent. 10 1 1 6 21 The causes underlying the 34 deaths from coma during 191G were as follows: (a) Diabetes untreated. (b) Neglect of preparatory treatment. (c) Obesity; high fat, low carbohydrate. (d) Patient abandoned treatment. (e) Imperfect supervision. (/) Increasing severity of diabetes. (g~) Ether anesthesia. (h) Carbuncle, (i) Unknown. (a) and (b) Diabetes Untreated and Neglect of Preparatory Treatment. From the above table it will be seen that three deaths from coma occurred during the first six months of the disease. One of these was Case Xo. 1143, who had previously had no treatment for diabetes, so far as I am aware, entered the hospital in coma and died fifteen hours later. Another patient, Case Xo. 1004, had been placed upon the old-fashioned fat-protein diet, according to the story of her 344 TREATMENT physician, a few days before entrance, was brought to the hospital in an ambulance, in beginning coma, and died in fifty-six hours. Details regarding the death of the third case are not sufficiently clear to me to assign a cause. (() Obesity. High Fat and Low Carbohydrate Diet. Case Xo. 974 died toward the end of the first year of the disease. This patient represents one of three patients Cases Xos. 974, 733 and 1058 who died respectively nine days, nine months, and five months after leav- ing the hospital, in the first, second, and third year of their disease. These cases were not intrinsically severe cases, but I believe were so by two factors: the low quantity of carbohydrate in the diet and the comparatively large quantity of fat whether body fat or food fat upon which they could draw. Just as Folin's non-diabetic obese patients readily developed severe acidosis upon fasting, fat dia- betic patients can do the same, though much more readily upon a fat-protein diet, as illustrated by Case Xo. 983. Case X"o. 983, a woman of 55 years, onset in December, 191.1, came for treatment in January, 1910; weight, 201 pounds; height, 5 feet 3 inches. An acidosis amounting to 4.3 grams ammonia and an alveolar air of 21 mm. Ilg, which she acquired upon a fat-protein diet, did not disappear for eighteen days, but she ultimately left the hospital eight, days later, with a tolerance for over 50 grains carbo- hydrate and 80 grams protein. If the acidosis is not checked, it can result within a comparatively short time in corn a, and it now appears that even if partially checked, and yet the patient is allowed to live upon a low carbohydrate diet for a period of months, eventually tolerance will be lost, as Allen has shown for his dogs, and acidosis will claim the victim. The treatment for these patients, therefore, is to withhold fat entirely at the beginning of treatment; as a result the carbohydrate toler- ance can undoubtedly be raised to a higher figure and the protein as well. Gradually small quantities of fat may be allowed, but the weight of the patient should be slowly decreased how far it is impossible to say, and only time will be a guide. I believe it seldom a good plan for normal obese individuals to lose more than one pound a month, and I doubt if these fat diabetics would need to lose weight any more rapidly. These 3 cases are so clear cut that I am confident they will awaken memories of similar cases in the minds of other physicians. Their deaths were' so regrettable, and to my present, thinking so easily to be avoided, that I consider it well worth while to give them in abstract. Case Xo. 974, with onset in December, 191 5, at the age of twelve years and eight months, came under observation in January, 1910. Height, 144 cm.; greatest weight, 38.0 kilos. It will be seen from his chart, Table 15."), that he became promptly sugar-free with THE FATAL CASES OF 1016 345 fasting and acquired a notable tolerance. In October, 1910, he returned to the hospital with considerable acidosis, and a high blood sugar, which did not yield to the short period of treatment, and the patient went home upon an inadequate diet. Through a mutual and unfortunate misunderstanding the patient left the hospital long before he was in condition to do so. The penalty was swift and severe, for he died in nine days. TABLE 155. CHART OF CASE Xo. 974. DEATH OF A TWELVE-YEAR-OLD BOY NINE DAY'S AFTER DISCHARGE FROM THE HOSPITAL. Urine. Diet in Ejrams. "Pi 1 Date, 1910. Diaretie j Ammo- rp , Carbo- Pro- Fat. Calor- Blood Alveolar Carbo- S "f ar ' air CO,, hydrate , mm. Hg. acid. total s u^ r ' tiydrate. tein. ies. balance. "' nt ' Knis. gms. Jan 18 (7.2%) 20 + + + + 20-21 1 . 4 30 Fas tin !I 18 hou rs 25 21-22 + 31 22-23 + 3 2 20 + 3 29 23-24 + + 6 3 36 + 6 24-25 + 15 21 6 198 + 15 24 25-26 + 20 34 11 315 +20 26-27 25 40 16 404 +25 0.17 32 30-31 25 50 48 732 + 25 31- 1 16 10 34 35 371 - 6 Feb . 1- 2 2- 3 10 34 35 71 + 10 3- 4 25 46 52 752 +25 0.19 8- 9 27 5'i 90 1142 + 27 Oct 4 - 5 + + 2 . 2 27 40 5 180 + 13 . . 24 5 6 + + 2.4 42 30 S 152 -12 0.43 6- 7 + + 33 20 10 120 -13 7- s -f- 3 15 8 92 + 12 8- 9 + 16 10 5 60 - 6 9-10 SI. + 10-11 0.33 11-12 10 5 60 + 10 12-13 20 10 120 +20 0.50 13-14 4 30 15 180 +26 14-15 2 10 5 60 + 8 0.33 15-16 16-17 5 37 53 645 + 5 26 Died at home in en ma Weight at entrance 68 pounds; at discharge 68 pounds. Phenolphthalein test, 22 per cent, in two hours. Case Xo. 1058, a woman, aged thirty-nine, years, developed diabetes in October, 1915, and came for treatment May, 1916. Height, 100 cm.; weight, 74.8 kilos, in contrast to the normal weight of about 00 kilos. The onset was acute, in almost a D day, though the exciting cause she considered a bad fright in August, 1915. She remained five weeks at the hospital, easily became sugar- free with fasting, but in the process acidosis appeared, but was nearly 340 TREATMENT dispelled by (10 grams levnlose. She acquired a moderate tolerance for carbohydrate, went home in good condition but died five months afterward. (See Table 1 .")(>.) TAHI.E 150. CHART OF CASE No. 1058. Total Date, Diacetic su^ar I'.llti. acid. in urine, RIT1H. Diet in grains. Blood Alveolar Kiifiar, , air ( '( ) 2 , per cent. nun. Jin ('arl)o- Pr hydrate. ' tei Fat. Calories. May is 10 I) 54 ' 10 20 2 20 21 4- + + 21 22 4-4-+ 20 ) , 100 29 22-23 4-4- 30 22 , 20S 23 24 4-4- 2.52 GO- ( ) 210 0. 14 21 25 + 40 ' 3 10 374 0. 14 25 2G SI. + 30 1( ) G 23S 32 20 27 ++ G 20 ) 2 us 33 27 2S SI. + 20 2: IS 330 3G 2s 2<) SI. + 25 30 22 451 29-30 25 r 33 4S5 .'!'' i June G- 7 33 G 117 1-12!) Nov. 21 Died in coma Weight at entrance 130 pounds. Weight at discharge 125 pounds TABLE 157. CHART OF CASE No. 733. Date. Diacetif acid. Total suwir in urine, grains. Carbo- hydrate. Diet in Protein. Drains. Fat. Calories. XallC'O. 1915 July 30 31 . . + + + !>3 3 1 14 31- 1 . . + + + !) 3 4 Aug. 1 2 . . 3 112 4 23.. ' 5 34. . 10 IS 15 SO. 25 5G 71 10-11 . 28 55 103 1259 11-12. 12 13 . . 2!) 04 120 191fi Apr. 22 . . Died in coma . .__ _ Weight at entrance 133 pounds. Weight at discharge! 12.~>i pounds. Case Xo. Too, onset of diabetes, at the age of sixteen years and eight months in May, 1913, came under observation in. June, 1014. Height Hi") cm., and greatest weight SO kilos! With fasting the patient. promptly became sugar-free. Not recogni/ing the importance two years ago of building up a tolerance for carbohydrate and keeping the fat low the patient was discharged with a low carbohydrate diet and 2 Loyulose. THE FATAL CASES OF in 16 347 high quantity of fat. Whereas this ease yielded promptlyto treatment and was undoubtedly mild in type, death ensued in nine months from coma. (d) Treatment Abandoned. Treatment was deliberately aban- doned not neglected, but actually given up by 8 cases, constitut- ing 25 per cent, of the deaths from coma and 10 per cent, of the total deaths for the year. It is a very poor excuse for a physician to claim freedom from responsibility from a death due to coma by saying that the patient abandoned treatment. The physician should make his patient so much more comfortable undergoing treatment than neglecting it that treatment will be followed throughout the course of the disease. Analysis of these 8 cases of patients who abandoned treatment shows that no one abandoned treatment during the first year, but that 2 cases did so during the second year. In one instance this was done by the parents of a little child seven months after leaving my observation, and at a distance of over one thousand miles, and by the parents of another child who, several months before the death of the child, ignorantly broke the diet and later allowed the child to break it. This child died one and a half months after he was last seen by me. During the third year there were two such deaths; both patients were adults and severe cases. One case repeatedly broke treatment in the hospital and gave up treatment after leaving the institution, and had various difficulties with physicians in other hospitals. The other patient was said to have broken treatment in the hospital; openly renounced treatment on discharge and refused to have the urine examined by her local physician. One of the two patients who abandoned treatment in the fourth year was a man whose diabetes was complicated by alcoholism, and the other, a nurse, who incident- ally was really at the start probably a rather mild case. She was fat and today could easily be kept in good condition upon a fairly liberal carbohydrate diet. A single case gave up treatment during the sixth year. This was case Xo. 1005, who felt that she had a full knowledge of the disease, literally wandered over five thousand miles in the course of comparatively few. months, broke treatment in the hospital, broke treatment outside, and reentered in coma and died. Last of all was a patient who gave tip all treatment after having the disease twelve years. Her case is next described. Case Xo. 927, with onset in 1903 at the age of forty-five years, came under my observation in October, 1915, at the age of fifty- seven years, marked diabetic heredity, greatest weight 105 pounds, but on entrance 11 1 pounds. Height five feet six and a half inches. A poor, feeble, nearly blind patient, tormented with pruritus, with an enlarged heart and liver, albumin and casts, entered with only a small quantity of sugar in the urine and a diet which was partially 348 TREATMENT satisfactory to her. In its ])l;u:e we gave a diet adequate in calories, rich in fat, and lo\v in carbohydrate, and produced thereby acid poisoning. The patient was unhappy in the hospital, and as I now look back on the case, had reason to be. Contrast the treatment this patient received with the treatment accorded Case Xo. 759 (see p. 489) and that of Case Xo. 979 as described on p. 308. It should be noted that the patient came in without acidosis but it followed fasting; with the acidosis it is quite probable that her consid- erable tolerance for carbohydrate, which must have existed, fell. TABLE 158. CHART OF CASE Xo. 927. ILLUSTRATION OF POOR TREATMENT OF A DiAi'ETie CASE OF TWELVE YEARS' DURATION. SEE TABLES 142 AND 143 FOR EXAMPLES OF BETTER TREATMENT. Date, 1915. Total Diacetie sugar in acid. urine. Diet in grains. farbo- Pro- r hydrate. toin. i Carbo- it. Alcohol. C'alories. liydrate. balance. Nov. 1 2 0- 7 + + 3 10 3 1 3 1 o 15 s o 15 8 15 20 15 15 35 10 15 11 21 15 11 21 15 41 40 3 21 i s 1 29 2S 1 29 2s 1 31 41 1 31 11 -23 40 _ 2 12 IS 142 + 3 IS 21S + 15 IS 218 + 15 is 325 + 15 18 470 + 15 18 539 + 15 IS 539 + 15 (I 080 + 15 (i 42 (i 206 + 25 414 + 1 (i 414 + 1 551 + 1 (i 55 1 + 1 The problem before the physician, therefore, in preventing abandonment of treatment is (1) tin' education of the patient and the relatives and C2) the improvement of treatment so that the greatest possible tolerance for carbohydrate can be acquired and the diet made easier to follow. To do this would have been possible both in Case Xo. 707 and Case Xo. 927. Two of the other patients were so dishonest that it is a question whether this tendency could have been controlled, but I believe that even with these, more care and the demonstration of the advantages of strict adhesion to rules and graphic illustrations of the harmful ness of breaking treatment by living examples might have led to the avoidance of an untimely end. It is notable that in only one death did alcoholic habits play a part. 1 Hloud auf^ar, 0.23 per cent. - Alveolar air, 35 mm. Tin. THE FATAL CASES OF 1916 349 (c) Imperfect Supervision. Imperfect supervision led to the death of 7 cases during the first and second year of the disease and three deaths in subsequent years. More deaths therefore were caused by imperfect medical supervision than in any other way. This is a reproach and an encouragement at the same time. It will radically affect my own treatment of diabetic patients; it shows the necessity first for better education of the patient, so that when not doing well the advice of a physician shall be sought; it shows also the need for closer cooperation between the family physician and the physi- cian who sees the patient in consultation or at the hospital, and, above all, it shows the necessity for simplification of methods of treatment, (loser supervision of the cases demands an enormous amount of time, but time so spent will be repaid by lives saved. All of us must devote more time to our cases, and we must run the risk of the opprobrium of forcing ourselves upon patients by calling to their attention or the attention of their physicians the need of constant supervision. Patients must understand the facts and be made to realize that lack of supervision often means needless death. Each physician has it in his own hands to protect the lives of his diabetic patients. (/) Increasing Severity of Diabetes. Increasing severity of the dia- betes may have been the cause of death of seven of the coma cases. TABLE 159. DEATH ix COMA DUE TO INCREASING SEVERITY OF THE DISEASE. Ago at Duration, Case Xo. oiisut. years. 363 ... 14 5.0 298 ... 10 7.0 181 ... 43 12.0 534 ... 9 5.G 552 ... 30 4.8 821 ... 23 2.4 775 ... 7 2.3 Though from four months to six years elapsed from the last time I saw these patients until death, a considerable share of the responsi- bility rests upon me, for in most instances I was frequently consulted about treatment, and in all of these cases treatment was quite closely followed; and though in no instance was the treatment as good as could be offered today; on the other hand, the treatment was infin- itely better than the treatment afforded my patients ten years ago. It will be noticed that nearly all these cases were of many years' duration. Certainly doubt arises as to whether the increasing severity of diabetes is a proper title for these deaths. AVith what is known about the prevention of coma today, I think it likely that this could have been deferred in at least two of the cases which were complicated with infections, and I think the criticism entirely Coma associated Period between with last visit and death. "Grip" 3 1 years G Bronchitis 4 months "Grip" 31 years 1 year 5 months 1 year 350 TREATMENT justifiable that though increasing severity of diabetes may have caused these deaths, the deaths occurred too soon. (()) Ether. During 191(> the death of one patient may have been due to ether anesthesia. This was a man, aged seventy-two years, with gangrene; duration of diabetes over six years. Though lie showed no aeidosis before the operation, he promptly developed it aftenvard and rapidly succumbed. I did not anticipate such a death, and I believe that it might have been avoided. (h) Carbuncle. A carbuncle this year, as in the preceding year (CaseXo. ">lo, p. )").")), was responsible for the death of one patient. This patient (( 'ase Xo. 904) had had diabetes for fifteen years, and was in quite fair condition when he entered the hospital. I had seen him off and on for several months outside when he had absolutely refused entrance to the institution, though offered free board and free treat- ment. When a carbuncle developed and he was forced to enter, the difficulty of treatment was greatly enhanced. The treatment of carbuncles is discussed on page 429. G. CASES UNSUCCESSFULLY TREATED BY FASTING. Under this heading will be discussed certain cases treated during the year and a h;'lf prior to December 1, 191"), which present features not sufficiently emphasized in the fatal cases reported in the preceding section. In addition to simply reporting the cases I have now added critical comments based upon the- experience of another year. Of 12.") patients treated by fasting beginning May, 1914, 14 patients had (lied by December 1, 191."). A study of the 14 fatal cases sho\vs plainly that several of the deaths might have been avoided or deferred. Thus '.> of these patients were among the first cases fasted and treatment was incomplete; the abandon- ment of treatment by the physician in one instance and by the patient in another accounted for t\vo more deaths and lack of medical supervision for a sixth. A seventh death was due to cancel 1 of the blad- der and death ensued because of a recurrence of the growth in the scar of the abdominal incision forty-one days after the primary operat ion ; the death of another patient also appeared to be entirely independent of the diabetes, and no other diagnosis than arteriosclerosis could be made. The remaining <> cases deserve a more extended notice. 1 . Death Caused by the Inauguration of a Fat-protein Diet in a Child.- ( HSC Xo. S( '."), ;iged two years and two months, developed diabetes three weeks before coining under my observation, May 24, 191."). A grandmother has diabetes. The patient was said to have had (')..") per cent, of sugar early in May, and was reported to have 1.2 per cent, of sugar on May -'.}, but the urine voided on the evening of May 24 showed 4.4 per cent, of sugar, with a marked CASKS UNSUCCESSFULLY THE AT ED BY FASTING 351 reaction for diacetic acid, a slight trace of albumin and an abun- dance of granular casts. Prior to admission to the hospital, the patient had been put upon a nearly fat-protein diet, and although his condition appeared fairly satisfactory when first seen at 10 o'clock at night on May 24, he had clear signs of diabetic coma the following morning. Despite fasting, which began upon entrance, salt solution by rectum and alkalis by mouth, he steadily failed, although on the following day the respiration was not as typical of coma as before. Later complete coma ensued with death on the evening of May 26. The last specimen of urine examined on May 2(5 showed no sugar and no diacetic acid. March, 1917, comment upon treatment of Case Xo. 805. The death of this child, in the light of our present knowledge, I consider absolutely unnecessary. Several days before the child came to Boston his physician telephoned me. I should have advised him to give the little boy a saucerful of 5 per cent, vegetables and an orange at each meal, or, if he preferred, either a small saucer of oatmeal or half a shredded wheat biscuit in place of the orange, with not over a gill of milk in the twenty-four hours. Even a little fish or lean chicken would be allowable. I should have forbidden fat in any form. As a result the sugar would have quickly dropped and a day or two later the total diet could have been curtailed if necessary and the child would have been sugar-free; acidosis would not have appeared at any time and tolerance could have been built up in easy stages. How needless the early death of this little child! The case has many analogies, notably the patient reported by Geyelin and DuBois (see p. 114), but with this difference: they saved their patient. If I should allow another such case to occur, I should blame myself as much as would a surgeon to have a general peritonitis follow the removal of an "interval appendix." Suppose the physician had not telephoned and that I heard of and saw the case for the first time when he entered the hospital at 10 P.M. The indications were plain. He had travelled one hundred miles and had eaten hearty food both late at night and on the train, and the bowels had not moved. The stomach and bowels should have been promptly emptied. Milk of magnesia or an enema, or both, should have been given at once. With a clear digestive tract the next morning a little oatmeal water gruel or albumin water and orange juice could have been given until the early stages of approach- ing coma had been dispelled. As it was I allowed the child to go to sleep with stomach and bowels full, gave alkalis the following morning, thus causing nausea and setting free acids which could not then be eliminated because of the difficulty of giving the child suffi- cient liquid, and in consequence coma ensued. I nder similar con- ditions today I believe the chances of saving this child would be 10 352 TREATMENT to 1 in his favor. Note that with the involuntary fasting, induced by my poor treatment, sugar and diacetic aeid disappeared before death ensued. For those who may be skeptical of my views about the treatment of such cases, I would refer to Cases Xos. 1)23 and 0:58, Tables 212 and 213 respectively, who were treated by fasting. 2. Intercurrent Infections. Intel-current infections played an important, role in the death of 2 cases. It has already been pointed out that an infection lowers tolerance for carbohydrate, in other words, increases the severity of the diabetes. The greatest caution, therefore, must be adopted in the management of such patients. The records of the following 2 cases illustrate several errors in man- agement: Case No. So"). Acute gangrenous appendicitis. A gentleman, aged seventy years, weighing 220 pounds, with diabetes of ten years' duration, was first seen four days after an operation for appendicitis at which, to the great regret of the surgeon and physician in attend- ance, it was necessary to use a large quantity of ether. This was in part due to one-fourth grain of morphin given subcutaneously preliminary to operation, which so reduced the respiratory excursion that nitrons oxide was a failure, and a large quantity of ether was required. The patient had developed an acute appendix while returning from the South, April 0, 1915. On April 1 1 the white count was 17,000 and the temperature 100. On April 12 the sugar in the urine was 11 grams and that in the diet IS grams. On April 13, the date of the operation, the patient took 14 grams of carbo- hydrate in the diet and voided 24 grams of sugar in the urine and 0.7 gram ammonia. lie underwent almost complete fasting for the four days subsequent to the operation, yet failed to become sugar- free, and general circulatory weakness increased. The wound progressed satisfactorily. Subsequently he was given whites of eggs and vegetable broths, with the result that the sugar in the urine rose. Later, further additions to the diet were made, with the same result, and he finally died, eight days after the operation, being conscious up to the last minute. March, 1017, comment upon Case Xo. S55. In view of the serious situation from the diabetic stand-point namely: (1) a case of ten years' duration; (2) a fat man; (3) an infeetion- fat should have been omitted from the diet April 9 and the food allowed should have been in the Form of water gruels, the quantity of carbo- hydrate approximating that taken by the patient in the immediate past. This diet would be continued after the operation, adding thereto protein in simple form as soon as it could be borne. There would be no need of adding fat, for he weighed 220 pounds. A relatively large quantity of carbohydrate is indicated because fat CASES UNSUCCESSFULLY TREATED BY FASTING 353 patients easily develop acidosis upon fasting, yet their disease is mild and with moderate restriction of calories their truly high tolerance for carbohydrate becomes manifest. From my knowledge of the patient I doubt if the operation could have been performed without ether, but the conditions which led to the large amount of ether being used could be avoided in another case. Under the best of conditions recovery would be exceptional. Case Xo. 513. Multiple carbuncles and septicemia. This case has few, if any, counterparts in the literature, and is therefore reported in detail. It is remarkable for: (1) an average loss of weight of 1.0 kilograms (3.5 pounds) daily for eleven days; (2) the excretion of between 31.6 to 37. S grams nitrogen daily for the six days pre- ceding death; (3) a dextrose-nitrogen ratio of 3.67 : 1, when nearly fasting; (4) a variation in the daily excretion of sodium chloride between the limits of 3.3 grams and 35.8 grams. Onset of diabetes in October, 1911, at the age of thirty-three years, and first came under observation in July, 1912. At that time, while at the hospital, he became free from acidosis, but not from sugar. During the next three years he worked hard in a market, living upon a partly restricted diet. lie was not seen except during August, 1912, and July, 1913. Upon July 29, 1915, he tottered into the office with a carbuncle on the neck of eleven days' duration, and at once entered the New England Deaconess Hospital. During the first two weeks (July 29 to August 12) in the hospital the change in the condition of the patient was favorable, and the surgeon said the carbuncle did as well as with a non-diabetic patient. Deep-seated abscesses then began to appear in various parts of the body, and the patient began to lose weight, and in the last ten days before his death, on September 2, rapidly failed. His condition was most pitiable, complicated, as it was, with the pains of multiple carbuncles, extreme weakness, and septicemia. Death occurred thirty-five days after entrance. (See Table 160.) It will be seen that he was fasted for six days, save for 48 grams of alcohol daily, and on the following four days took 18 to 50 grams protein and from 15 to 27 grams fat each twenty-four hours. He was then allowed a slightly greater quantity of protein and fat for two days, then again fasted for two days, when the urine became free from sugar and the acidosis disappeared. The diet was again increased, but sugar returned, even amounting to 40 grams in twenty-four hours, although the carbohydrate in the diet was 1 gram, the protein 73 grams, and the fat 98 grams. After four days of fasting the quantity of sugar in the urine fell to only 46 grams. From this time on he was given no carbohydrate and from 37 to 12 grams of protein and 46 grams of fat on four successive days; then several more fasting days were employed, yet at no time did 23 oo[;[ - r- r "5 o o , o >fi> * o o us to . o . c-. o cr. ~. o cr. cr. c-. c: ~ cr. ~. ~ o c-. c. cv c- r-^-fOCO^OOOCOOCO I I I I I I I I I I I I -i co ^-i o + + -f +- i- 00 O! ( 354 ) M -t "c m M ~i sr. y. c; c c: cr- '^ -co -co ; o >< r- ?i -r -r o o -* -* c-i ~r ~> c ^t K -r ?. -* cr. * < c o o o + + + + H -J- + + + + + + ( 355 ) 356 TREAT Ml': XT he become sugar-free. Aeidosis steadily increased and he finally died, undoubtedly with acid poisoning, but not in coma, on Sep- tember 2, 191o. A blood culture made immediately after death showed a pure growth of Staphylococcus aureus. Carbon Dioxide. The carbon dioxide tension of the alveolar air was 29 mm. in the morning and 25 mm. during the afternoon of the day of admission, July 29, and gradually rose after the daily administration of 32 to 4S grams sodium bicarbonate in seven days to 54 mm., and it will be seen that this persisted even when smaller doses of alkali were given. On August 1 1 the carbon dioxide tension amounted to 4.~> mm. mercury, although no sodium bicarbonate was given that day. Later it decreased, coincidently with the development of metastatic abscesses, and despite SO grams more of sodium bicarbonate, it kept in the neighborhood of 22 mm. mercury. IHacctic Acid. Diacetic acid was present throughout, except for two days. Ammonia. The ammonia was 3.7 grams during the first days at the hospital; later decreased to the neighborhood of 1 gram, but during the last few days of life rose, reaching 5 grams in twenty- four hours, four days before death, although 90 grams of sodium bicarbonate were taken daily during this period. Xitroyen. The nitrogen remained practically constant, between 9.3 grams and 12.5 grams, during the first ten days of fasting and a low diet. In the following feeding period two observations showed that it rose slightly, but the striking feature of the nitrogen excretion was the marked elimination during the last six days of life, when it varied between 31.0 grains and 37. S grams, although the patient took almost no food. How long the nitrogen had been high it is impossible to say, but the marked increase in nitrogenous excretion can hardly be explained by the washing out of retained nitrogen, for at this period, although there were, no carbohydrates in the diet, the amount of sugar in the urine was considerable, from 72 to 125 grams, and it will be seen that the dextrose-nitrogen ratio at this time was approximately 3 : 1. Xodiinii Chloride. -During a portion of the period of observation the salt in the urine was determined quantitatively. It will be noted that this was very lo\v --4.7 grams to 3.3 grams ten days after entrance to the hospital. During the latter portion of the fasting period, August 24 -20, it was about normal -S. 7 grains to 12J') grams but on the following days it, rose with remarkable rapidity to 20, 35. S, and 35.0 grains. The great increase in the excretion of nitrogen was also striking, and would lead one to con- clude' that the increased excretion of both salt and nitrogen was due to a washing out of the system of loosely stored nitrogen and salt, were it not for the fact that the sugar in the urine rose as well. One CASES UNSUCCESSFULLY TREATED BY FASTING 357 cannot logically attribute the increased excretion of sugar to a washing out of body sugar, because the percentage of sugar in the blood throughout this period remained practically constant, and the diet contained no carbohydrate. It is more logical to conclude that the sugar and nitrogen stand in relation and that the sugar in the urine really comes from body protein, but before accepting this conclusion the remarkable changes in body weight must be considered. Body Wciglii. The weight of the patient July 30 was 62.7 kilo- grams (138 pounds), and on August 6, GO. 7 kilograms (147 pounds), although the total calories ingested by the patient during these seven days amounted to 2531, or 362 per day. Therefore this gain in weight must have been due to a retention of water. Explana- tion of this is afforded by the 288 grams of sodium bicarbonate given during these seven days. Further evidence that sodium bicar- bonate was retained in the body is furnished by the coincident gain in the COo tension of the alveolar air. On August 5 the quantity of sodium bicarbonate had been reduced to 24 grams, and during the following five days was reduced still more, until on August 10 it was entirely omitted. Yet during this period the weight continued to rise O.S kilogram (If pounds). Edema was manifest, but not very marked, even when the weight was greatest. During the following nine days, August 11 to 19, no sodium bicarbonate was given, and the weight rose 0.3 kilo (f of a pound) more; but from this period on it steadily fell, and in eleven days, August 19 to 30, dropped from 67. S kilos (149^ pounds) to 51.8 kilos (113| pounds), a loss of 1.6 kilos (3.5 pounds) per day. This rapid loss in weight was not detected in the first few days, because the patient was so feeble and in such pain from multiple carbuncles that it was difficult for him to be moved; but on August 30 this remarkable loss was proven. The actual loss may have been greater, for during this period SO, 96 and 96 grams sodium bicarbonate were adminis- tered on respective days, and before this alkali was given the weight may have been lower. As a matter of fact, it did rise during the following forty-eight hours to 52.4 kilos (115| pounds), when 144 grams of sodium bicarbonate were taken. The weight of the body immediately after death dropped to 51.8 kilos (114 pounds). The striking loss of 1.6 kilos (3.5 pounds) body weight daily for eleven days, and the coincident increase in sugar and nitrogen excreted, suggest very important changes taking place in the metabolism. They indicate in a striking manner the importance of maintaining reserve fluid in the body. So shrewd a therapeutist as Hodgson has recognized the advantage of having patients drink large quantities of mildly alkaline water. The therapeutic benefit which his clinical sense showed him to result from this procedure 358 TREATMENT secures striking confirmation in these figures. Furthermore, they suggest the importance of studying the weight of severe diabetic patients, preceding, during and after corn a. We recall no figures which show so rapid a loss of weight in a diabetic patient. March, 1017, comment upon Case Xo. 513. This man had 14,S31 calories in twenty-five days, and if we deduct the calories from alcohol, he had only 11,57(5 calories in twenty-five days, or an average of 4(50 calories per day the entire period. He should not have been allowed out of bed! The low caloric diet may to a con- siderable extent have accounted for the marked protein disintegra- tion immediately preceding death. Case No. 110(5, without an infection, showed the same during extreme inanition and yet recov- ered. The diabetes was not extremely severe at the beginning. J am inclined to believe that with omission of fat, gradual reduction of carbohydrate and protein, and the avoidance of alkalis the patient would have become sugar-free more rapidly. The increase of sugar coincidently with the increase of fat in the diet is beauti- fully shown August 15-1(5 to August IS- 10. Unquestionably gradual increase of protein and fat would have produced better results and the gathering cloud of acidosis might have been dispelled. The cardinal failure here on repeated occasions was the too rapid increase of food. Very likely this patient \umld have even shown some tolerance for carbohydrate. With acidosis conquered, a greater caloric intake rendered possible by a progressively increasing diet and the avoidance of so much fasting, the resistance of the patient might have been better and metastatic abscesses and car- buncles might not have developed. I am not inclined to follow Christie's' example and fast another such case eleven days to make the, urine sugar-free, but I shall try first to conquer acidosis, recom- mend the simplest possible surgical measures for the carbuncle and endeavor to maintain nutrition. Xo alkalis! 3. Diabetes of Long Duration. Changes in the diet and regime of patients who have suffered from diabetes for many years are a 1 \vays dangerous, and particularly so in the cases of extraordinary length. Treatment of such individuals should never be undertaken lightly nor without a full realization of the gravity involved. This has been illustrated by Cases 205, \\(\\ and 3U). 2 These three were hereditary cases and this Fact may have accounted for the long duration of their disease, which was nine, seven, and nineteen 1 Christie: Lor. fit., |>. \W. "Sec Josiin: Dialiclic Standards, Am. Jour. Mod. Sr., 101.3. cxlv, p. 474; ;xlwo Konodirt and Josiin: ( 'arneuie Publication No. 130, 1910, pp. 1LS and 120. CASES UNSUCCESSFULLY TREATED BY FASTING 359 years respectively. They finally came under my supervision in a much debilitated condition. Even at that time the danger of changing the diet was appreciated, and unusual care was taken to prevent a fatal issue. The 3 cases died in coma, 1 upon a sea voyage, 1 after unusual exertion, and 1 for reasons unknown, in periods of two and a half years, two months, and five months, respectively, after being seen. Consequently, when Case No. 887 came for treatment, with a duration of the disease of twenty-nine years, unusual apprehension was felt. For nine days the patient was not allowed to make the slightest possible change in the regime, for, although she was wretched, she was alive. She was then admitted to the hospital, but with much foreboding, although no alarm was felt by other physicians, experienced in diabetes, who saw her. Table 1G1 shows the progress of the patient. It will be seen that fasting for nine days failed to rid the urine of sugar; indeed, upon the ninth day the quantity of dextrose in the urine was 17 grams, and the diet contained but 8 grams of alcohol. It is true that upon the second day after admission the urine was reported as sugar-free. Acting upon this supposition, on this day 5 per cent, vegetables (150 grams) were allowed, but except for this day fasting was complete from July 31 to August 7. On August 8 three eggs were allowed; on August 9 three eggs, lean meat 30 grams; on August 10, 200 grams of orange, about 15 grams oat- meal and the whites of three eggs, but otherwise the patient took no food from her entrance to the hospital on July 30 to death on August 12. March, 1917, comment on Case Xo. 887. This case constitutes a distinct failure. Contrast the treatment of this case with that of Case Xo. 1151, p. 171, and with other cases cited in the discussion of the treatment of Cases Xos. 979 and 1007, Tables 142 and 143. Table 101 shows that the quantity of urine varied little for the first five of the thirteen days. From then on it increased in amount, and it is here again significant that with it rose the nitrogen, except upon the second day before death. With the increase in the quan- tity of urine, there was an increase in acidosis, as shown by the ammonia. The carbon dioxide tension of the alveolar air did not change materially during the whole period, for it was 32 mm. Ilg upon July 30, and 29 mm. Ilg the morning of August 11, but upon the following day, shortly preceding death, it was 14 mm. Ilg, but an estimation made at this time may not have been wholly reliable. The inferences which were drawn from the behavior of the alveolar air were misleading. X'o great change took place in the quantity of sugar excreted. The dextrose-nitrogen ratio at the beginning afforded an index of the severity of the disease, though this state- 3(>0 TREATMENT r.( i. ) 'jn: spiinocl -p;4 TREATMENT Xo. SS7, described under Diabetes of Long Duration. This is evident, because the patient retained no food for many hours and yet voided during this period at least 2000 c.c. urine. March, 1017, comment upon Case Xo. 8f>9. The cardiorenal condition was quite as severe as the diabetes. By treatment the diabetes ought to have been improved, not made worse. The patient entered the hospital without acidosis and should have left without its development. The omission of fat followed by the cautious restriction of carbohydrate and protein would have rendered this possible. By the stay in the hospital both the diabetes and the heart disease should have been benefited. 5. Syphilis. The treatment of syphilis may have been responsible for the deatli of one patient, but changes in diet very likely were an equal factor. Case Xo. S0(i, male, aged thirty-eight years, came under observation December 17, 1914, reporting an onset of diabetes two months before. The urine contained 7.8 per cent, sugar, but after two days' fasting glycosuria disappeared. The tolerance gradually rose to -4 grams carbohydrate, 50 grams protein, and enough fat to maintain Aveight. Following discharge, the patient remained in good condition under the observation of his physician, and, so far as 1 can learn, was sugar-free. In June, 1915, I heard from his physician that another doctor had obtained a positive Wassermann reaction, that three injections of salvarsan had been given, and mercury and iodide of potash as well. Shortly after, the patient became suddenly ill and died in coma. March, 1917, comment upon Case Xo. 800. The possibility of syphilis should have been proved or disproved when the patient was first seen and the proper treatment of the syphilis instituted under the favorable surroundings of the hospital. H. SPECIAL DIETETIC METHODS. 1. Fasting Days. Looking back upon the treatment of diabetes prior to Allen's introduction of prolonged fasting, it really seems, to use an expression of Xaunyn's in reference to the care of severe cases, that our diabetic, patients were nursed rather than treated. Only those who have cared for many patients by the older methods can appreciate the advance which Allen has given to diabetic therapy. Many years ago Xaunyn strongly urged the use of starvation days in the treatment of diabetes. lie repeatedly called attention to the advantages derived therefrom, and said one should not fear temporary undernutrition if thereby it was possible to remove the sugar from the urine. Over and over again a day of SPECIAL DIETETIC METHODS 305 starvation would render a patient sugar-free when all else had failed. Such days were usually indicated in the treatment of the severest cases of diabetes, and in a case of moderate severity such a day would do what many days of low diet had failed to accomplish. But it was not alone to render the urine free from sugar and to place the patient in a better condition to acquire tolerance for carbo- hydrates that Xaunyn used these days, for he observed that they helped to lower the acidosis in a striking manner. Von Xoorden agreed with Xaunyn that these days were never disadvantageous. lie writes: "I make use of these, especially when there is high acetonuria. It is astonishing how strikingly the acetone falls on a hunger day. Its effect stretches out for a number of days later. In numerous severe cases a hunger day has been instituted every week with excellent results." A third reason existed which also accounted for the good effects of a starvation day, to wit: the patient's digestion was given a rest. X'aunyn, in an indirect way, protected the digestion of his patients by emphasizing the value of a low caloric intake. Hodgson, too, is precise upon this point. He says: "Again it should be stated that the quantity of all food, even if it is carbohydrate-free, must be greatly restricted. The number of calories that the body ordinarily requires is no safe criterion of the amount of food that should be given a diabetic. It is not the quantity of food that should be metabolized, but the amount that can be metabolized that should determine the amount given to the patient. All in excess of what a patient can actually use burdens the already overtaxed excretory organs and retards improvement." Hodgson should be given the credit of having published this article in 191 1. 1 Guelpa, in 1910, 2 reported his success in the treat- ment of diabetes by the employment of several days' fasting com- bined with purgation: "I do not pretend that this cure has been put as yet on a definite scientific basis. Much further and more generalized experience of it is required before we can arrive at clear conclusions as to the extent of its sphere of usefulness. From the observations I have already made, however, of the action of the cure, it is possible to draw certain inferences of capital importance. "1. There is absolutely no danger and no serious inconvenience in abstaining entirely from food for three or four days, or even longer; the period of abstinence, also, may be repeated several times without risk or inconvenience, if each day a large dose of a purgative is given to insure intestinal disinfection. " 2. While there may be some slight discomforts during the period of abstinence, these never persist after food is resumed. On the other hand, undoubted and durable benefits are always gained, in 1 Hodgson: Jour. Am. Mecl. Assn.. 1911, Ivii, p. 1187. = Guelpa: British Mod. Jour., 1910. ii, p. 1050. 36() THE ATM EXT the shape of increased freedom of movement, greater clearness of ideas, amelioration of all congestive conditions, and a true feeling of general well-being. In a word, one is always better in all respects after the cure than before. ".'!. The cure is a perfectly safe procedure if controlled by exami- nations of the blood and the urine. It insures the maximum of benefit being derived from any concomitant therapeutic measures." (iiielpa reports many cases of diabetes and the astonishingly good results which followed the method he employed. He deserves great credit for his originality and courage in applying this method as well as for his modesty, and his name should always receive a prominent place when fasting is mentioned. The advantages of the fasting method at the outset of treatment are many: (1) It is a simple method and enables the practising physician to render all save exceptional cases of diabetes sugar- free and to keep them so. (2) The treatment is such that acidosis is almost invariably diminished rather than increased. How much such a gain in treatment means is plain when one remembers how large a percentage of the fatalities of diabetes during the first twelve months of the disease is due to coma. Furthermore, should acid poisoning occur in the course of treatment, fasting will almost invariably lessen and eventually dispel it. In consequence, alkalis have become seldom, if ever, necessary. (.'>) The method is suit- able for a patient to learn, and he is thereby placed in a position to protect himself. (4) The expense of treatment to the patient is reduced because he can readily perform the urinary tests essential for treatment, and by keeping sugar-free, visits to a physician are rendered infrequent. (5) The responsibility of treatment rests upon the patient, and this makes the chances of his following direc- tions far more probable. But the most satisfactory part of fasting treatment is the sound basis upon which it is founded. Others have employed and advised fasting; Xauuyn and (iuelpa tried it exten- sively; Hodgson used a very low diet with great success, but Allen }>rnml prolonged faxtiny efficacious. Before hr treated ani/ human jtaftentx hi/ faxting, lie demonstrated that flux method icax iixefnl for animal*. To accomplish this he performed partial pancreatectomy in dogs, preserving that portion of the pancreas adjoining the duct, so as to avoid atrophy of the remnant. Allen found that, according to the amount of pancreas removed, simple lowering of the sugar tolerance without actual diabetes results, or a diabetes varying in intensity from the mildest to the very severest type can be produced. Ivcstriction of carbohydrate or protein, or brief fasting, would render animals with the mild type sugar-free, but to accomplish this with the animals of the severest type fasting must be prolonged for many days or even weeks. Eventually, however, it could be SPECIAL DIETETIC METHODS 367 successful, and along with the disappearance of the sugar, acidosis would also pass away. Allen's experience convinced him that the glycosuria of even the severest type of diabetes might be cleared up advantageously by one initial fast, though it might be necessary to prolong it exceptionally for eight or ten days. And he goes on to say: "Broadly speaking, freedom from glycosuria seems obtain- able in all cases of uncomplicated human diabetes before there is danger of death from starvation. Even wasted and emaciated patients have borne fasting with apparent benefit, giving the impres- sion that they have been suffering more from intoxication than from a lack of nutrition;" and in another place, " among the patients treated thus far during a variable number of months, in the hospital and at home, spontaneous downward progress has not yet been observed." Today I doubt if he would feel justified in making quite so strong a statement, but certainly it is true that downward progress where the patient is under a careful regime is far less common than hitherto, and I feel more like agreeing with Allen today than a year ago. 2. Vegetable Days. "Vegetable" days or "green" days "rab- bit" or " Nebuchadnezzar" days as the patients sometimes call them were suggested by von Noorden as useful and less exacting than fasting days in the course of diabetic treatment. Their intro- duction was of much value, because many physicians would pre- scribe vegetable days who were afraid to order a fasting day, and many patients would take a vegetable day who refused to go entirely without food. The sense of satisfaction which one obtains from the bulk of vegetables is far out of proportion to the little nutri- ment contained. Unfortunately, in the use of vegetable days, large quantities of fat were often employed, and the benefits which might have been derived therefrom were to a considerable extent counteracted. So employed, a vegetable day, instead of being a day of undernutrition with a small amount of carbohydrate which would have favored the lowering of acidosis, became in reality a fat day, and in consequence the acidosis did not always decrease. Today vegetable days are employed without fat to lessen the acidosis, and probably with justification. 3. The Oatmeal Treatment. The use of oatmeal as a special form of treatment in diabetes was introduced by von Xoorden in 1903. l The sudden disappearance of sugar from the urine, despite the administration of so much carbohydrate, was at first looked upon with awe. Gradually with a better understanding of the storage of carbohydrate in the body and its utilization or non- utilization, as disclosed by the respiratory quotient, the mystery 'Von Noorden: Bcrl. klin. Wohnschr., 1903, xl, p. 817. 3C.S TREATMENT began to disappear. The actual "cure" is seldom employed today, but the experimental problems associated with it are so many that it deserves extended discussion. Formerly I occasionally employed it in the treatment of very severe cases of diabetes, but the apparent lack of the utilization of oatmeal as disclosed by the respiratory quotient discouraged extensive use of the same. Notwithstanding this fact, and although 1 have not treated a ease in this manner for over two years, I believe the oatmeal treatment has taught us much, and that eventually we shall benefit by the new idea it introduced. The "cure" was originally prescribed by von Xoorden 1 as follows: "The oat c.ure, as now prescribed by me, consists in the daily administration of 200 to 250 grams of oatmeal, best given in the form of gruel every two hours, and 200 to '500 grams of butter, and often about 100 grams of vegetable proteid or a few eggs may be taken in addition. Otherwise, nothing else is allowed, except black coffee, or tea, lemon juice, good old wine, or a little brandy or whisky. Such a diet is often disliked by the patient, but I have ahvays succeeded in getting over this difficulty. After three or four days upon it the purpose for which it was intended is often found to have been attained; in other cases the same program must be repeated two or three times. It is apparently advisable to precede the oat cure with a few days of restricted diet, or even one or two vegetable days, for when the cure immediately supervenes upon a mixed diet the desired effect follows rather late. TABLE 103. 'Vox NOORDEX'S ILLUSTRATION" OF THE "()AT CUKE." Diet. 1. Strict diet 2. Strict diet .'i. Strict diet 4. Vegetable day fj. Vegetable day (). < >at meal ( 7. ( )at meal 5. Oatmeal (t. Oatmeal 0. Veiietabli 1. Vegetal, 1. 2. Oatmeal :;. Oatmeal 1. Oatmeal :>. Vegetable f + Ammonia, Kins. 3.2 3.8 4.15 2.9 2.8 2.1 1.0 1 .5 1 .1 1 .:; l.s 0.9 (). ( .) 1.0 0.8 o.s 0.7 1 .0 0.9 0.8 "At the commencement of the oat-cure treatment one notices, it is true, even in the most favorable cases, an increase of the Von Nnnrden: I). H. Treat & Co., 1900, p. 190, SPECIAL DIETETIC METHODS 309 glycosuria; but after a few days the excretion of sugar diminishes and the acetonuria even more so. During the oat days the urine may often be quite free from sugar, and if it is not entirely free, one may be fairly certain that it will be so in the succeeding vegetable days. Table 103 gives the details of one such favorable case. "The estimations made before the oat cure was begun show plainly enough that it is a case of severe glycosuria combined with excessive acetonuria. With the most restricted diet it had not been possible to bring the sugar below 40 grams; even on vegetable days more than 20 grams were excreted. In the course of the oatmeal treatment the urine became free from sugar, and it remained so on the subsequent return to the restricted diet. It even appeared that small quantities of carbohydrate could be well tolerated, whereas for several months previously there had been no question of such a thing. 1 "I believe that a glance at Table 103, which is only one out of a large number showing the same thing, will suffice to show that a result has accrued which formerly would have been deemed impos- sible to obtain. Unfortunately, however, there are only relatively few cases in which the result is quite so surprisingly beneficial; in many others it is incomplete, although still satisfactory; in others again no result at all is obtained. The following fact is noteworthy: cases in which the results of the treatment were most beneficial relate without exception to the very severe forms of glycosuria; many of them were in children or young people. On the other hand, the result was almost without exception a failure in cases of slight glycosuria, the exact opposite of what might a priori have been expected. 2 The oat cure rendered me immense service in severe cases, and I may even say that I have often suc- ceeded in fending off incipient coma by its use." The positive proof of the value of an oatmeal day \vould be the demonstration that the carbohydrate in the oatmeal was oxidized in the body. As yet, such a proof has not been satisfactorily estab- lished. It does not suffice to show that the sugar represented by the oatmeal does not appear in the urine during the oatmeal period. One should demonstrate a rise of the respiratory quotient following the use of oatmeal. This would appear an easy problem. Thus my own respiratory quotient before breakfast on September 30, 1914, was 0.82, but forty-five minutes later, after I had eaten 00 grams carbohydrate in the form of oatmeal it rose to 0.90. With dia- betic patients the results were irregular but the evidence on the whole 1 From the data furnished in Table 163, this case would not appear severe today. This may be one reason why the improvement was so marked. 2 Perhaps the mild cases were not as energetically fasted beforehand as were the severe. 24 370 THE ATM EXT suggested a slight utilization. These are reported in detail in Section 1 1 . Benedict and I have found the problem most difficult and so have other investigators. Roily in a series of experiments tested the com- parative effects of oats, rye, wheat, lentils, and green cornmeal on diabetic patients. Unfortunately, few of the experiments were pre- ceded by control periods. Two of 'his cases lie considers severe. In Case I, at three, five, and six hours after 70 grams of oatmeal were administered, the respiratory quotient was 0.73. After 70 grams of wheat meal it was 0.76. The respiratory quotient of his Case V after SO grams of wheat meal was 0.71, after SO grams of rye mealwas().73, and after SO grams of oatmeal was 0.71 . \Ve have repeatedly failed to find any increase in the respiratory quotient immediately after con- siderable quantities of oatmeal were ingested. (See Cases Xos. 194, '2 K't. 332, ")(') 1 , 591 , 740 and 773.) We soon felt that experiments of a few hours' duration could not be decisive, for the oatmeal might be simply stored in the body and later oxidi/ed. The length of the experiments was therefore extended to twenty-four hours, but even then the oxidation of oatmeal was not satisfactorily proven by the rise in the quotient. Again our experiments were lengthened and by means of the respiratory apparatus a series of some eight observations was made daily upon two cases during two oatmeal days, but here, too, no consistent effect from tlie oatmeal was demon- strated. Such a series, however, is not complete, and the fasting quotient was determined on the morning of the third day with the result that in this case (Case No. 501) it was 0.70. These data are all recorded in the charts on page 145. It is very clear that much work must be done before decisive conclusions upon the utilization of oatmeal in diabetes are reached. The introduction of fasting treatment will simplify the problem by furnishing a method by which the patients can be placed in a more stable state of equilibrium prior to the test \vith the oatmeal. Allen and DuBois 1 in studies upon several severe diabetics find "no special influence of oatmeal in diabetes or special readiness of oxidation. . . . The respiratory exchange fails to account for all the carbohydrate that disappears. The behavior of the respira- tory quotient showed no important difference on the first day and on the third day of the oatmeal treatment." In the light of present knowledge the frequently striking suc- cesses of the oatmeal treatment in diabetes as practised during the last few years are seen to depend upon several factors of which the most important is the fasting or greatly restricted diet which preceded and followed the cure. A similar result could be obtained with any carbohydrate as Blum pointed out and as Klemperer's 'Allen and DuBoib: Arch. Int. Mcd., 1010, xvii, p. 1010. SPECIAL DIETETIC METHODS 371 experiments with dextrose and Benedict's and my experiments with levulose show. During fasting the sugar level in the body is lowered and the body is therefore in a more favorable condition to store carbohydrate and possibly to utilize it than under ordinary circum- stances. The fasting or semifasting which followed the oatmeal days may be of very great advantage. It is quite possible that carbohydrates stored for a long period in the body may ultimately be better oxidized than carbohydrates representing more nearly an overflow of the carbohydate reservoir which escape quickly into the urine. With this in mind it is easy to understand why Blum found small quantities of oatmeal worked better than large amounts in severe diabetics. He also showed it was more successful in mild than in severe cases. Yon Xoorden also made another observation which indicates that the oatmeal is simply stored. A patient showed no sugar during the two preliminary vegetable days and the three following oat- meal days, yet in the next three vegetable days 96, 100 and 32 grams of sugar were excreted respectively. The quantity of protein upon the oatmeal days is extremely slight. Thus, whereas the carelessly treated patient in diabetes ordinarily consumes far more than 100 grams of protein, if he takes even the full quantity of oatmeal, namely, 250 grams, he will receive not over 40 grams. The low quantity of protein is undoubtedly an important factor in the success of the treatment. The oatmeal is given without animal protein. It has been thought that animal protein is somewhat less well borne by diabetic patients than vegetable protein, but I do not believe the evidence hitherto submitted to be convincing. The difference in the rate of absorption of the two kinds of protein may be of importance. Interesting observations bearing upon this point were found in the literature by F. M. Allen in the experiment of Kichhorst, 1 who showed that dogs fed on nothing but starch gruel for a considerable period and then suddenly given a large quantity of meat show glycosuria on the first or second day of meat feeding, but not thereafter. Five of Eichhorst's dogs behaved thus, and the glyco- suria resulting from meat diet ran even above 2 per cent. Straub and Rosenstein found that preliminary meat feeding favored the occurrence of glycosuria in animals poisoned with carbon monoxide, but this did not take place if the animals were previously fed with pure carbohydrate, or abundant carbohydrate with little protein. Similarly, Selig noted that ether anesthesia was followed by glyco- suria in dogs fed with meat, but that it did not appear with a pre- liminary feeding of carbohydrate. In his study of the oatmeal 1 For this and other references in this paragraph, see Allen: Glycosuria and Dia- betes, 1913, p. 441. 372 TREATMENT treatment, Fulta observed tnat ])rotein added to the oatmetd treatment tended to the excretion of more sugar in the urine than could come theoretically from the added protein, and that meat was especially harmful as compared with vegetable protein. Indeed, some patients were more sensitive to protein than to car- bohydrate. In a recent publication, Bernstein and Falta 1 have shown that a carbohydrate-fat diet lowers the metabolism, and they explain this as a result of the attendant decreased protein metabolism. A lowering of the metabolism is therefore, according to them, to be expected in the oatmeal treatment, because of the small quantity of protein therein contained. This has been observed, but they point out that such a lowering is to be expected only when the carbo- hydrate is utilized and thereby body protein spared. Their data are most interesting. I am not ready to draw conclusions from Benedict's and my own experiments upon this same topic as yet, but I would point out that invariably in our levulosc experiments the metabolism rose after levulose (see pp. 3XO-3S4). The article of Bernstein and Falta is of much value, and I recommend its perusal not the less because of the criticisms it contains upon Benedict's and my work. These criticisms will receive careful analysis in our forth- coming ( 'arnegie publication. A gain in weight is usually coincident with the employment 01 the oatmeal cure. This is not peculiar to oatmeal, and is said to occur as a result of any carbohydrate day, though I now have some doubt as to the accuracy of this statement. Already attention has been called to this phenomenon of gain in weight on changing from a fat to a carbohydrate diet in normal individuals. Contrary to many observers, who have considered it a disadvantage that the patient develops edema, in most cases I think it is distinctly helpful. I cannot forget that patients with edema seldom or never develop diabetic coma. Falta noted a remarkable retention of protein during the oatmeal cure. The significance of this has not been explained, but it is probably in some way connected with the reten- tion of fluids by the body. A second explanation of the gain in weight may be the behavior of the kidneys, according to Barren- scheen. Tie injected human subjects intravenously with '20 c.c. of a 10 per cent, solution of lactose, and upon each of the following two days he gave 2.~>0 grams oatmeal. On the third day he gave a mixed diet, together with a repetition of the injection. Ipon the oatmeal days the excretion of the lactose* was delayed from one to five hours, which Barrenscheen attributed to slight renal changes, not otherwise demonstrable, caused by the oatmeal. A third reason for the gain 'Bernstein and Falta: Doutsch. Archiv f. klin. Mod., 1910, cxxi, p. 95. SPECIAL DIETETIC METHODS 373 in weight may be the high caloric value of an oatmeal day. This is not generally appreciated. It is shown in Table 164. TABLE 1G4. NUTRITIVE VALUE OF AN OATMEAL DAY. Carbo Substance. Quantity. hydrate. Protein. Fat. Alcohol. Calories. Oatmeal ... 240 100 40 10 . . 944 Butter .... 240 ... . . 200 . . LSOO Whisky ... GO ... . . ... 30 210 Total .... 160 40 21(3 30 2954 A fourth cause of gain in weight upon an oatmeal day is the considerable quantity of salt taken by the patient. For example, in the preparation of 240 grams of oatmeal the usual quantity of salt added by the cook is 10 grams and the patient may take even more. If to the oatmeal an equal amount of butter is added, according to the original advice of von Xoorden, the quantity of salt is increased by 0.3 grams, for butter contains on the average 2.51 per cent. salt. It would therefore seem quite likely that the edema which is associated with the oatmeal treatment might in great part be explained by the unusual quantity of salt given upon that day, and it is conceivable that along with the salt a considerable quantity of the carbohydrate of the oatmeal might be retained as well. A striking characteristic of the oatmeal treatment is that car- bohydrate is administered in only one form and it has been con- tended that this is one reason for its apparently better assimilation than the same amount of carbohydrate in several forms. This supposition may be true but there is little sound evidence behind it. Such a phenomenon might be explained by the simplicity and blandness of such a diet leading to very slight stimulation of the digestive glands in general, and the pancreas in particular. In fact Cohnheim and Klee have noted this peculiarity in oatmeal. This may explain why boiled oatmeal acts better than baked oat- meal. Allen has pointed out that if the external function of the pancreas is relieved of work the internal function may act more vigorously and the diabetic condition be correspondingly benefited. The explanation agrees with facts. Allen cites the work of Cohn- heim and Klee, who observed that the foods which caused the greatest activity of the external pancreatic function are the ones which give rise to glycosuria, and the foods which stimulate the external pancreatic function least are the ones which have least tendency to glycosuria. These writers suggest that the internal function of the pancreas is strengthened by relieving the strain upon the external function. Various other theories have been put forth to explain the oatmeal "cure," but none have been entirely satisfactory. Some have believed that there was a specific body in the oatmeal which might 374 TREAT ME XT act upon the liver or more indefinitely upon the general sugar metabolism and favor its combustion, but the evidence in favor of such a view is unsatisfactory. Again, the effect of oatmeal in rendering the- kidneys less pervious to carbohydrate has been urged as a reason for the apparent utilization of oatmeal, but even granting this to be the case, it does not seem probable that all the carbohydrate il2.")0 grams oatmeal daily for three days =750 grams oatmeal = ">()() grams carbohydrate) given upon three successive oatmeal days could be held back in the body, although it is true that Case No. '144 (see page 410) showed a positive carbohydrate balance while undergoing an oatmeal cure under von Noorden's care. It is noteworthy that this patient never became sugar-free after this cure save for occasional days, despite vigorous dieting. Furthermore, an explanation of this character would not explain why the oatmeal treatment occasionally fails. Still another explana- tion advanced by Lipetz in 190") is based upon the possible rapid fermentation of the oatmeal in the intestines, preventing its absorp- tion as sugar, but instead as decomposition products of the oatmeal. \ arious writers have looked upon this explanation favorably, but the general trend of opinion discountenances this theory. You Noorden has emphasized the absence of flatulence and has pointed out the excellent digestion of the normal stools with the oatmeal treatment. Allen has critically examined the evidence of Rosenfeld, Klotz, and others who have endeavored by a series of experiments to support the fermentation theory, and concludes that "von Noorden is correct in holding that bacterial processes are comparatively slight in the upper part of the small intestine, and that a practi- cally complete fermentation of the large quantity of carbohydrate administered, without symptoms on the part of the patient, is unthinkable." The acidosis frequently decreases or disappears following or during the oatmeal cure. If such a reduction was constant it could be used as an argument in favor of the utilization of oatmeal, but unfortunately it is not. Furthermore, the tests for acidosis were far less accurate when the oatmeal cure was in vogue and con- sequently the data upon this point are not as complete as one would wish. 1 nfortunately, too, for settlement of tins question the treatment was complicated by the use of fat in the form of butter which would probably act exactly opposite to the oatmeal. Y\ith Case No. :!44 the mild acidosis disappeared upon the oatmeal days in April, 190S. but changed only slightly on a single oatmeal day in September, 1910, and August, 1911. There was decrease in Cases Nos. L'lli' and o71, J though this mav have been due to defective SPECIAL DIETETIC METHODS 375 elimination for the urine fell from 4035 c.e. before the oatmeal day to 3330 c.c. upon it in the former instance and from 2520 c.c. to 1430 e.c. in the other. There was also a striking decrease in the aeidosis with Case Xo. 441, l also with Case Xo. 289. 2 The ferric chloride reaction fell in intensity and upon a subsequent test August 23 and 24 the ammonia dropped from 3.2 grams to 1.7 grams. (a) The Place of Oatmeal in Present Diabetic Therapy. It must be recognized that the employment of von Xoorden's original method for the administration of oatmeal in severe diabetes has often been followed by good results. Taking advantage of what has been learned since the introduction of the oatmeal treatment, the indications (?) for its use in diabetes are as follows: 1. To cause a retention (increase the storage) of water, carbo- hydrate, and protein in the body. 2. To decrease the aeidosis. 3. To afford rest to the digestive organs, especially the pancreas, with the hope thereby of improving the tolerance for carbohydrate. 4. To study the problems involved under the three preceding headings. For this purpose it is an ideal food. 5. In a modified form it may be useful in replacing a fasting day once a week. The employment of oatmeal in the "oatmeal cure" has stimulated its use in smaller quantities, and it is undoubtedly one of the most valuable additions to the strict diet of diabetic patients. It can be given in many different forms, it serves well as a vehicle for butter and cream, and, finally, is useful as gruel in the treatment of indi- gestion or diarrhea. It occasionally happens that a patient likes oatmeal to such an extent that he is willing to live almost exclusively upon it for weeks or even months. I have seen one such patient. Upon this diet he did very well. Unfortunately, such cases are rare. In prescribing oatmeal the dry weight should be the measure employed, because different brands vary enormously in bulk and hence in content of carbohydrate when cooked, because of the water employed. How wide these variations are the following table illustrates: TABLE 165. WEIGHTS OF DIFFERENT VARIETIES OF OATMEAL UNCOOKED AND COOKED. UNCOOKED. COOKKD. Quantity, Weight, Xo. tablespoonfuls, Weight, Xo. tublespoonfuls, 1 Kill. grains. heaping. grains. he-aping. H-O oatmeal . . 32 4 353 3-| Quakor oats. . . 37 5 325 4 Scotch oatmeal. . 81 5 Sol 10 American oatmeal 84 3| 690 6 Irish oatmeal . . 96 5 853 15 1 Benedict and Joslin: Carnegie Publication No. 176, Case P, p. 42. 2 Ibid., No. 136, Case J, pp. 23, 132, 176. o/O TREATMENT 4. Other Carbohydrate Cures. (//) Wheat. Blum 1 believes wheat flour acts just as efficaciously as oatmeal in a carbohydrate cure, and Csonka 2 has observed no difference between the utilization of starch of wheat and the starch of oatmeal flour by completely phlorizinized dogs. Blum demonstrated that the effect of the oatmeal or wheat was most marked in the mild cases. He noted that patients having a positive carbohydrate balance of 70 to SO grains could take 200 to 250 grains of oatmeal with a similar quantity of butter and occasionally three or four eggs or 50 to 75 grains of vegetable albumin, and after living upon the diet for the customary three days, and then having a vegetable day, the last trace of sugar would disappear. Should the diabetes be a little more severe he employed 125 to 150 grams of oatmeal, but the same amount of butter. In still severer forms of diabetes with acid intoxi- cation only 100 grams of oatmeal were allowed for a day or two, then only 75 grams for a few days, and finally a vegetable day. In other words, he was fasting his patient to a moderate degree. (l>) Potatoes. A potato diet was advocated years ago by Mosse, 3 and in fact was the first of the carbohydrate cures recommended in diabetes. From what is known now it is easy to understand why a potato diet frequently worked well. (1) The potato diet was an undernutrition diet because no emphasis was laid upon the simultaneous use of fat; (2) there is little protein in a potato; (>) potatoes arc 1 a bulky food, and so satisfy the patient's appetite. The claim that considerable quantities of alkali are thus intro- duced into the system does not rest upon a firm foundation. I have had little experience with potato, but in 2 cases (Xos. 7(15 and 806) iu which it was employed, respiratory quotients gave no evidence that it \vas assimilated. (Table S2.) In small quantities potato, like oatmeal, is most valuable. It carries butter well, contains only about one-third the quantity of carbohydrate in bread, and is easily measured. A potato the si/e of an egg weighs about 00 grams, whether cooked or uncooked, and contains 12 grams carbohydrate. \\ith a good many mild cases of diabetes it is advantageous to exclude bread entirely from the diet and to substitute' potato, of which the quantity need be only slightly restricted. One hundred grams of carbohydrate in the form of 100 grams bread are quickly eaten, but it is not so easy to take the same 100 grains carbohydrate' in 500 grains of potato. Little potatoes, carefully cleaned, when baked are often eaten with the skins by patients with much relish, and with relief to their constipation as well. 1 Blum :.Soinuino Modioulo, 1!*13, xxxiii, p. .'-il.'i. s Csoiika: .lour. Am. Mod. Assn., 1910, Ixvii, p. 1114. 3 Mo.-or: ROMIO de mod., l'JU_', xxii. up. 1U7, 279, 371, 020. Cited by Xaunyn. SPECIAL DIETETIC METHODS 377 (c) Bananas. Bananas have been advocated by von Xooraen as a substitute for oatmeal. Like potatoes they contain about 20 per cent, carbohydrate. Most of the carbohydrate is in the form of starch, but when the banana thoroughly softens and ripens this changes to sugar. Bananas are usually eaten in the starch stage. The carbohydrate in one banana is about 20 grams. The average weight of 12 whole bananas was 141 grams and the weights varied between 119 and 167 grams. When peeled these same bananas weighed on the average 87 grams, the minimum weight being 73 grams and the maximum weight 101 grams. It is therefore approximately correct to consider that a whole banana weighs about 140 grams with the skin on and not far from 90 grams peeled. (See Table 127, p. 264.) For the same reasons that I employ potatoes, I prefer to give bananas rather than bread to the patients. It is so much easier to eat too much bread than banana. ((/) The Milk Cure. Milk was advocated as an exclusive diet for diabetic patients by Donkin 1 some years ago, and recently attention has been again called to it by Williamson. 2 It has not come into favor. Unfortunately, it is so simple to prescribe a milk diet that this is often done, and it is not uncommon to see patients taking large quantities of milk with distinct harm to themselves. It was not intended by the original promoters of the milk cure that other foods should be simultaneously eaten, but, as so often happens, the original directions have been overlooked. Still more harmful than the indiscriminate use of milk has been the advice to drink buttermilk. Buttermilk contains all the carbohydrate which is in ordinary milk, but the fat which contains so much nutriment for the diabetic has been removed. Buttermilk might be employed temporarily, but as an addition to the strict diabetic diet would be harmful. Case Xo. 17 is the only one in my series which has shown an apparent tolerance for milk: Male, teacher, onset of diabetes at the age of fifty-five; came under my observation in August, 1900, at the age of sixty, having lost 27 pounds, 1(5 per cent, from his highest weight. Upon a diet of 3000 c.c. milk, containing 150 grams carbohydrate, which he had employed with only trifling additions for a period of six weeks, the sugar in the urine was only 13 grams. During this period weight fell 1 kilo. I endeavored to make him sugar-free, and lowered the carbohydrates in the diet to 25 grams, but the sugar in the urine fell only to 9 grams. I well remember having greatly increased the protein and fat in the diet when the milk was omitted. Undoubtedly the secret of the favor- 1 Donkin: British Mcd. Jour., 1S74, i, p. 838. 2 Williamson: British Med. Jour., 1915, i, p. 456. 378 TREAT MK XT able course of the patient upon the milk diet was the comparatively small number of calories which he obtained and, conversely, the harmful effect of the rigid protein-fat diet was due to the large number of calories it contained as well as to the acidosis which my records of sixteen years ago show it brought on. I low plain the explanation of this case is today, but for a long time it was a puzzle. (c) Levulose. Since the publication of our second monograph in I!)!.?, F. G. Benedict and I have continued our studies upon the respiratory metabolism of diabetic patients. Save for references to this in my Harvey Lecture, 1 the material has not been published. We have felt it was better to accumulate data and correlate the whole in a single report. The utilization of carbohydrates has been our chief problem, but we have also included the study of other subjects if they appeared of sufficient interest to warrant divergence from the main theme. In the interim of these five years, much helpful work has come from other laboratories, and the measurement of body surface, the studies upon fasting and other individuals have introduced new conceptions as to the methods of determining the basal metabolism. It was not my intention, and indeed I did not think it wise, in this book or at this time to refer to the work which Benedict and I and our associates have done, but the interest which the more recent experiments with levulose must awaken in thinking patients and doctors led me to ask Dr. Benedict for the privilege of reporting 4 severe cases of diabetes in which levulose has appeared to do good. To my request Dr. Benedict instantly and whole-heartedly assented. The significance of these experiments lies in the fact that whereas these patients with whom the tests have been tried formerly fasted one day a week, we substituted for the weekly fasting day a levulose or levulose and oil day, thereby securing for them additional calories; that in each case the total metabolism and in most cases the respiratory quotient rose; that ni)on these days the urine remained sugar-free, or if any sugar appeared it promptly disappeared even if the full diet of the day preceding the levulose day was resumed the following morning; that in some instances the blood sugar was lower the day after the levulose than it was before the levulose was taken and seldom at all higher; that following the treatment or in spite of it, - cases of diabetic inanition, from being unable to stand, were in the course of several weeks able to get out of bed and walk, and that in each of these cases the eU'ect upon the carbohydrate tolerance was favorable. In these cases acidosis, though formerly existing in i>, was absent when the levulose was given. How le\ ulose acts in the presence of acidosis will be later discussed, as will also the eil'ect 1 Jo^lin: Luc. oil., p. HO, SPECIAL DIETETIC METHODS 379 of the substitution or addition of protein. In nil, Benedict and I have made 44 experiments with leviilose. These must be analyzed as a whole before conclusions can be drawn, but to the keen eyes of those working upon the respiratory metabolism, other hints of peculiarities in the metabolism of these patients will be disclosed than those already mentioned, should a careful examination of the data in the charts be made. The selection of leviilose for these tests was partly by accident. Another form of carbohydrate may act as well. I have resisted the advice and temptation to try such until recently. I have seen so many methods of treatment fail in diabetes that if anything promises well, I prefer to abide by it until its utility has been proved or disproved. In our early cases leviilose was given alone one day a week, and examinations of the respiratory metabolism as well as of the blood and urine were made both before and after its adminis- tration. In later cases the same plan was followed, but upon the theory that if carbohydrates are burned fat would be burned, too, olive oil was also given. In no instance has the patient shown an acidosis upon such a day, or an increase of sugar in the urine above that of a levulose day pure and simple. By no means do I recommend the general use of levulose in all cases. It is quite possible that these cases have done better than would others. To show how conservative I am in regard to the whole question, I am selecting a new case for observation only every two or three weeks, and each case is being investigated by Dr. Benedict and me with all the facilities in our control. Indeed, one of the chief reasons for publishing these cases at this time is to evoke constructive criticism which will help in the laborious analysis of the past five years of work, and in future experiments. From the therapeutic stand-point I wish others would repeat or continue in a better way our experiments. If they do not, it will take too long to find out all the faults or advantages of this plan of treatment. Furthermore, I dislike the thought of having my patients bear the brunt of all the tests. These experiments remind me of my Italian diabetic years ago at the Boston City Hospital, who went through his typhoid fever successfully on a diet of oatmeal and olive oil. The story of the 4 cases follows: Case Xo. 1190, having had diabetes for seven years, came to me in December, 1910, minus fifty-three pounds of his former best weight, with G grams of ammonia and an alveolar air of 22 mm. mercury. For his treatment at this early stage my assistant, Dr. Ilornor, deserves all the credit. Proof that alkalis are unnecessary unless the ammonia exceeds 6 grains, and the alveolar air is below IS is evident, because the soda he had been receiving was omitted w M t- U >> g 'juoponb Ajo; o -uiui jod -j-a Ji H C 'BUISBJjJ 1 IN -^C^Tfl'IA\ 4110,) jod 'ji:3ns souopiQ -r 7i .- c~. -r -t 71 71 -H O rH ,-H o coo o d +T i 7 i 77 i = i i = i +++++++ - co X' cr. o; o o o o ]oqoo[\- i | o cooo -o oo7iiooo-7st^ >s-2 ^BS"?-^ ai u 'a a S>< C a)^ 1 - c P5.-K B C5^.l grams in one period of twenty-four hours. It is true that he had a few vagaries, such as a passion for saccharin tablets (21, each of one-half-grain strength per day), bouillon cubes, cocoa butter, salt, celery salt, various peppers and similar delicacies furnished by thoughtful angel visitors, but he assures me that nothing was taken to account for so high a nitrogen, and he fully realizes the importance of the statement. 1 believe the nitrogens trustworthy. It will be seen that all the levulose was not tolerated, but the quantity of sugar appearing in the urine was comparatively small. The chart shows the progress of this patient; he is gradually gaining tolerance, and I have transferred his case to Dr. Brigham and Miss Lundholm. In the following cases levulose was given on one day each week, replacing the fast day, for several weeks. The quantity was limited to one gram or one gram and a half per kilogram body weight. Case Xo. 032, Table 107. A young officer, aged thirty-five years, with diabetes of one and a half years' duration came to me first in 1913. At the hospital diacetic acid showed repeatedly, and the am- monia was 1 .7 grams, but the tolerance for carbohydrate lay between 1 5 and 30 grams. Nevertheless he was discharged with 0.5 per cent, of sugar in the urine, and diacetic acid was present, with a diet of 30 grams carbohydrate and a limited quantity of protein, though with an unlimited amount of fat. He returned in February, 1910, and it required twelve days to rid the urine of sugar and twenty-one days to rid it of acid. Table 107 contains the result of the adminis- tration of levulose, levulose and oil, and protein and dextrose on one day weekly instead of a fast, day. Kxcrcise was utilized to the limit, and, as to be expected of an army man with a Victoria ( Yoss, obedience was implicit, cooperation ever present, and system exact. I have permission to publish this letter received eleven months after leaving the hospital. M.vucu 8, 1!)17. "1 have really been wonderfully well, feel splendid and everyone remarks ho\v well I am looking. Tests have shown a, slight trace of sugar on three mornings since October S, last; all other times absolutely sugar-free. Mv weight doesn't change at all SPECIAL DIETETIC METHODS 383 if anything I have gotten very slightly lighter. I weigh -from 1242 to 12o2 pounds. I still stick absolutely rigidly to my routine, but I have gotten up to 30 grams carbohydrate per diem; that is, on the last five days of the week I take 30 rest of diet the same. The last three weeks I have been taking 15 grams oatmeal for breakfast on Monday, Tuesday, Thursday, Friday and Saturday mornings, TABLE 168. CASE No. 1213, P. 384. LEVULOSE, CHANGE JUICE, ADDITION OF PROTEIN, CLIVE OIL. Date, 1917. _o M _M 'o & 31 'l 30.7 30 . 6 30.5 30.3 34.2 31.4 30.8 31.3 Urine. Diet in grams. Blooc Respiratory 1MOOC1 - metabolism. oJ 3 [3 -r a -, o o - '~ 5 3 c a c h Z 5 C SL -j. 3 tr. 2 2 2 _c - -- > C 1 C l ( ( .':' ( c ' c r. PK c ,c L *L* _0 O u F* a c - bl 5. Fat. o _S 3 >, 'E So Total fatty acids. L -t>- ^S" | 1 _o "3 1 d C a E c ~:. n 8 O o c 1 g _rj ^ t- c & c o & c, "3 5 5. - % ~ J fe 6 o ^ S o o Before experiment Before levulose Levulose given After levulose . Before experiment Before levulose Levulose given After levulose, oil 30 Before levulose Levulose given After levulose, oil 60 Before experiment Before levulose Levulose given After levulose, fish 240 ... Before experiment Before levulose Levulose given, oil 60 After levulose . Before experiment Before levulose Levulose given, oil 60 .... After levulose . Jan. 5 6 7 8 9 10 11 Feb. 3 4 5 8 13 14 15 1C 20 21 22 27 28 M-ir 1 2700 1500 1950 2500 2800 2350 2000 2100 1XIK, 1900 1400 13.2 T'.G 12 30 12 50 5s 5s 39 :>7 16 37 1-' _>L> II n 50 (12 5s 24 41 ( ( 14 14 LI 14 48 228 366 560 (119 682 824 768 308 615 .26 .20 .21 0.47 O.C5 .21 .31 . 32 .30 .27 .36 .14 7s 93 31 il 31 '. 3 2454 2450 2500 2 GO 240 101 120 104 134 99 110 93 11- 0.81 0.8 O.M) 0.87 0.97 0.91 91 8.5 2 5 29 38 37 76 11 1-1 555 946 . ~O S4 1295 . . 1 60 . . 30 5 545 113 31.7 31.0 31.8 31 '3 31 ' 8 33.3 1400 2800 1700 6.6 1 5 3959 14 S05 4 '23 35 14 421 649:83141061 8fi 1535 1700 1600 i'.b 1 tr. (30 6 2 0. . . . 60 4063 4679 5 11 1 1 815 813 .']-> 1001 . . . 2S- 107 84 . . 1885 7 6 IK 5 347 OS- 1 17 58 3'1 :\:, 47 17 CM iq'in." 2 3 4 5 (> 8 14 32.71900 32 . 5 2800 31.82300 32.41,800 32.62100 2 3 2 tr. tr. 640 648 045 037 037 74 14 948 81 141043 7514 953 7614, 930 7614 930 .2P 88 105 118 111 132 0.84 1.00 . 84 1.91 1210 2 60 6 60 5 815 . lie W.7 1700 5 _'7 61 1 1 775 .23 . ' . 93 31.7 11(10 (] tr. 60 OiiO 5 815 . . I . 1 . 39 . 24 .43 ,4s .32 .52 .OS .21 . 119 . 91 1 . 18 . 34 . 42 . 39 . 48 .51 . 62 . 28 384 TREATMENT Wednesday all carbohydrate in .") per cent, vegetables and cream, Sunday (fast day) all carbohydrate in 5 per cent, vegetables." Case No. 11213. This little patient of seventeen years, so weak that the students carried her up to the amphitheatre in the Peter Bent Brigham Hospital that I might show them a case of inanition, and I carried her upstairs on her return to the Deaconess Hospital, first came to me on January 5, 1 ( ,)17, weighing sixty-eight and a half pounds, with the story of having lost fifty pounds in the preceding year, the onset of diabetes being placed in May, 1010. She showed 3 grams sugar on a fasting day, but no acid, and said she could not tolerate protein. I accepted the challenge, gave protein, and found she had 50 grams tolerance therefor, in the presence of 40 TABLE 160. CASE Xo. 123.'}, P. 3S5. LEVULOSX, OKA.\<;E JUICE, ADDITION' OF OIL ou BUTTER. Before experimf Before levulose Levulose tnven After levulose Before exporin Before levulos. Levulose tiivei After levulose, After levulose I'jrul of experi Before experi Before levulo: < (ranges (jivei After leviilo.-e Next inornin SPECIAL DIETETIC METHODS 3X5 grams of fat. But she lingered along, as her chart shows, and I decided to give her levulose on her fasting day and with what result the record makes plain. Apart from the documentary evidence this afforded, I may add that in April she walked between one and two miles in a day. Each day massage was given by Mr. Sundelius or the patient's mother. Her weight is now 71 1 pounds and her diet carbohydrate 10 grams, protein 36 grams, fat 83 grams. Case Xo. 1233, aged twenty-seven years, had been fasted and fed alternately since his diabetes began fourteen months ago. He differed from the other 3 cases in that he showed edema, and the blood-pressure was 110 systolic and 80 diastolic. The urine showed only the slightest possible trace of albumin, and no casts; the phenolphthalein excretion varied between zero and 24 per cent., and the non-protein nitrogen was successively 110, 95 and 120 mg. per 100 c.c. blood, but eight days later 28 milligrams. He, too, tolerated levulose and levulose and oil, later oranges and oil, oranges and butter, to a surprising degree each scheduled fast day, and his edema finally disappeared. He was discharged upon the following diet: carbohydrate 30' grams, protein 70 grams and fat 112 grams, with the advice to take carbohydrate 20 grams and oil 20 grams three times in the day on one day each week instead of a fast day. (Table 1G9.) 5. Rectal Injections of Sugar. The experiments of Arnheim in 1904 showed that glycosuria in diabetic patients was not increased after enemata of solutions of sugar. Simultaneously, acetonuria diminished, and this appeared to be good proof that the sugar was absorbed and oxidized. This favorable action was attributed to the slow absorption of the rectal injection. Since this time others have studied the problem, notably Bingel, Reach, Ballint, Liithje, Jahnson-Blohm and Petitti. 1 The subject is, however, by no means settled. The whole matter should be reinvestigated, for the conditions are now much more favorable for its solution. In the first place, it is questionable whether in former experiments one sufficiently considered the effect of fasting, and secondly, the possibility of carbohydrate storage following fasting. Finally, today the respiratory exchange is far more generally employed, and this is really the ultimate test. Experiments dealing with rectal alimentation of normal individuals will throw much new light upon the question, and these have been in progress for some time under the direction of Dr. Carpenter of the Nutrition Laboratory. Dr. Carpenter has been kind enough to give me the following preliminary summary of some of his tests. From these it is evident that absorption of the sugar is between 31 and 72 per cent. (Table 170.) 1 SOP Allen, Loc. cit. 25 3Mi TREATMENT TABLE 170. PRELIMINARY SUMMARY OF RESULTS OF INJECTION BY KECTUM OF LEVULOSE AND DEXTROSE SOLUTION.' (T. M. Carpenter, Ph.D.) Amount given, gins. Amount of liquid, Time retained. llrs. Min. Amount in washout, Kins. Amount absorbed, Kins. Remarks. E XPEli I MEN T.S WITH C OMMERCIA L LEVULOSE. 50.0 500 3 30 18.0 32 . One washout at end. 50.0 500 2 38 25 .4 24 . 6 Defecation and washout. 50.0 1000 3 20 16.7 33 . 3 Defecation and washout. 50.0 1000 2 13 21.6 28 . 4 Defecation and washout. 25.0 500 3 12 7.8 17.2 Two washouts at end. 25.0 500 2 48 3.8 21 .2 Two washouts at end. 25.0 500 3 12 1.8 23 . 2 One washout. 25.0 500 3 00 2.7 22 . 3 Two washouts. 37 .5 750 2 37 0.7(?) One washout. V. YPT.'DTM I.' NTS W TTTr K MILBAUM <1T rnrx /rcKKR (GEREINIGT). 00.0 1000 5 00 24.0 36.0 Two washouts at end; much cramps. 30.0 500 3 49 11.3 1S.7 Two washouts; subject com- fortable. 30.0 500 4 46 S.O 22.0 Two washouts; no cramps; so7iie peristalsis. 30.0 500 3 47 5 . 6 24.4 Two washouts; comfortable. 30.0 500 3 33 11.1 18.9 Hard to retain solution; one washout. 30.0 500 6 11.0 19.0 Very comfortable; two wash- outs. 30.0 500 4 9.0 21.0 Reported gas pains; two wash- outs. 30.0 500' 4 5.2 24. 8 Comfortable; two washouts. Liithje 2 repeatedly witnessed the disappearance of aeidosis during the treatment of patients with enemata of sugar. As evidence of the absorption of sugar, lie observed an increase of the percentage of sugar in the blood, and controlled the observation by noting no such increase following the administration of salt solution. The slow absorption of the sugar did not account for its better utilization when given by the rectum, because when sugar was introduced into the body at the same rate by being slowly sipped, the difference in favor of rectal alimentation persisted. Liithje found that with good technic patients may absorb one or two liters of solu- tions of sugar containing 5.4 per cent., and thus may get ">() to 100 grams a day. Ten patients were treated by him by this method. Hergmark also showed that after enemata of dextrose there was an increase in the carbon dioxide production, and that the experi- mental aeidosis diminished as well. Recently, Jalinson-Blohm found that with healthy subjects the instantaneous as well as the 1 Nearly all of the respiration experiments showed a more or less positive rise in (lie respiratory <|iioticnt after the injection of the sutrar solutions (February 2, 1917). Liithje,; Therapic tier Gcjrfiiwart, 1U13, liv, p. 193. TREATMENT OF ACID INTOXICATION 387 repeated doses by mouth gave inside of the first three and one- fourth hours a strong increase in the glycemia, which rapidly sank in several cases so far that hypoglycemia took place. Dextrose injected by rectum gave blood-sugar values which lay very near to those at the beginning, and inside the limits of error of the method. The experiments with the diabetic subjects showed a strong rise of the blood-sugar curve with ingestion of dextrose by mouth. As with the normal subjects, the rectal experiments gave no increase in the percentage of the blood sugar. An increase in the urinary sugar did not occur either with healthy or diabetic subjects after the rectal feeding. This work well shows the unsettled state in which the question now rests, but Benedict's and my experi- ments with levulose and other carbohydrates will help in its solution, as will also the simplification of treatment which fasting has created and the more accurate knowledge of diabetes which refined analyses of blood, urine and the respiratory metabolism afford. I. THE TREATMENT OF ACID INTOXICATION AND DIABETIC COMA. In the plan of treatment of diabetes outlined thus far, emphasis has constantly been placed upon the prevention of acidosis and coma. To this end it has been recommended to omit fat from the diet as a preliminary to fasting, then fast and resume the diet, by beginning with carbohydrate, following with protein and ending with fat. These procedures often suffice both to prevent acidosis and to abolish it if present; but when the acidosis threatens to be severe, the following rules, now in force for my cases at the New England Deaconess and Corey Hill Hospitals, are suggested. They will usually avail even at a stage when the patient is conscious, but ex- hibits the labored breathing of diabetic coma. They have not been successful when the patient has been unconscious, or the CO 2 in the alveolar air has been below 15 mm. mercury, though I am confident that some cases even of this type will eventually be saved if the orders are faithfully followed. These rules are: 1. Nursing. Provide a special nurse for the patient for both day and night, and preferably one trained in diabetic work. 2. Bed. Keep the patient in bed and warm. Avoid loss of calories through exertion or exposure; if restless, protect from be- coming chilled by flannel night-clothes. Every effort should be made to allay nervousness and discomfort. 3. Care of the Bowels. Move the bowels by one or more enemata. Cathartics should usually be avoided for fear of causing diarrhea. 3SS TREATMENT 4. Administration of Liquids. Give 1()(K) c.c. of liquids within each six hours. The liquids are to be given slowly, hot, as coffee, tea, thin broths, \vater; if the prospect is dubious of giving so much liquid by mouth, salt solution or tap water is to be given by rectum; if this resource fails, the nurse should call the doctor to give intravenously the balance of the liter, which remains not given for the period. (It will seldom be found necessary to give more than 1000 c.c. liquids, thanks to the avoidance of alkalis.) In order to secure the introduction of sufficient liquid in the first six hours, the cleansing enema at the beginning of treatment should be followed after half an hour by an enema of 500 c.c. salt solution in all cases as a matter of precaution. f>. Diet. ~ If the patient has been accustomed to the fasting method of treatment, begin or continue the fast, but if he has been upon a full diet, give a gram of carbohydrate per kilogram body weight during the twenty-four hours in the form of orange juice or gruel (oatmeal) made with water; whichever course is adopted, it is to be followed until danger is over. (i. The Heart. Sustain the circulation with the help of digitalis. ( 'affein may be given subcutaneously or as black coffee by the rectum. 7. Alkalis. Avoid alkalis. If such have been previously given, omit at the rate of '.>() grains a day. The results obtained with these measures will be briefly sum- i 1 . Table 171 shows that there has been a steady fall in the number of patients who have developed acidosis after entrance to the hospitals and is compiled from tables 1-47-140. TABLE 171. ILLUSTRATIVE OF THE PKEVEXTIOX OF ACIDOSIS IN TMHEE SUCCESSIVE (lnori's OF THIRTY PATIENTS KACH ADMITTED TO THE NEW ENGLAND DEACONESS AND COKEY HILL HOSPITALS. Group. 1. April. HI1.1, to March, 1010 ... 7 If) . . 4 (i 2 112 ~1. The following cases of severe acidosis have been treated in general according to the plan outlined above, and have been dis- charged alive. (Table \~'l.) 1 Sec also Causes of Death in Section I, p. fi-1 ; the Fatal Cases of 1'JKi, p. 330, and Ca.-c. 1 .- rnsiicressfuily Treated l>y Fasting, p. 3.10. TREATMENT OF ACID INTOXICATION 389 TABLE 172. -RECOVERY OF CASES THREATENED WITH COMA. No ALKALIS EMPLOYED. Case Xo. 7tl,-wwl rvi Alveolar air Date. FcCl. XI h. | mm Hg C 2 ' 983 .... Feb. (sec p. 114) 1, 1910 ++ 4 2 ++ 4 4 2 23 22 3 + + + 4 25 1209 Jan. 2, 1917 32 24 (see p. 402) i 780 June 11, 1910 + + 3 9 i K /on (see ]). 217> 12 + 3 2 24 21 14 81. 2 o : 55 25 705 .... Dec. 0, 1915 + + + 3 3' 21 (sec p. 314) : Jan. 24, 1910 + + + + 2 21 25 + + + 2 24 942 July 12, 1910 4 4 20 20 (see p. 395) 13 ++ 3 ~ 17 900 Jan. 5,1910 + + + + 3 ~* 25/24 974 Jan. 24, 1910 + 23 (see p. 304) 990 .... Feb. 27, 1910 ++ 1 5 23 1005 .... Mar. 0, 1910 1 1 22/24 (sec p. 347) 1011 .... Mar. 29, 1910 ++ 1 8 23 Apri 13, 1910 22/22 1085 .... Oct. 30, 1910 ? 1 7 28 20 (see ]>. 340) 1190 .... Dec. 8, 1910 + + + + 4 0-' 21 (see p. 379) 9 + + 20 22/14 10 + + + 3 3 18/20 11 + 3 1 20 20 12 + + 3 3 21 13 + + 2 24 1070 .... June 23, 1910 ++ 1 9 : >1 20 (see ]). 217) 1103 . . . . AUK. 9, 1910 + + + + 2 1 22 21 1120 3 .... Sept 0, 1910 + + + 21 (see p. 338) 7 + 18 Oct. 11, 1916 ' ++ 1 9 22 j 14 20 27 + 15 3. By former methods of treatment in which alkalis were generally employed to combat acid intoxication, G4 4 per cent, of all my fatal cases of diabetes succumbed to coma; but with the partial adoption of the present method the total figures for my cases have already fallen to 00 s per cent., and for the fatal cases during this last year to 44 per cent. It should be noted that many of these I did not see for some months before death. 1 Fourteen hours. 2 Eighteen hours. 3 Acidosis disappeared, died of inanition November 26, 1910, siijiar- and acid-free. 4 Up to December 1, 1915. 5 December 1, 1916. 300 T RE ATM EXT 4. When such measures as those indicated here have been adopted, I know of no case of diabetes, either of my own or reported in the literature, which has shown a dextrose-nitrogen ratio of 3.(>5 : 1. Far be it from me to say the present treatment is ideal, but I believe it a step in the right direction and do not intend to abandon it until I know of a case of diabetes of seven years' duration in an individual of thirty-five years of age or over, who shows an aeidosis of greater intensity than that of Case No. 11% (see p. 3rr ,sr which is harmful, but the sudden- ness with which the ratio of fat and carbohydrate in the diet, whether in health or disease, is changed. Six grams of fat in an egg were given to a diabetic child, recently made sugar-free, and caused an alarming aeidosis, but a few weeks later the same child thrived on .">() grams of fat a day. (See p. 4S.~.) (c) Administration of Substances which Favor the Combustion of Acid Bodies.- For many years the sudden change from a diet poor in carbohydrate to one rich in carbohydrate has been observed to occasionally result in the apparent decrea.se of acid poisoning. Von Xoorden employed the oatmeal cure for this purpose, and I think all must allow that desperate cases of diabetes have been rescued from immediate coma by such means. It is a question, however, whether this improvement has been due to the better oxidation of the oatmeal during such a cure, or secondarily to the retention ot body fluid which it involves. Then, too, there is always the possi- bility of the decrease in urinary aeidosis being simply apparent TREATMENT OF ACID INTOXICATION 393 and due to a diminished excretion of urine. The evidence accu- mulated is not extensive enough to settle this point. Experiments referred to by Falta, 1 as well as several experiments conducted at the Nutrition Laboratory suggest that upon the last days of an oatmeal cure a certain, although not large, amount of carbohydrate is burned, as shown by the rise in the respiratory quotient, which is our best evidence upon this point. This rise in the quotient is so slight when compared with the clinical effect sometimes obtained from oatmeal that it hardly seems probable that the two stand in relation. It certainly does not seem rational to give a diabetic patient with a blood sugar three times that of a normal individual still more carbo- hydrate, and I hesitate to do this. On the other hand, if the blood sugar is low and nearly within normal limits and the acidosis is increasing, the administration of 150 grams oatmeal (100 grams carbohydrate) or 100 grams of levulose might be indicated. For the present, bedside clinical experience must settle the question and incomplete laboratory data should not be given undue weight. (See Cases Xos. 1196 and 032, pp. 380 and 381.) The administration of alcohol was shown by Neubauer 2 to decrease the acidosis of a diabetic patient, but this writer pointed out that it did not have this effect in health. Allen, in his early report of severe cases treated with fasting, utilized alcohol for this purpose, desiring to avail himself of every means which had been advocated with reason to combat acidosis. I understand that in later cases he has employed it less. My experience has not been encouraging and I have prescribed it but little. My data do not show whether its use is advantageous or not, but my impression is that it is not of great value. I give it, if it is a comfort to the patient, but always bear in mind the possibility that it will upset the stomach, which it frequently does, and often wonder whether, when given during fasting, it may not have an injurious action upon the cells of the liver and pancreas. If used at all, it should be well diluted. It can l)e given, according to the taste of the patient, in the form of whisky, brandy or the sugar-free wines, a list of which will be found on pages 51G and 517. Dr. John R. Williams, of Rochester, showed me a patient who was apparently brought out of impending coma by the use of whisky and honey. It should be remembered that for every 30 c.c. of whisky given a patient the combustion of 12 grams fat may be prevented, granting alcohol and fat act in an isodynamic manner. It is possible that it will be eventually shown that the effect of alcohol in reducing acidosis is simply that due to its sub- stitutive caloric value for fat, thus reducing materially the source of acid formation. The favorable effect of the administration of carbohydrate in the 1 Bernstein and Falta: Deutsch. Arch. f. klin. Mod., 1916, cxxi, p. 95. 2 Neubauer: Miinch. mcd. Wchnschr., 1SS2, v, p. 346. 394 T RE ATM EXT prevention of coma has been recorded by Epstein and Felsen. 1 1 pon the first day 70 "Tarns carbohydrate in the form of skimmed milk and 5 per cent, vegetables were given and subsequently oO grams carbohydrate. It would be fortunate for the diabetic patient if the researches started by Schwartz 2 might, have been continued. He found in glnconic acid a substance which was capable of combustion by the diabetic, and as a result an existing acidosis could be 'decreased., Unfortunately, this substance is not available, due in great part to the expense of production. 1 know of no other substitute for carbohydrate which has this antiketogenic function. (<7) Dangers of Alkalis. The dangers attendant upon the use of alkalis in the treatment of acid intoxication far outweigh their advantages. This is a strong statement, but experience has forced this confession from me. I believe that often a patient threatened with diabetic coma is sent into actual coma by the careless adminis- tration of alkalis. Between September 22, 191"), and March 12, 1917, but ?> of the cases of diabetes in hospitals under my care have received alkali in any form. During April and May, 191C>, I gave a total of 44 grams to one patient and 20 grams to another. One of the three patients had been receiving alkalis prior to entrance and I hesitated to suddenly omit the soda and thus disturb the water balance of the body. These three patients all died of coma. The results obtained since the routine administration of alkalis has been abandoned have been so satisfactory that T shall not willingly return to their employment. A diabetic patient theoretically needs alkalis. \Vilenko, :! has shown that glycolysis in blood and organs is hampered when the blood becomes even a trifle less alkaline than normal. Hut one must bear in mind that it is possible that the administra- tion of small quantities of an alkali over long periods may set free acid bodies existing combined, quiescent and harmless in the body, and thus do harm. Furthermore, the administration of alkalis over long periods may deplete the body salts, such as chlorides, which are distinctly useful. Another danger from the use of alkalis is the occurrence of nausea and vomiting, and this is real, though greatly lessened when chalk is combined with the sodium bicarbon- ate. When alkalis are given large quantities of urine must be voided to remove the salts of the acid. The quantity of liquid which must be ingested is so large as to overburden the stomach and the excre- tion of so much acid frequently overwhelms the kidneys and they cease to act. Finally, the constant use of an alkali appears to promote the 1 Epstein and Folson: Am. Jour. Mod. S<\, 1917, ciliii, p. OS. j 2 Schwartz: Dcut. Arch. f. klin. Med., I'M).'!, Ixxvi, p. 247. ' Wilunko: Med. Klin., Berlin, l'Jl-1, x, p. 1003. TREATMENT OF ACID INTOXICATION 395 constant excretion of acid bodies. I have known a moderate acid- osis of months' duration to vanish with the omission of soda. It is frequently to be observed that when an alkali is omitted in the convalescent stages of a diabetic cure that acidosis, as measured by the urine, entirely disappears, but will be brought back by resuming the alkali. A very small quantity of alkali may cause the appearance of a positive ferric chloride reaction in the urine. Patients have observed this after a Seidlitz powder, and one patient, Case Xo. 942, thought it followed the use of a brand of saccharine which was combined with sodium bicarbonate. In other words, the administration of an alkali may give a false idea as to the severity of the case if one is guided by the urine alone. This shows how necessary it is to study the blood. Van Slyke in the discussion of a paper by Fitz at the Association of American Physicians in May, 1917, pointed out that not only acids remaining in the body might do harm but also those excreted in the urine by removing bases. It is by no means, therefore, an unmixed blessing to favor the re- moval of acids from the body by the use of alkalis. A few years ago I wrote the following sentence: "Few instances in medicine occur which show more strikingly the benefit of the administration of a drug than the change from the drowsiness and exaggerated respiration of beginning diabetic coma to the reawaken- ing which follows the administration of large doses of sodium bicarbonate." Since I have given up alkalis I have seen fully as drowsy patients recover without their use. In the ordinary treatment of diabetes alkalis are certainly not needed. It is safer, more agreeable to the patient, and easier to bring about the dis- appearance of a slight or moderate acid intoxication by the omission of fat followed by fasting than to attempt to neutralize acid intoxi- cation with an alkali. In earlier days when alkalis were given to diabetic patients as a matter of routine it was found better to administer the same in fairly large doses, 6 to 8 grams six times a day, and after continuing this amount for three or four days to gradually decrease the quantity. The clinical experience of various authors, as well as my own, favored the administration of doses fully as large as this, rather than to begin with small doses and then increase the same. Possibly the reason for this was that by this means alkalis were given over a shorter period of time and the total amount was less, so that the good accomplished was not offset by the harm of prolonged administration. If alkalis are employed they should be administered in dilute form both to avoid irritation of the gastro-intestinal tract, and to favor the ingestion of large quantities of liquid. The greatest pains should be taken not to give an alkali for several hours before or after 306 TREATMENT acid fruits. Generally, the stomach tolerates alkali better when not given within an hour of alcohol, so that when alcohol and alkalis are given freely it is best to alternate the same. It is undesir- able to give alkalis in milk, because nausea and vomiting may result, and nothing is more harmful in severe acid intoxication than the production of vomiting. Rather than bring a patient to this stage, give up all medication by the mouth. A quart bottle of Celestins Vichy contains approximately 4 grams of sodium bicarbonate, and patients are often glad to take their alkali in this form. In some cases an additional 4 grains or more of the bicarbonate may be added to the chilled Vichy. It is generally agreed that the sodium salts are preferable to the potassium salts if the doses given are to be large. Sodium bicarbonate was the alkali most commonly and advantageously used in my practice. Sodium citrate is of the same bulk as the bicarbonate, is said to be less irritating because carbon dioxide is not set free in the stomach, less apt to cause diarrhea and less disturbing to the appetite, but I have not found that it made much difference which preparation the patient used. The citrate is said to be oxidized in the body and in this oxidation to favor the combustion of acid products, but I question whether this property is of enough importance to outweigh the advantages of the use of sodium bicarbonate, which can be so easily obtained. Sodium carbonate is too irritating, and, furthermore, on account of the water of crystallization, the actual amount of sodium which is contained in 100 grams of sodium carbonate is equivalent to only 5S grams of the bicarbonate. This is not generally recognized. Magnesium citrate is an excellent preparation when constipation is associated with the acidosis, but naturally large amounts cannot be given in this form. ( 'halk, calcium carbonate (creta preparata) is valuable because it will not only act as an alkali, but because it tends to render the sodium bicarbonate less irritating to the stomach and thus is less apt to cause nausea or diarrhea. Alkali may be given by the rectum in the form of a 3 per cent, solution of sodium bicarbonate in salt solution. This is occasion- ally well borne, but I attach so much importance to the liberal administration of liquids that I always hesitated to use the rectum for alkaline medication. There is too much danger of its becoming intolerant for any kind of fluid. Sodium bicarbonate may also be given intravenously. I nfor- tunately, at times following the intravenous administration of sodium bicarbonate, convulsions and collapse occur, so that one dislikes to resort to this method if others will suffice. It is my impression that if the intravenous method were adopted freely at TREATMENT OF ACID INTOXICATION 397 the beginning of threatening symptoms, the alkali treatment of coma would be much more successful. The harmful effects attrib- uted to the intravenous injection of alkalis may in part be connected with the late stage at which they are given. Sodium bicarbonate is the best form of alkali to give intravenously, and although 10 per cent, solutions have been given, it is probably safer to use a 3 per cent, solution, and to give a liter of the same and repeat it within twenty-four hours. I gave 1500 c.c. of such a solution several times to Case No. 4, with the result that he came out of coma only to relapse into it. Young individuals present a better prognosis than adults and much better than elderly individuals with damaged kidneys. Blum suggests that a second injection be avoided unless increased diuresis has shown that the body was capable of dealing with the first. The sodium bicarbonate may be sterilized along with the salt solution. It is true that some of the carbon dioxide is liberated through sterilization, but this does no harm when injected intra- venously. Under no conditions should soda solution be given sub- cutaneously. Magnus-Levy 1 pointed out that if a current of carbon dioxide should be passed through the solution of sodium bicarbonate after it has been sterilized and the sodium carbonate has been precipitated until the red tint with phenolphthalein disappears the sodium bicarbonate would be reformed. Sterilization, however, is probably not necessary. My former assistant, Dr. F. A. Stan- wood, demonstrated by a series of experiments that if sodium bicar- bonate was taken with a sterile spoon from a previously unopened Squibb's package, weighed in a sterile watch-glass, and then added to sterile salt solution, the resulting solution was sterile. He also found that if infected with staphylococci and typhoid bacilli the solution sterilized itself in three days. (?) Importance of Liquids. A study of recovery from coma in diabetes invariably shows, so far as I am aware, that large quanti- ties of urine frequently 10,000 c.c. daily for several days were voided in old days when the alkaline treatment was employed. Seldom if ever does diabetic coma occur in a patient showing considerable edema unless severe cardiac or renal trouble is also present. The exclusion of salt from the diet of the severe diabetic patient apparently hastens coma, and it is possible that the adminis- tration of a diet poor in salts for long periods may be harmful for this reason. Conversely, the favorable effect of some mineral waters may possibly be attributed to the large quantity of salts which they contain. It occasionally happens that a case of diabetic- coma is rescued by the free use of salt solution. All these observa- 1 Magnus-Levy: Therap. Monatschr.. 1913, xxvii, p. 838. 398 THE AT ME XT tioiis point to the importance of having the diabetic patient secure liquids in abundance. Hut it is not enough to offer diabetic patients liquids. The liquids must be given in such form that they will be retained. A stomach bombarded alternately with alcohol and soda each hour will not tolerate even plain water long. Recourse should be had to rectal injections of salt solution early rather than late. The salt solution may be given intravenously provided it is given slowly. Only once have I known of salt solution being given subcutaneously to a diabetic without causing an abscess. In giving salt solution one must be careful not to give so much as to embarrass the kidneys in excreting the salt. Taj) water may be given alternately with salt solution. J. THE CONSERVATION OF ENERGY IN THE DIABETIC INDIVIDUAL. The untreated severe diabetic loses much energy through unoxi- di/ed products (sugar and /3-oxybutyric acid) voided in the urine, and by his increased metabolism. Every effort should be made to prevent this waste, and this is accomplished by the treatment above outlined. So soon as sugar and acid disappear from the urine the metabolism falls. Patients upon a low diet, should avoid extremely low temperatures as well as rapid reductions in body heat, which occur in automobiling or in prolonged cold bathing. So soon as sufficient calories can be given, and the body, by treatment, is brought back into good physical condition, the necessity for these precautions disappears. There is another loss of heat not generally appreciated. Patients drink cold water, and this must be raised to the temperature of the body before it is excreted as urine. The amount of energy expended in warming this water is considerable. Thus if three liters (kilos) of water at 10 ('. (;">() F.) are drunk, the temperature must be raised to :>7 ('., (l>X. kilos of water 1*7 (\, there would be' required SI calories. In oilier words, the patient has employed enough energy in raising the temperature of this liquid to that of body heat to convert approximately two pounds of water at (\ into boiling water. Body heat is also lost by the large quantities of liquids drunk and water voided. Thus < !()()() c.c. of urine would represent a loss of approximately "200 to -\00 calories. The loss of energy could also be prevented by limitation of exercise. But it has already been pointed out that there are objections to such a course. \\hile it is true that exercise favors the oxidation of carbohydrates, there is alwavs danger that the exercise mav be overdone unless the THE USK OF DRUVX 399 patient is prepared, as lie should be, for it. Even cases of diabetes of moderate severity should not enter into contests or place themselves in other situations where they cannot instantly rest if the need is felt. Case No. 344, who was a model patient, and withstood his severe diabetes for years, exercised quite freely, both horseback riding and at tennis, but he invariably exercised for short periods and took a complete rest immediately afterward. lie was able to live up to a rule of Dr. F. C. Sbattuck that one hour after the exercise ceased the patient should feel better than before it was begun. Attention has already been called to the need of maintaining the body warmth of patients threatened with diabetic coma. Just as much pains should be taken to avoid exposure of such a patient as is taken with patients suffering from acute Bright's disease. K. THE USE OF DRUGS. Drugs may be very beneficial in the course of treatment of a diabetic case, as in any chronic disease, but this is not because of any specific action upon the diabetes. Nevertheless, scores of drugs have been employed with such a purpose. I use none of them, and I think the same custom is followed by those who have had a greater series of cases. The older writers considered opium with its derivatives and jambul and aspirin the best of the drugs. Even these drugs fail to increase the tolerance of the patient for carbohydrates more than temporarily. A good rule to follow in regard to drugs for diabetics is that of Ralph Waldo Emerson for novels wait a year. This saved me once from giving corrosive sublimate to Case No. 1. The employment of drugs by the profession is far too common. Why this should be so, when men like Naunyn, von Noorden, and Hodgson, with an enormous experience in diabetes, discountenance drugs, deserves analysis. The chief reason for this which I have heard was that the patient could not be controlled, and will not be satisfied unless he is given medicine. 1 have had no trouble upon this score. The reason why it is hard to retain the patient's confidence is not that drugs are not given, but that results are not obtained with the diet prescribed. A properly devised diet will keep the patient in hand better than all the drugs in the pharmacopu'ia. A proper diet implies that the urine has been examined and the carbo- hydrate, protein and fat in the diet for the corresponding period reckoned. It is my impression that physicians who use drugs the most are those who examine the urine quantitatively the least. If one wishes to know the effects of a drug upon the course of diabetes a careful metabolism experiment should be carried out. 400 TREATMENT The patient should be put upon a constant diet and maintain this for a period of at least fifteen days. During the middle five days the drug would be administered, while the preliminary and after- periods should serve as controls. It would really be advisable to have the experiment last twenty days, so that the drug could be given during one of the middle five-day periods, thus concealing its administration, in order to discount any mental effect. Such an experiment would at least give a hint of the value of a drug. From a considerable experience with metabolism experiments in diabetes I must add that data of this sort can only be satisfactorily collected when the patient is under close supervision. Even with the best of intentions to cooperate in the test, the patient may break the prescribed diet without knowing it. Finally, before resorting to such drugs as opium or aspirin one should consider the probable duration of life of the particular patient under dietetic treatment. If such consideration shows that the patient has a good prospect of living ten years, advice to use opium would be very different from such advice if the expectation of life was ten months, ten weeks, or ten days. The older plan of treat- ment with drugs lost sight of this. The perspective was too narrow. Thus Case Xo. 10, male, bank teller, came to me in October, 1899, with the history of onset of diabetes at thirty-nine years of age, March, 1S9S, while on an important financial trip to New York City. The urine contained S.I per cent, sugar, and with moderate restriction of carbohydrates showed, four days later, a moderate ferric chloride reaction and 3.4 per cent, sugar. For weeks previous to his visit to me he had taken large quantities of opium, and had reached a most deplorable condition until he gave it up of his own accord. It is true that he became sugar-free only for a few days, with the treatment of eighteen years ago, but for the succeeding twelve years he not only continued his occupation, but was promoted to the position of cashier, only giving up work in 1909 upon development of mild signs of pulmonary tuberculosis. The utter impracticability of treating such a case for twelve years with opium is obvious. Opium may decrease the sugar in the urine by about 1 per cent., if it has already fallen to the vicinity of 2 per cent. The quantity of sugar is seldom lowered more than 10 to 15 grams, and at the most only 15 to '20 grams. Small doses of opium are not efficacious, and the larger doses, /. c., 5 to 7 grains, exert too injurious an influence to be continued more than for a short time. Von Noorden occasionally uses opium temporarily, but never for more than two weeks. Kleiner and Meltzer, 1 following the use of rather large doses of morphin to dogs, found the elimination of intravenously injected 1 Kleiner and Meltzcr: Proc. Nat. Acad. Sc., 1910, ii, p. 369. THE USE OF DRUGS 401 dextrose considerably increased and that it retarded the return of the sugar content of the blood to its previous level. Aspirin is still less advantageous than opium, for though it may be taken with little discomfort for a short period of time, it eventually is sure to disturb the digestion. No one pretends that aspirin will act after it is omitted, and I expect my patients to live too long to use it permanently. I cannot acquire any enthu- siasm for the other drugs, but in the works of Naunyn and von Noorden data will be found about many preparations which they have investigated. An exhaustive description of the older prep- arations will also be found in the article by Kaufmann. 1 Organic preparations made from the pancreas and the pancreatic gland itself have been frequently employed, but thus far the data do not appear to me to warrant a continuance of their use. So far as the digestion of fat is concerned, I fear my patients have absorbed it too well. The Bacillus bulgaricus has been extensively advertised and tried. Henderson 2 investigated its action and concluded it was of no advantage in the treatment of diabetes. Caramel, originally recommended by Grafe 3 was employed later by Umber. In mild or moderate diabetes he noted no increase in the sugar content of the urine, but in severe cases the glycosuria increased. 4 It should not, therefore, be employed save with the same control as any carbohydrate. Substitutes for Suf/ar. Saccharine may be used as a substitute for sugar. It is far better, however, for the patient to forget the sweet taste. Children especially should grow up without it, learning to depend on the natural flavors of vegetables and fruit. If sac- charine is employed, advise the patient to render foods less sour with saccharine rather than to make them sweet, because the sweet taste persists. Foods should not be heated after the addition of saccharine, as they are apt to acquire a bitter taste. Various other preparations have been employed instead of saccharine. Such are crystallose and dulcin. Attention is called to the possi- bility of saccharine prolonging a positive ferric chloride reaction in the urine if it is prepared with sodium bicarbonate. So far as I am aware, substitute preparations have no advantage over saccharine and cost more. Glycerin is not desirable as a sweetening agent. From it sugar may be formed. 5 1 Kaufmann: Zentralbl. f. klin. Mod.. 1903, xlviii. p. 260. - Henderson: Jour. Am. Mod. Assn.. 1915, Ixiv, p. 4950. 3 Crafo: Munch, mod. Wchnsohr., 1914, Ixi, p. 1433. 4 Umber: Deutsch. mcd. Wvlmschr., 1915, xli, p. 181. 6 Naunyn: Loc. cit., p. 441. 402 TREATMENT Hediosit, Ilediosit, the lacton of glycoheptonic acid, is a sugar with 7 carbon atoms. It. is oxidi/ed by the patient, and so might be helpful, not only in combating acidosis, but to sweeten the food. Unfortunately it is expensive and often disliked by the patient, and Linels 1 has found that upward of a quarter of the quantity adminis- tered may be lost in the stools under normal conditions. If a laxa- tive effect is present, SO per cent, is thus lost in addition to a portion excreted in the urine. L. THE TREATMENT OF COMPLICATIONS. The complications or sequelae of diabetes are quite as important as the disease, but we should bear in mind that it is the diabetes which is originally at fault and not the complication. A duodenal ulcer may perforate and cause peritonitis, but the real cause of the death is the neglected ulcer and not the complication peritonitis. It is the inadequately treated diabetic who develops complications. Diabetic patients should remain free from complications because diabetics are or should be constantly under the doctor's eye and his prophylactic care. In the past the duration of life of diabetic patients has been so short that the dangers which might result from peculiarities of treatment have been neglected, for in the space, of five or six years these would not manifest themselves, but now with the duration of the disease being so much prolonged, all these factors must be taken into account. It is a well-known fact that disease of the kidneys may occur in diabetics who have had diabetes for many years. lias this been brought about by the protein-fat diet? Those interested in the study of diseases of the kidney could ask no better experimental basis for the investigation of etiological dietetic problems than these patients. Here are experiments of a most accurate type going on before their very eyes, not for days or months, but for years. 1. Infectious Diseases. Never try to cure the diabetes in the presence of an acute infectious disease. This is where the family doctor has often scored on the supposedly wise consultant. The latter would order a fat-protein diet and death would result. If the case is doing well, either leave it alone or make gradual changes and there will be more chances of having the individual for a patient during years to come than as if brilliant dietetic orders were given. \\ith my diabetic children, realizing their tendency to acidosis, I always try to avoid acidosis during an infection by keeping the fat low, otherwise the treatment is plain sailing. 1 Linels: Arch. f. cxp. P;n1ml. u . Pharin., HM-l, Ixxvii, p. :M~>. TREATMENT OF COMPLICATIONS 403 Carbohydrate in the form of oatmeal gruel, orange juice and even as puree vegetables and protein as whites of eggs, oysters and fish, will tide over many emergencies and it very likely will be found that the patient can bear a moderate quantity of fat as well. There is no reason for not giving what the full tolerance will allow. Could it be that the tolerance during fever would not be lowered if fat was restricted in other words, acidosis avoided? TABLE 173. CASE No. 12G4. ILLUSTRATING How EASILY A DIABETIC DOCTOR BECAME SUGAR-FREE AND WHAT HAPPENED WHEN TONSILLITIS DEVELOPED. Sugar Diet in grams. i 'c.* in urine .. ^ 2 Date, ;_' -r fs .*! Diet ordered, 1917. 5 - 1 -2 >: " r? f *a gms. Cj '*3 '3 5 5 JL - ~~ t -s' -~ R ~ | ^ 3 ~ ti1< ~ ^ -5, o Is ' Jsl > S - ^ 1 ^i < c C X !5 Spec. v. si. + tr. Mar. S- 9 1300 + 0.1 1 153 + 80 + + . 950 152+ 177 5 per cent, veg., 450; oranges, 3; fish, 300; potato, 270; bread, t 90. 9-10 1400 1 1 153 98 , 1004 153 + 178 Same as 89. 10-11 550 II 153 75 47 ; , 1335 153 + 179 Make potato 180 and bread 120; add meat, 120; bacon, 30; eggs, 2 11-12 1200 105 ! S3 47 ) 1415 105-r Add milk, 240. Tonsi llitis . 12-13' 1100 tr. 40 20 IS 508 40+0 Add oatmeal, 30. 13-14 800 0.1 1 104 41 10 090 103 + 1 Same as 12-13. 14-15 700 si. + 0.1 1) 108 01 2 534 108+ 179 Make 5 per cent, veg., 300; bread, 90; fish, 180; oatmeal, 15; milk, 30. 15-10 SOO v. si. + 0.1 2 100 68 28 948 104 + 174 Omit fish; make meat 120. 10-17 1100 II 52 ; 50 3 435 52 + 173 Make potato 90; oat- meal, 30; milk, 60; add fish, ISO. 17-18 1300 1) 77 , OS 41 ) 919 77+0 172 Omit fish; make meat 150; potato, 120. 18-19 1300 II 77 71 74 3 1258 77 + 174 Add butter, 30; make bacon 15. 19-20 1400 77 79 87 'o 1407 77- 173 Make bacon 33; eggs, 3. 20-21 1200 70 , 79 109 , 1001 70+ 174 Omit milk; add cream, 00. 21-22 1000+ II 70 ' 79 99 1511: 70+ 173 Same as 20-21. 22-23 900 II 70 79 109 ) 1001 70 + . . Same as 20-21. (a) Pneumonia. By no means do all diabetic patients die when they have pneumonia as Table 174 will show. 1 am quite sure there have been other patients in my group who have recovered, for it is only recently that my attention has been directed to this point. Case Xo. 105 recovered from mumps, Case No. 93S from chicken- pox, and Case Xo. 12(>4 from tonsillitis. (Table 173.) I have already referred to an Italian diabetic at the Boston City Hospital who recovered from typhoid fever upon a diet of oatmeal and olive oil. A patient in diabetic coma whom I was asked to see 404 TREATMENT showed a diphtheritic membrane (culture positive) and died twelve hours later. During an attack of herpes zoster the carbohydrate tolerance of Case No. o21 fell markedly. This is interesting because it suggests an infectious etiology of this disease. TABLE 174. RECOVERY FUOM PXEUMOXIA ix DIABETES. Ago at Duration. Case onset of Before After No. diabetes. pneumonia. pneumonia. 8 . . . . . 60 12 years 6 months 1 year 1 month 10 . . . . . 47 7 months 9 years 30 . . . . 42 7 years 3 years 3 months 40 . . . . . 62 20 years 4 months 358 53 1 7 years 1 year 3 months 435 3-3 23 years 2 months SOo . . 55 12 years 12 years 2 months 1271 . . . . . 56 4 years 7 years 9 months 2. Tuberculosis. Tuberculosis formerly was responsible for about one-half of the deaths of diabetic patients, but today, in private practice, it seldom occurs. Montgomery, 1 who has published an excellent critical summary of the literature upon diabetes mellitus and pulmonary tuberculosis, concludes that tuberculosis is not more frequent among diabetics than among the rest of the popula- tion of similar age. Undoubtedly this is due to the well-known decrease in the mortality from tuberculosis everywhere. Yon Xoorden, writing later than did Xaunyn, records a still lower percent- age among his cases, though in Yienna it was nearly twice as common among hospital diabetics as in Frankfort. This is a striking illus- tration and a warning against exposure of diabetic patients to tuberculosis. Prophylaxis is important. Unfortunately a tuber- culosis latent for nearly a lifetime may break out in the presence of diabetes, as it did in Case Xo. (521), at the age of sixty-nine years, and in Case Xo. 344, in whose family there had been much tuberculosis, though he himself had never shown recognizable signs of it. Xo matter how well the diabetic patient may appear, he remains a vulnerable individual. A glance at Table 17") shows that the number of cases of tuber- culosis was far more numerous among the early than among the late case numbers, another illustration of the disappearance of tuberculosis in the community and probably an illustration of the effects of treatment of diabetes, for I take it that in the last t'e\v years diabetic patients have not been so debilitated with the progress of the diabetes. On the other hand, an argument could be raised that recently diabetic patients have died prematurely of coma, whereas, according to the old regime, they would have 1 Montgomery: Am. Jour. Me Tuberculosis. 1.3 2.S 0.3 1.6 1 Tubercular heredity. - The fixation of the onset of tuberculosis is of course to a considerable degree uncertain. The statistics are so compiled that the onset is earlier rather than later than reported in this table. 400 TREATMENT Six of the 2S fatal eases, or 21 per cent., died in coma. This is far below the percentage of deaths from coma in the whole group of diabetics. It suggests that the diabetes decreased in severity, due to the emaciation and lack of food consumed. Seventeen patients apparently died in the ordinary manner of an uncompli- cated tuberculous case. It is interesting that of the iU cases of tuberculosis, my records show that tuberculosis was present, in only 7 instances in other members of the family. Contrary to my expectation, my records show that frequently tuberculosis occurred before the diabetes. Proof is not convincing, but my records indicate that I felt this to be true at the time I took the history of the individual. The explanation lies in the fact that the tuberculosis, though it existed for a longer period, was of a latent or mild type. Fie. 14. Tlll>pri-: ;uid diabetes. The tcmprrat lire, chart of Case Xo. '.1 10. and rr>>pirn1ion Tuberculosis comes on in diabetes even more insidiously than u>ual; more than once I have been surprised at discovering not only its pre.MMice, but the advance which it has made. One should always be on the watch for it, taking For granted that it will not appear in an open manner, and therefore making complete physical examinations of all diabetic case- several times a year. Indeed, this is necessary for other reasons, for it is a very easy mistake for the physician to be so wrapped up in the treatment of the diabetes that he neglects the general condition of the patient. Loss of weight TREATMENT OF COMPLICATIONS 407 is attributed to the diabetes, and the absence of temperature throws the physician off his guard. (Fig. 14.) I can confirm older writers in the rarity of hemoptysis. A suggestive family history and an .r-ray examination help to solve the diagnosis of tuberculosis in diabetes in the exceptional case who has no sputum. In Case No. G33, long under close observation in the hospital, I must confess that the presence of tuberculosis was first disclosed at autopsy. It was supposed by those of us in attendance that the patient's decline was wholly due to severe diabetes, ultimately culminating in pneumonia. The postmortem examination revealed extensive tuberculosis. Such a mistake is not extraordinary, for shortly after, bearing it in mind, I saw a similar case in the wards of one of the best hospitals in the country where this diagnosis had not been entertained. The tolerance of the patient in question was gradually improving, but he was losing weight. On the other hand, an error was made in Case No. 1083 by neglect to observe the rise in temperature upon two days during his hospital stay. Tuberculosis, angina pectoris, old age, and diabetes are com- patible with a useful life and a fair degree of health. In Case Xo. 029 diabetes developed at the age of seventy-one, tuberculosis became active at seventy-five, and angina pectoris at seventy-seven. The patient came under observation in July, 1913. Six months later tubercle bacilli were found, though previous physical examina- tions had disclosed no lesions. Undoubtedly tuberculosis had existed in latent form since youth. Under the close supervision of a devoted nurse, thoroughly trained in diabetes, the patient attended to active duties, and is now living in comfort at the age of seventy-nine, an illustration of the hopefulness which one is justified in assuming even in the presence of diabetes, tuberculosis, and angina pectoris. My friend does not enjoy fasting. A summary of his case is given below: TABLE 176. TUBERCULOSIS, ANGINA, OLD AC;E AND DIABETES COMPATIBLE WITH A USEFUL LIFE AND FAIK DEGREE OF HEALTH. Case Xo. 629. Onset of diabetes at seventy-one years. Tuberculosis became active at seventy-five and angina pectoris at seventy-seven years Present age seventy-nine years. Urine. Volume, Nitrogen, Total sugar, Weight, Date. c.c. Albumin. grams. grams. pounds. Sept, 24, 1913 . . 1450 11.0 58 155 Feb. 7, 1914 . . 1175 s. p. t, 10.2 9 15S Sept. 3, 1914 . . 1050 s. p. t. 15 147 Jan. 11, 1915 . . lf>50 s. p. t. 13.0 10 145 Nov. 30, 1915 . . 1200 7 140 Feb. 26, 1916 . . 1350 13S Mar. 10, 1917 . . 1320 si. tr. 7.3 1 145 2 1 Feb. 1, 1917. 2 Note the pain in weight. 408 TREATMENT Case No. 8(51, male aged fifty-two years, developed diabetes while traveling in Europe in 1914. Tuberculosis superimposed itself upon the diabetes, but apparently neither was discovered until considerable progress had been made in both. Through the courtesy of Dr. Lawrason Brown I saw this patient, and it is gratify- ing to record that after less than a day of fasting he became sugar- free, and gradually acquired a tolerance for (57 grams carbohydrate, and hemorrhages ceased. Tubercle bacilli did not disappear from the sputum, and he subsequently succumbed to tubercular menin- gitis. In the intervening eight months the weight increased twenty pounds. How extensive an improvement can take place in pul- monary tuberculosis complicated by diabetes is shown by Dr. Brown's letter: "The patient presented dulness with broncho- vesicular breathing and increased vocal resonance below the fifth vertebral spine on the left side. There were also numerous moder- ately coarse rales. Over the right side there were moderately coarse and fine rales. Five months later the lesion at the left base cleared up almost completely. There were a few fine rales to the third vertebral spine, when he left, on the left side and possibly below the fifth rib. The right side still showed signs over most of the side." The improvement in the last two cases is not exceptional. Xaunyn records a case of a physician, aged fifty-three years, with diffuse pulmonary tuberculosis, and also a small ulcer on the left vocal cord, in whose urine was found 4 per cent, of sugar. Careful treat- ment of the diabetes and the tuberculosis gradually brought about a gain, although some months after the' beginning of the treatment there was slight hemoptysis, with tubercle bacilli present. The condition, however, kept improving, and sixteen years later both diseases were apparently arrested. The advent of diabetes upon tuberculosis is seldom reported, as observed in cases of tuberculosis treated in sanatoria. Mont- gomery 1 found that of ol,8o4 cases of tuberculosis treated in sana- toria there were 101 (about \ of 1 per cent.) cases of glycosuria and ol (about I, of 1 per cent.) cases of diabetes, and he states that practically not a case of diabetes following tuberculosis has occurred at the Phipps Institute, in the sanatoria or under the care of physicians with whom he is acquainted. Case No. 44o, a woman, aged fifty-five years, first seen on Sep- tember 120, 1011, gave a history of congestion of the lungs at the age of twelve, hemoptysis at sixteen, and again to a marked degree at the age of twenty-four. Tuberculosis was confirmed by Drs. Flint and Loomis, of Xew York. Tubercle bacilli were never found, but the evidence of tuberculosis was undoubted. An .r-ray examina- 1 Montgomery: Amer. Jour. Mod. Sc., 11)1-', clxiv, p. 643. TREATMENT OF COMPLICATIONS 409 tion on September 21, 1911, showed general peribronchitis with numerous broken-down glands with healed borders, increased density and mottling with an appearance of healing at the right apex, small areas of calcification, slight emphysema. The patient was sugar-free in September, 1911, and is alive today. The reason for the advent of the diabetes in these two instances may be explained by the latency of the tuberculosis, and the explanation for the rarity of diabetes in tubercular sanatoria is shown by the course of the case described in the next paragraph. Diabetes, even of the severest grade, may practically disappear when tuberculosis intervenes and reaches an advanced stage. Many writers have observed such instances. Case No. 344, cited hitherto as having an extremely severe type of diabetes, developed tuberculosis after eight years. Eventually the acidosis, which had amounted to the most extreme grade which I have met outside of coma, completely disappeared, and a minus carbohydrate bal- ance of 40 grams, which had existed for months, changed to a positive carbohydrate balance of 60 grams. The complete record of the case is given below. (Table 177.) The disappearance of the acidosis is explained by the development of a positive carbohydrate balance. The explanation of the body's regaining power to utilize carbohydrates was difficult until I showed the records of the case to Dr. Allen and told him that if he could explain the improvement of the diabetes coincident with the development of tuberculosis, a key to a valuable method of treatment of diabetes would be found. A few days later he said to me that he believed he could duplicate the case with his depan- creatized dogs. How well he succeeded is known today by his experimental work upon animals, which showed that fasting and loss of weight will improve the tolerance of even the severest diabetic patients. Older writers laid emphasis upon the rapid extension of the tuber- culosis in diabetes after it first appeared, and for this reason gave an especially unfavorable prognosis. Examples of this character are only too many, but while they are not encouraging from the therapeutic point of view, they should not be considered too dis- couraging. It is possible, indeed probable, that many of these cases had poor treatment not only for the diabetes but for the tuberculosis as well. The duration of the tuberculosis as well as of the diabetes in my cases absolutely controverts this former pessimistic view. Of the 24 cases in which the duration of the tuberculosis is recorded, in only 3 instances lias the patient died during the first year of the tuberculosis. In one instance only did the diabetes last for less than one year. These facts are shown in Table 175. 410 TREATMENT S~ n c 5 5 . ? 2, -i ^S 2 *^ 5'6 ?. -f -r -r-ff -H - OO .OO TREATMENT OF COMPLICATIONS 411 O COO CO CO CO CO CO CO M I 0) 01 01 01 + + + + + + oi a ot 412 TREATMENT I II + I I I I I II III I I + I ' I I I TREATMENT OF COMPLICATIONS 413 3. Arteriosclerosis. Bright's Disease. Arteriosclerosis is of com- mon occurrence in protracted cases of diabetes. It is of the senile type. It seldom occurs even in the severest cases of diabetes in youth or middle-age a strong argument against diabetes being a direct causative factor of arteriosclerosis. A study of the blood-pressure of a series of my diabetic cases favors the view that the presence of sugar and acid in the urine does not injure the arteries or kidneys. Thus, only 19 per cent, of the cases occurring between the ages of twenty-one and fifty exhibited a blood-pressure of over 150. What is still more confirmatory of this point is the fact that 07 per cent, of the patients of fifty-one years of age and over showed a blood-pressure between 100 and 150. The blood-pressures of diabetic patients between the ages of twenty-one and fifty, and fifty-one and over are shown in Table 178: TABLE 178. THE BLOOD-PRESSURE IN DIABETES. 100-125 120-150 151-17; No. of cases 21-50 years of age 174 191 50 No. of cases over 50 years of a^e 172 283 129 176-200 Above 200 19 8 62 29 The blood-pressure was determined with the Riva Rocci apparatus and a Janeway cuff 14 cm. wide. Occasionly the Tycos apparatus was used, but its reading was frequently checked up with the mercury manometer. The blood-pressure was taken with the patient lying down except in the rarest instances. TABLE 179. THE AVERAGE BLOOD-PRESSURE OF NORMAL AXD DIABETIC INDIVIDUALS. ARCS. 15-20 . . . 21-25 . . . 2(5-30 . . . 31-35 Normals, Fisher. 281 . . 785 . . 791 689 Diabetics Joslin. 38 33 56 39 36-40 . . . 41-45 . . . 40-50 . . . 51-55 . . . 56-60 . . . Over 60 . . . . 2,111 . . 6,740 . . 4,471 . . 2,371 . . 1,100 64 75 116 127 103 163 Average blood-pressure, mm. Hg, Normals, Diabetics, Fisher. Joslin. 120 121 123 122 124 121 124 120 127 125 129 139 131 143 1)52 154 135 154 156 1 19,339 814 127 139 In Table 179 the blood-pressures of normals and diabetics are compared. Fisher- gives the blood-pressure for 19,339 accepted candidates for life insurance, and additional information is added ' Not included in average. - Fisher: Jour. Am. Mod. Assn., 1914, xliii, p. 1752. 414 THE AT ME XT o . .10 TREATMENT OF COMPLICATIONS 415 416 TREATMENT in a subsequent letter quoted by T. C. Janeway: "Dr. Fisher has written me personally that since this publication he has been in- formed of a record of 9000 readings by another company which did not vary 1 mm. in the average of any age from those he has tabulated making an average standard -based upon the observation of over 2S,()()() healthy adults." The blood-pressures of S14 of my diabetic patients are placed alongside Fisher's table. The striking similarity of the blood-pressures in Fisher's and my tables up to the age of forty years makes the immediate divergence of the two groups in subsequent decades noteworthy and demands an explanation. Janeway's conclusions 1 coincide with my data. He writes: . . . it is clear from my own observations and those of Elliott- that diabetes itself is without influence on the arterial pressure. Hypertension is therefore presumably an expression of the well-known tendency of diabetics of suitable age to develop arterial disease; or possibly in some cases arterial disease is the cause of the lesions responsible for the diabetes." Cadbury, 3 at the Peter Bent Hrigham Hospital, found in a study of ))()") patients that a systolic blood-pressure of 100 mm. mercury or more was accompanied by definite signs of chronic nephritis in 73 per cent, of the individuals examined. Low blood-pressures are frequently encountered with diabetic patients. My records are not full upon this point, but I shall pay more attention to this question in the future. In Table LSI arc grouped 10 cases showing such low blood-pressures. Cases with inanition stand out prominently among the group of those in which the blood tension was low. It was partly for this reason that in these cases massage and exercise have been considered important. TABLE 181. -CASES OF DIABETICS WITH BLOOD-PRESSURE OF 100 MM. HG OR BELOW. Case Xos. Number of cases 220231 Another source of information upon the influence of diabetes on blood-pressure can be found by comparing the blood-pressure of a series of diabetics during subsequent stages of the disease. For this purpose Table ISO has been arranged: ' June-way: Johns II. i,, kins Hosp. Hull.. I'M'), xxvi. p. :ill. * Elliott :' .lour. Am. Mr0 at the age of sixty-five without quantitatable sugar in the urine, but in whose urine three years later .1.12 per cent, of sugar was found. After treatment of the diabetes sugar dis- appeared, and the blood-pressure fell in succeeding years to 230, 220 and 210. Coineidentally there has been marked amelioration of the anginal pains, and the patient says he feels better than for years. The improvement may be due to the gradual loss of weight which occurred following the onset of the diabetes. Few cases of arteriosclerosis in diabetic patients appear to result from syphilis. Syphilis may be far more common among diabetic patients than has hitherto been proven, and I suspect exists far more frequently among my own cases of diabetes than my records show: there were 1(1 case's with a history of syphilis among a total of 1187 patients. The routine introduction of the Wassermann test will furnish much valuable information upon this point. Wassermann tests have been performed upon 107 cases of my series. These were selected partly by routine, but special pains have been taken to have the test made in suspicious cases. Yet of the total number of 107 patients whose blood was tested, only gave a positive result. This is confirmatory of clinical and pathological observa- tion that the' type of arteriosclerosis which one sees in diabetes is not that which is usually attributed to syphilis. The symptoms of which diabetic patients suffering from arterio- sclerosis complain are chiefly related to the lower extremities and the heart but less commonly to the brain, characteristic symptoms of senile rather than syphilitic arteriosclerosis. Pains and numb- ness of the legs and cold feet are very common. Intermittent claudication occurred in Case Xo. 8 one year before a cerebral hemorrhage. The whole train of symptoms leading up to gangrene often extends through a period of months, and these patients sutler more than those with actual gangrene. The cardiac symp- toms are more of the order of angina pectoris than of cardiac incompetency. Indeed, angina pectoris is fairly common, though my records show only 7 cases in which there was sudden death from this cause. Cases Xos. 11, lid, 2Sf), :5.V>, -Jlo, -11M and 7.")!). Case Xo. 11, whose brother died of diabetes at fifty-two years, developed diabetes at fifty-eight and came under my observation six years later when I was called to see her during a midnight attack of angina pectoris. Some weeks later she went through a course of restricted diet, became temporarily sugar-free, but ultimately sugar returned. After three months death occurred in a second attack. At the time 1 was self condemnatory and inclined to attrib- T 'RE ATM EXT OF COMPLICATIONS 419 ute the final attack of angina pectoris to the change in diet, but in the subsequent fourteen years this experience has not been duplicated. Cardiac incompetency is rare. Case Xo. 354, with onset of diabetes at fifty, died with such symptoms at the age of seventy, but for the preceding one or two years the sugar in the urine had been replaced by albumin. Case No. 234, who developed mild diabetes in addition to his obesity at the age of seventy-seven, died with symptoms of heart failure. The cardiac asthma often referred to in the older diabetic literature has not been recognized by me. 1 can recall no case of diabetes under fifty years of age in which the patient complained of shortness of breath or presented the signs of an incompetent heart due to uncomplicated valvular disease. I know of but one case of rheumatic fever in a diabetic patient. Case Xo. 694, a mild diabetic, had a slight attack in the spring of 1917. Bri ght's Disease. Albuminuria is frequently observed in the urines of diabetic patients, but actual Bright's disease, like arterio- sclerosis, is practically unknown except in cases past fifty years of age. The relation between Bright's disease and diabetes is shown by a study of the blood-pressures of diabetic cases during successive years. (Table ISO.) The course of the disease in a series of my patients who presented undoubted evidence of Bright's disease is indicated in Table 1S3. It is seldom that one sees death from uremia. Confusion rarely exists between uremic and diabetic coma except in cases of pregnancy. In other words, the renal compli- cations of diabetes have been unimportant in the past, but with the prolongation of life which modern treatment is bringing about they will deserve attention. Cases Xos. 20, 57, 211, 233, 287, 354, 410, 421, 425, 470, 725, S59, S72, 1030 and 10SG were cases in which the condition of the kidneys was quite as important as the diabetes. Cerebral arteriosclerosis (including under this term such diag- noses as cerebral hemorrhage, apoplexy, hemiplegia, or paralysis) was responsible for the death of 20 cases. Treatment, The treatment of diabetes in the presence of arterio- sclerosis, whether general or primarily localized in the kidney, must not be undertaken lightly. Years ago this was impressed upon me by Case Xo. 347, who became sugar-free under the old regimen of restricted carbohydrate and increased fat-protein diet. Soon after discharge from the hospital, however, acidosis increased, and coin- cidently with this a latent nephritis broke out, and death in coma followed. The greater sensitiveness of patients with diseased kidneys to acidosis was pointed out by Goodall and Joslin. 1 It should also be remembered that an acidosis develops in the 1 Goodall and Jo.slin: Boston Mod. and Rurg. Jour., 190S, clviii, p. 799. 420 TREATMENT course of Bright's disease quite independently of the acidosis which occurs in diabetes. This was noted by Frothingham. 1 TABLE 183. THE COURSE OF A SERIES OF CASES OF DIABETES ASSOCIATED WITH BRIGHT'S DISEASE. Duratio i in Blood years fr >m Case No. Age at onset of Albumin in urine. sugar, Klood- per pressure. | onset >f diabetes mtil Cause of death. diabetes cent. death >r Dee. 1, 1910. 16 66 Slightest possible trace Uremia. 26 51 Very slight trace 3 Bright 's. 57 57 Very slight trace to s Hemiplegia. slight trace 85 58 0.3 to 0.7 per cent. 12 Bright'a. 211 45 0.05 to 0.1 per cent. 3 Bright's. 233 54 0.1 to 0.9 per cent. 150 -220 15 Coma. 287 57 Slightest possible trace 190 5 Hemiplegia. to very slight trace 334 GO Very slight trace to 20 Uremia. slight trace 337 40 Slight, trace to trace 100 15 Uremia. 35 \ 50 0.15 to 0.2 per cent. ISO 11 Bright. 's. 3SS 69 Slight trace 170 3 Bright's. 416 54 Slightest possible trace 0.23 , ISO -200 9 Hright's. to slight trace 421 53 0.15 to 0.2 per cent. 150 ISO 9 ( 'ardiac. 425 47 Slightest possible trace 180 8 ( 'ardiac. to very slight trace 725 52 Very slight trace to 190 1 Hemiplegia. slight trace 75S 54 Slightest possible trace 215 2 Coma. to trace 859 52 Slight trace (0.1 per 140 1S5 15 Cerebral cent.) hemorrhage. 1015 47 Slight trace 1..37 SO Annria. 703 54 Trace 0.1S 1S5 S Alive. S72 63 Very slight trace 200-250 3 " 910 47 0.3 per cent, 0.27 175 20 " 070 33 0.05 to 0.1 per cent. 0.24 120 17 " 1030 55 0.05 to 0.1 per cent. 0.22 150 12 " 10SG 07 0.05 to 0.1 per cent. 0.15 215 3 " 1157 59 Very slight trace 0.23 190 (i That marked changes in the diet seriously affect the circulation I am also led to believe by Case Xo. 859, who, in the course of fasting treatment, developed symptoms which indicated marked disturbance of the cerebral circulation, though no actual paralysis developed. These examples show how careful one should be in attempting to better the condition of arteriosclerotic patients who have diabetes. They furthermore show why the older generation rrothiaghain: Arch. lut. Mod., 1916, xviii, p. 717. TREATMENT OF COMPLICATIONS 421 of doctors hesitated, and at that period with good reason, to alter the diet of such of their elderly patients who were leading tolerably comfortable lives. Today the dangers attending the beginning of treatment of such individuals are gradually lessening, and my ideas of the treatment of diabetes in conjunction with arterio- sclerosis have changed much in the last year and a half since Dr. Ilornor and I have studied the cases with a duration of fifteen or more years. At present, after much reflection and conversations with members of the families of these patients, I am convinced that treatment of diabetes can take place without detriment, and that, on the contrary, it may be of distinct help (see p. 468). This change in view-point is due to the ability to prevent the appearance of acid poisoning in these individuals while they are becoming sugar-free and the avoidance of changes in the water content of the body, which takes place so commonly when alkalis are given. Acid poisoning may be avoided in the treatment of arteriosclerotic diabetics by the elimination of fat from the diet. This reduction in the total caloric value of the food may be sufficient to render the urine free from sugar, but should this not be the case at once, no change in the diet need be made for the following two or three days unless the total quantity of sugar fails to decrease. During this interval the quantity of carbohydrate in the diet should be kept the same as upon the day before treatment was commenced, or if this was as much as 100 grams or over, dropped to 100 grams and then reduced 20 grams daily. The quantity of protein should be limited to 1 gram per kilo. If the sugar is not steadily decreasing, after three days the protein should be omitted and the carbohydrate halved, and if necessary, halved again the second day. It will not then exceed 10 grams, and there would be no danger in fasting the patient for twenty-four hours if the urine was not sugar-free. So soon as the urine is sugar-free the quantity of carbohydrate may gradually be increased until sugar returns or the carbohydrate amounts to 30 grams. Protein should then be added, and the plan of diet followed as in other cases. Change in the water content in the body is avoided during the alteration of the diet by watching the patient's weight and increasing or decreasing the fluids and salt ingested. It is particularly neces- sary during a fast to give the patient liquids containing salt. Case Xo. 941 informed me that he lost 13 pounds during four days of fasting in one of the best clinics, but inquiry showed that during these days he drank only pure water. Had he taken broths freely, or a mineral water, most probably weight would not have been lost, because the fluid content of the body would have been increased. A loss of 13 pounds in four days could obviously only be accounted for by fluid loss, This is plain, because 1 kilogram 422 TREATMENT ('2:1 pounds) of fat would furnish 0000 calories, and undoubtedly the patient did not use up six times this amount of nourishment in the four days. It should he remembered that there is no need for hurry in the treatment of diabetes in the presence of arteriosclerosis, because the diabetes has probably existed for a long time, and under treat- ment it is hoped that the patient will continue to live for many years to come. Therefore the few days spent in preliminary treat- ment are no deterrent to its employment. The improvement in the condition of patients showing Bright's disease and diabetes combined is often striking. So frequently is this circumstance noted that it cannot be lightly passed over. Presumably the restricted diet and the low protein in the modern diabetic treatment conduce to this result, quite apart from the fact that the Bright's disease has previously been made worse by the diabetes. In the presence of Bright's disease the quantity of protein in the diet should be kept as low as 1 gram per kilogram body weight, and frequently quite under this amount. Repeatedly patients with Bright's disease acquire an increased tolerance for carbohydrate as the Bright's disease increases in intensity. This often is quite a comfort to the patient, and the physician should be always on the watch for it, so as to give these individuals the benefit of the freedom of diet of which they have so long been deprived. The possibility of renal "block" should be borne in mind. The appearance of acid poisoning with change to a protein-fat diet is illustrated by Case Xo. 75*), Table 217, p. 479, with onset of diabetes at the age of forty, who later developed angina pectoris and came under my observation August 2S, 1914, at the age of fifty-four years. After moderate restriction of diet at that time the patient was temporarily made sugar-free. The diet was not. followed, sugar returned, and the patient was weak, uncomfortable, able to walk but short distances. Prior to returning for treatment in 1915, through the advice of a niece, who also had had diabetes, she placed herself upon a fat-protein diet in anticipa- tion of her visit, to me, with the result that marked acidosis appeared, as shown by the urine of September 27, 1915. During the following days fat was excluded from the diet, but carbohydrate and protein allowed. The sugar decreased, but far more important, the acidosis promptly disappeared. On October 1 she entered the New England Deaconess Hospital and quickly became sugar-free when carbo- hydrate was gradually reduced. At the end of fifteen days the patient left the hospital with a tolerance for carbohydrate of 4S grains, protein 05 grains, fat 52 grams, and an ability to walk three and a half miles without bringing on any symptoms of angina pectoris. Nine months later she died in an attack. Had this patient TREATMENT OF COMPLICATIONS 423 been treated in the same way as Case Xo. 347 (see p. 170), serious or even fatal complications might have ensued. The surprising improvements which occur in the circulatory symptoms of these patients quite likely may be due to the decrease in weight which ensues upon the restricted diet. Thus along with the treatment of diabetes very desirable treatment of the arterio- sclerosis is accomplished. The extent of the influence of the kidney has been determined in several cases by the phenolphthalein test. The excretion of the drug, however, has varied to a considerable extent, not only in different cases but in the same patient. This is brought out especi- ally in Case Xo. 1233, described on page 385. Undoubtedly the retention of fluid, as shown by moderate edema, accounts for part of this variation, but very likely not for all of it. At present my assistant, Richard Ohler, is making a special study of the renal conditions of various groups of my cases, and it seems best to defer further statements until more data have been collected and these have been carefully analyzed. 4. Gangrene. Gangrene has occurred in 37 cases, or 3 per cent, of my series in 28 males and 9 females. Two per cent, of my living cases of diabetes have had gangrene. It was present in (> per cent, of the fatal cases. This gives a wrong impression, however, of the frequency of gangrene, and a much better idea of its importance in diabetes is seen by a study of Table 184, in which the frequency of gangrene according to the age of the patient is recorded. The youngest patient save one, who developed gangrene at twenty-eight years, was fifty and the age of the oldest seventy-four years. In Boston during the years 1895 to 1913 Morrison' found that gangrene appeared in 178, or 23 per cent., of 775 fatal cases of diabetes. There was no occurrence among my cases of gangrene in an upper extremity. In two instances gangrene occurred in both lower extremities, and amputation of each leg was required, and the remaining leg of Case Xo. 895 is now infected, subsequent to recovery from pneumonia. Case X'o. 343, with onset at fifty-eight years of age, lost one leg from gangrene when sixty-two years of age, and the other leg two years later, death resulting at the age of sixty- five. Case X"o. 355 died of angina pectoris four years and eight months after the second operation, sugar remaining absent from the urine during nearly the whole of this period. Case Xo. 58 is also remarkable for the long duration of the disease following amputation of the leg. He was seen by me while undergoing treatment for the morphin habit five years subsequent to the onset of diabetes. Gan- 1 Morrison: Loe. cit., p. 34. 424 TREATMENT grene occurred, and a leg was removed, but the patient lived eight years after the operation. The age at onset of the gangrene is good proof that the disease is not due to the diabetes alone, for only S of .'57 cases occurred under the age of fifty. Gangrene may come in the diabetic with advancing years, as is proven by the statistics recorded in Table 184: TABLE 184. (!ANGKE.\E ix DIABETES. A ire at onset Total rases of Cases of cantrrene. of diabetes. diabetes. No. of eases. I'er eent. Under 30 years . , . . . 201 1 0.4 3 !-")() " 434 7 2.0 51-00 . . . . . . 2of> 9 3.0 01-70 . . . . . . 12!) 13 10.0 71-SO . . . ... 28 7 2f, . When gangrene occurs in diabetics, it is usually after the disease has existed for several years. In Table 1N5 it will be seen that 11 cases appeared before the disease had lasted four years, but two- thirds of the cases developed after that period. The duration of the existence of the diabetes preceding the gangrene is here shown: TABLE 185. DURATION OF DIABETES PRECEDING GANGRENE. Years. No. of eases. Under 1 year. 7 209, 252, 202, 2SS, 470, 517, 725 1 3 " 1 403 3 4 " 3 3M>, 3S9, 955 4 5 " 3 S4, 250, 343 5 10 " 9 150, 340, 420. 4SO, 525, 035, 1010, 1053, 1072 10 20 " 12 47, 5S, 210, 237, 355, 404, 505, 093, 749, SOS, S30, 1111 Uncertain 2 25s, t s93 The concr .sv but a mortifying process dependent upon extensive arterial disease. . . . In short, characteristic for so-called diabetic gangrene is the presence of the typical lesions of athero- or arteriosclerosis. These differ in no way from the lesions of the arteries of the arterioselerotic or senile gangrene, and justify the conclusion that in diabetic gangrene we are dealing with an atherosclerotic or arterioselerotic process." The causes of gangrene develop slowly, and consequently gan- grene itself is usually preceded by a long duration of symptoms. ( harcot pointed out intermittent claudication as a precursor of ^er: Archives of Diagnosis, April, 1915. TREATMENT OF COMPLICATIONS 425 gangrene, and coldness of the extremities and numbness and pains in the feet are common signs of approaching danger. The prophylactic treatment of gangrene is seldom preached, but it is important. All the measures which are suddenly advised when gangrene appears should have been adopted in a modified and appropriate form years before its appearance. First and foremost the emphasis in prophylaxis must be placed on treatment of the diabetes, and those conditions which might predispose to arterio- sclerosis. The treatment of the diabetes follows the same lines which have been laid down in the preceding sections, and the treatment of the arteriosclerosis need not be discussed here. The most important dietetic considerations are the avoidance of acid poisoning, the use of a low protein diet, and the maintenance of the weight of the patient by varying amounts of fat. Treatment of the patient as a whole should be directed to improvement of gen- eral circulation, especially that of the lower extremities. Patients predisposed to gangrene must be urged to walk for short intervals several times a day, and then regularly three times a day go through such gymnastic exercises as will bring about a free now of blood throughout the feet. Patients should not remain long in one posi- tion. When sitting it is desirable to use a cricket (in New England, but a "foot rest" in other sections of the country). Hot foot-baths should be encouraged, and hot-air baths may be employed. In carrying out these procedures care should be taken to avoid exhausting the patient by so arranging beforehand that the patient will be comfortable during the application of treatment. Beware of burns. Whereas active hyperemia will improve the circulation, passive hyperemia is dangerous and should never be employed. The treatment of actual gangrene demands the closest cooperation oetween physician and surgeon. Treatment should be prompt and energetic delay is dangerous. Stetten, 1 a surgeon, in an excellent paper, reports remarkable results in the conservative treatment of gangrene, but it is evident that his results were due first to the fact that he was a surgeon with an extensive surgical experience, and not a general practitioner, and second to the fact that the medical treatment which he gave his patients was along modern lines. His results show what can be done. While undoubt- edly early cases may do far better than formerly with medical treatment, and often apparently hopeless cases may clear up, the physician who expects quick results should control his own opinion as to the favorable progress of the disease by that of his surgical colleague. Don't try to see how near the danger-point you can go before surgical interference. Too often surgical treat- 1 Stetten: Jour. Am. Med. A^n., 1913, Ix, p. 1120. 42() T It K ATM EXT ment applied by physicians is atrocious, and I always long to see a real surgeon called in. Procrastinating medical methods often do far more damage than prompt surgical intervention. Even in the old days early operation usually resulted in recovery, but when the patient was put through a fat-protein diet for weeks and then the surgeon was called upon to amputate in the presence of acid poisoning, death frequently occurred, as we might expect; but it was not the operation which caused the death, but the poor medical treatment prior to the operation which the physician prescribed. Stctten recommends (\) treatment for the diabetes, and this has already been described; ('2) the use of active hyperemia already referred to, and (o) local treatment. Absolute rest of the part is essential and the strictest asepsis. Moist dressings of sterile physiological salt solution are to be constantly applied and changed twice a day; antiseptics are to be avoided; beware of tight ban- daging; the extremity should be immersed twice a day in sterile salt solution, and in this the dressings may be soaked off. If the extremity is to be baked the dressings may be left on, but the greatest precaution should be taken to prevent a burn. Treat the wound tenderly. Avoid forcible removal of sloughs. In general, rely upon expectant treatment, but do not hesitate to open pockets of pus. If operation is necessary, use nitrous oxide or nitrons oxide and oxygen or spinal anesthesia, but avoid ether and under no circum- stances use chloroform. Local anesthesia is objectionable because of trauma to the part. Spinal anesthesia may be employed. 1 hiring the operation avoid the use of an Esmarch bandage. The absence of a palpable pulse in a popliteal space is not a contra-indication to amputation below the knee, and a low amputation is by no means to be shunned. The postoperative care of the patient, demands the best of nursing. The patient should be frequently turned from side to side and should be urged to take deep breathing exercises. Older writers discredited operative measures when acid poisoning was present. Today we know that the appearance of acid poisoning should be prevented, and that if it occurs it can usually be dispelled by feeding small quantities of carbohydrate ."> per cent, vegetables and one or two oranges but otherwise no food, and subsequently the operation may take place. TAISLK 1X0. Dnt.vno.v OF I, IKK OF DIAHKTIC PATIK.VTS AFTF.K OXSF.T OF ( !A.\<;KI:\K. Living. s I'ndcr 1. 1 :; 1 ~T~ : 7T'T TREATMENT OF COMPLICATIONS 427 Results. The duration of life of diabetic patients after the onset of gangrene has varied in my 37 cases from less than one year to eight years. The details are shown in Table 180. From the above table it will be observed that among the fatal cases, two-thirds succumbed to the disease within the first year, whereas among the living cases two-thirds have already gone beyond the first year. I see no reason why statistics for this com- plication in diabetes should not rapidly improve. In my recent fatal cases the explanation of failure lies in manifest errors in treat- ment, either made by myself or by other physicians. Mistakes have been so plain that they should readily be avoided in the future. (Seep. 71.) Operations were performed upon 18 of the patients, and the remaining 19 were treated medically. Table 187 leaves little doubt as to the efficacy of the two methods : TABLE 187. OPERATIVE AXD XOX-OPERATIVE TREATMENT OF DIABETIC PATIEXTS WITH GANGRENE. Duration of ' Surgical treatment life after onset of gangrene. Fatal. Living. Fatal. Living. Years Uncer- | Under tain. 1. 12 13 13 111 1 Of the 18 cases treated surgically, 4 are still alive, but of the 19 treated medically I know of but 2. Furthermore, of the 18 cases treated surgically, 7, or 39 per cent., died during the first year, but of the 19 cases treated medically, 15, or 79 per cent., suc- cumbed during the same period. The surgical treatment of diabetic gangrene, in my experience, though executed by many different surgeons (under all sorts of conditions), and with gangrene extending over varying lengths of time has been twice as successful as medical treatment. There is little doubt in my mind but that if the beginning of gangrene was as noisily ushered in as an attack of biliary or renal colic the results of treatment would be far different. Death from gangrene is frequently now, as in the past, the result of procrasti- nation on the part of the physician and patient, the inauguration of a fat-protein diet and ether anesthesia. Surgery often receives, but seldom deserves, the blame of a fatal issue. 5. Care of the Skin. Pruritus, Furunculosis, Carbuncles. Pruritus frequently occurs in diabetes. Pruritus pudendi invariably will vanish within a few days after the disappearance of sugar from the urine, but general pruritus may be annoying and persist for weeks. 42S THE ATM EXT I do not remember to have observed its presence in young diabetics. Yon Xoorden has found sodium salicylate internally almost a specific in the treatment of general pruritus. In using it bear in mind the confusion which may follow in examining the urine for sugar and diacetic acid. I have not employed it. If pruritus pudendi does not clear up promptly, as the urine becomes sugar- free, an examination will probably disclose a prolapse, leucorrhea or urinary incontinence. Ixest in bed, absolute cleanliness, simple douches, and the simplest of ointments, are indicated. The free use of oil to prevent irritation during micturition is helpful. I have known applications of yeast and water (yeast cake rubbed up with 1200 to 500 c.c. of water) to be efficacious. It is well to bear in mind that patients with chronic nephritis are often relieved of pru- ritus when the protein in the diet is reduced. Infections of the skin are apparently less common now than formerly and this may be attributed to the improved hygiene. Such infections are and should be rare in diabetic patients under treatment. They demand immediate, thorough, yet gentle, treat- ment. One of the first duties of the physician is to tell diabetic patients to keep the skin clean and to report the beginning of an infection at once. Patients should be warned of the danger from slight wounds, should specifically be advised not to allow mani- curists or chiropodists to draw a drop of blood, and cautioned to promptly report any injury to the skin. Absolute cleanliness of the body is essential. Formerly the increased percentage of sugar in the tissues was held responsible for the presence of infections, but recent work has tended to disprove this theory. It is of interest that von Xoorden has observed such infections to be more common soon after the development of diabetes than when the cases were well established. One would expect the reverse if either the sugar or acid in the tissues predisposed to infection, because' these tend to increase as the diabetes progresses. Subcutaneous injections, whether of water, salt, solution or drugs, should be avoided. It is common for salt solution or solutions of sodium bicarbonate', when injected subpectorally, to result in abscess. If there is the slightest tendency to fnrunculosis, 1 at once adopt simple 1 measures analo- gous to those described by Bowen. 1 The patient is advised to wash the whole body twice a day with soap and water, using a wash cloth or piece of flannel, and to dry the skin without rubbing, so as to avoid breaking open any pustule; the whole body is then bathed with a saturated solution of boracic acid in water, with the addition of a small proportion of camphor water and glycerin. I have often used a solution of two parts alcohol and one part water to advantage, 1 Hovvon: Jour. Am. Mt'd. Assn., 1010. Iv, p. 201); Boston Mod. and Sur. Jour., 1017, <-lx ;vi, [). 90. TREATMENT OF COMPLICATIONS 429 but I notice that Bowen in his second paper still prefers the boracic acid. Individual furuncles may be treated with the following ointment according to Bowen: Boracic acid 4 Precipitated sulphur 4 Carbolated petrolatum . 30 One should be careful, however, not to overtreat the skin. Harm may result from frequent dressings. The simplest lotions should always be employed. In severe cases the patient should be put to bed, all linen changed twice daily, and the patient treated in as aseptic a way as possible. In a few cases vaccines have appeared to be of marked benefit. "This procedure, thorough bathing and soaping, the application of the borated solution, and the dressing of the individual furuncles, is repeated, as has been said, morning and night. A further point of vital importance relates to the clothing that is worn next to the skin. Every stitch of linen worn next to the skin should.be changed daily, and in the case of extensive furuncu- losis all the bedclothing that touches the individual, as well as the night-clothing, should be subjected to a daily change. Naturally, this treatment must be continued for several weeks after the last evidence of pyogenic infection has appeared, and this fact must be emphasized to the patient at the outset." (Bowen.) Energetic dietetic treatment does not show so satisfactory resul.ts in the presence of furunculosis as in pruritus pudendi. Infections greatly increase the difficulty of rendering the urine sugar-free, and a small boil may cause a trace of sugar to persist for days. It is easy to keep the patient on a low diet too long. These patients require the most careful management, and as acidosis easily develops all the measures adopted to prevent and combat acidosis should be employed. Neglected cases are the most dangerous. Carbuncles cause the death, directly or indirectly, of many a case of diabetes. It is remarkable, however, how satisfactorily these can be treated, provided that prompt measures are adopted from the beginning. The urine of the patient should at once be made sugar- and acid-free, and the very best surgical skill invoked. Bungling medical surgery is painful to behold. In connection with dietetic and surgical treatment, vaccines may be used with advantage. Fourteen cases of carbuncles were readily found in my records, but I am sure that others exist. Of this number death followed the carbuncle in at least five instances. It is my impression that a good prognosis can be given provided treatment is begun early. The surgeon should be encouraged to adopt the same measures which he would adopt in a non-diabetic individual. 430 TREATMENT Metnstases resulting from the carbuncle were numerous in two of the cases: No. ">lo and NO. 77(>. Case No. ")!:) is reported under Cases Unsuccessfully Treated by Fasting, p. '.ft'.}. Case No. 77(1, age at onset fifty-two, li)()S, first seen thirty-four days after the beginning of a carbuncle in Octo- ber, 1914. At the time there was extensive ulceration in the neck; '.}.() per cent, of sugar in the urine. lie died five weeks later with metastatic abscesses and undoubted septicemia. Case No. 901 (see p. '.ftO) died finally of coma with extension of the carbuncle. An apparently hopeless case may recover. Such was Case No. Sl 7; age at onset of diabetes, seventy-two; first seen a year later, in -January, 191"), ten days after the beginning of a carbuncle. Urine 2400 c.c., 4.2 per cent, of sugar. Examination showed two car- buncles on back of the neck, extensive ulceration on the right side of the neck and another area of ulceration in front. The physician described him as "drowsy by day and delirious by night." On account of his age and pitiable condition, it was debated by those in attendance as to whether it was justifiable to even attempt further' medical or surgical treatment. Finally fasting was begun, with the result that he became sugar-free, and, except, for a brief interval, has remained so ever since (December 1, 1910). In July, 191"), the patient was eating vegetables, a little fruit and potato twice daily, and felt well save for extensive pruritus. The insomnia which this caused made him miserable, but he was well enough to relieve it by spending his evenings at Revere Beach picture shows. Christie's 1 case was severe, frequently showing high dextrose- nitrogen ratios while taking 12 grams sodium bicarbonate daily, was fasted eleven days, and was finally discharged sugar-free upon a diet containing carbohydrate' DO grams, protein ">."> grams, and fat 22.") grams. No operation upon the carbuncle was performed. (i. Care of the Teeth. Many diabetic patients have poor teeth or no teeth at all. Approximately 41 per cent, of .'>()() cases recently seen had either poor or false teeth. Poor teeth are by no means necessary, even in diabetes of long duration. Cases which are carefully treated from the outset and are cautioned about care of the teeth have little more trouble than healthy individuals. Then, too, patients with diabetes of long duration are occasionally seen with perfect teeth. It is my impression that the changes in the teeth an' subsequent to the onset of the diabetes, but I am open to conviction upon this point. It is never advisable, however, to advocate radical treatment of 1 Christie; Jour. Am. Mod. Assn., 1017, Ixviii, p. 170. TREATMENT OF COMPLICATIONS 431 the inflamed gums at the outset, or to allow seemingly poor teeth to be extracted. Over and over again, as a result of constitutional treatment, the condition of the mouth improves, and loosened teeth become firm. Only when it it seen that further progress is not to be expected should operative measures be employed. The extraction of teeth has probably often resulted in the death of the patient, due presumably to the anesthetic. In illustration of this point I would cite Case No. 729. (o) Death Subsequent to Extraction of Teeth Ether Anesthesia. Case No. 729, female, married, teacher, first seen May 1C), 1914; no diabetic heredity. Greatest weight 163 pounds, but 124 at the first visit. One child born in 1905 and well. In September, 1907, sugar first found while five months pregnant. Confinement was normal, and the child is living and well. The patient became sugar- free and remained so until January, 1911, when she was again pregnant, but miscarried in the following March. In July, 1911, she became pregnant and the child was born at term and is also well. Sugar was absent at this time, but it soon came back and persisted. A miscarriage occurred in February, 1914, and there was a history of two other miscarriages. Under dietetic treatment at the New Kngland Deaconess Hospital in June. 1914, the sugar decreased from G.2 per cent, to zero, and the patient acquired a tolerance varying between 15 and 45 grams. On April 2(>, 1915, three months after her last visit to me and without my knowledge, all her teeth were taken out with ether as an anesthetic. The extraction took place at the dentist's office. " She was very sick after getting home, and all that night and the next day complained of the awful weakness and pressure, so that it was hard for her to breathe. April 28 her mind was wandering, and it was difficult to bring her back to con- sciousness." She died in coma April 30, 1915. In contrast I would cite Case No. 753, age at onset thirty- three, who first came under observation at the age of thirty-nine, in July, 1914. He was seen only once, but on October 28, 1915, entered the hospital and the course of the case is shown by Table 188. On [Monday, November 15, the patient was given gas and oxygen by Dr. Freeman Allen and the teeth extracted by Dr. Ashland, who took great pains to avoid trauma. The patient was up and about the ward the same evening, and so comfortable as to attend the nine o'clock clinic the next morning, and two days later started for home with gums practically healed. (See also p. 341). The same rules which govern the use of anesthetics in surgery apply in dentistry. Gas or gas and oxygen are undoubtedly the best anesthetics. Novocain may be advantageously employed and obviates in these days the dangers of general anesthesia, but it should be used with caution because the infiltration of the tissues 432 TREATMENT may be distinctly harmful. The following suggestions are based upon notes made for me by Dr. W. I. Ashland, to whom 1 am much indebted for tins and other favors. TABLE 1SS. ILLUSTRATION' OF USE OF GAS AND OXYGEN AS AN ANESTHETIC WITHOUT SUBSEQUENT Aciuosis. CASK No. 7/53. >. = Diet (grains). -^ 1. ~ C - r v; u - ^ 5 c L. Dato. tj " - S ^3 ~ ~ & " .S M 7 i C S ? _E' - *-' r. ~ ~ -~ _ ^ " J c * ~ = o - ^5 ^ >. ^ o - - 1 - c ~ - - X . ' ^ "3 ~* "i ". ~- <~ *~ -- -f. O ?H t- c ^ K < 1014 Julv 20 2400 1035 98 30 3420 10H7 138 Dor. 11 40SO 103(1 08 L915 Fob. 14 3300 1030 30 Oct 2S 1 ''OO 1031 1 ! ! 1) ()O 193 30 20 000 14 20 5 100 + 124 30 SOO j . . 3 12 123 0.18 31 050 1021 3 12 Nov. 1050 1023 20 01 72 1 (X)s +20 120 14 1000 ... 15 , 1050 1010 , . . . . 110 Tooth extracted an cstlu lie, u ,_ as an d oxy iron. 10 1OOO . . n 2; 30 ,)!.> J-M "So soon as diabetes is diagnosed, dental prophylaxis should begin. Regular visits at frequent intervals should be made to the dentist in order that the teeth may be thoroughly cleaned. The dentist should take great pains not to injure the gums. Free food de-posit should be removed from the teeth surfaces, but no attempt made to scale under the gums. The teeth should be carefully polished with a rubber disk and not touched with a rotary brush. Later, when the diabetes is under control, the teeth may be more thor- oughly cleaned, and all calculus removed. As this is an exceedingly difficult task, two separate appointments should be made. The gums may then be lightly touched with tincture of iodine. "A tooth-brush should be medium sized, with soft bristles. A tuft arrangement of bristles is excellent or those spaced about an eighth of an inch apart, because either of these types of brushes will allow the spaces between the teeth to be reached. A rubber tooth-brush may he employed in cases with inflamed gums, although the best results cannot be obtained with this type of brush. The material which cannot be removed in this manner had better be left Ifor the dentist. A second brush with two or three rows of bristes might advantageously be used for cleaning isolated teeth. "Any tooth powder or tooth past*' may be used providing it is not gritty. A paste is the more convenient, economical, and less likely to contain gritty substances. Pure precipitated chalk, Castile soap, or even salt and water may be employed. TREATMENT OF COMPLICATIONS 433 "Lime water or any simple mouth-wash may be used, but for their cleansing properties rather than antiseptic quality, which they do not possess. The mechanical cleansing of the teeth is the essential feature of oral hygiene. "The patient should be told that in cleaning the teeth the chief problem is to cleanse the interproximal spaces with a minimal amount of injury to the adjacent tissues. The exposed teeth surfaces require little attention. Do not brush across the teeth. Do not begin the brushing in the same region twice in succession, as when the brush is first used it will be found harsh. " Before brushing the teeth, rinse the mouth with tepid water, in order to remove free deposits. "The brush should have a rigid handle, but be held in the hand with a yielding grip. Brush the teeth thoughtfully, with an up-and- down, rolling or rotary motion, touching the gums hardest on the downward stroke. This will be found difficult to accomplish, but practice will bring very satisfactory results. The teeth should be brushed after each meal, upon rising and on retiring. "After the use of the brush the mouth should be rinsed with tepid water, to remove any loose material remaining in the mouth. "As an adjunct in the care of the mouth and teeth, floss silk is an absolute necessity. The floss silk employed should be fine and preferably of the 'ribbon' type. The silk should be carefully forced with a sawing motion between the teeth, being especially careful not to strike the gum a sharp blow. Push the floss down the side of one tooth and up the side of the other. Continue this until you have cleansed between all teeth. This should be done at night after the last brushing. "After the teeth have been brushed and flossed, the mouth should ne rinsed with lime water, and this should complete the toilet." 7. Constipation and Diarrhea. Constipation is the rule in diabetic- patients. It is less common with restricted than with excessive diet. Occasionally one is gratified by the abundant use of vegetables relieving an existing constipation, but this is seldom the case. If more care were taken to select those which lead to a slight gas forma- tion, such as cauliflower, onions, cabbage and spinach, better results might be secured. The gas prevents hard fecal masses. Thus one of my patients who has a diabetic garden regulates his bowels by beginning his breakfast with a slice of raw cabbage. Constipation is a serious menace if the caloric intake is large; it has caused no dangerous symptoms in my experience when the patient has been fasted. Constipation commonly precedes coma and seriously embarrasses treatment of the same. Hodgson empha- sizes the importance of keeping the bowels free in diabetes, and his large experience deserves attention. 28 434 TREATMENT Physical exercise is the best treatment for constipation. Since I have encouraged the patients to walk outdoors five times a day and to take up other forms of exercise, there has been less need of drugs for constipation. The constipation of Case No. 34S, whose prostate was successfully removed by Chute, vanished when lie began to saw wood. Musfidye. Mr. Gnstaf Sundelius, who has been of so much help in giving massage and exercises to my patients with inanition and gan- grene, and in training them to massage and exercise themselves, has furnished me the following sets of exercises for the relief of constipation. HOME EXERCISES FOR CONSTIPATION (SUNDELIUS). Group 1. Eaxy to Execute, Suitable for Weal; and Elderly People. 1. Abdominal Kneading and Stroking. Kneading. Lying down, with knees slightly drawn up. Place hands one on top of the other on the abdomen at the right groin; with small circular movements and deep pressure work upward until the ribs are met, then across toward left, following the boundary line of the chest, then downward to the left groin. Repeat twenty to fifty times. Stroking. With hands similarly placed, make long, steady and deep strokes following the same route. Repeat twenty-five to one hundred times. 2. Ley Rolling. Lying down, take hold of both legs just below the knees, press the knees up close to the abdomen, then carry them apart, then down and inward until they meet again, thus letting the knees describe two circles. Repeat ten to twenty times. '.\. Abdominal Compression. Standing against the wall with hands clasped behind neck, draw the abdomen forcibly in, using the abdom- inal muscles, hold a second, then let go. Repeat ten to forty times. 4. Trunk liolling. Standing with hands on hips, feet apart and legs well stretched, roll the upper body in a circle on the hips by bending forward, to the left, backward and to the right. Then reverse, and repeat six to twelve times each way. (iron]) 2. lieiptlre More Force ami Should be Uxed by Strong People Only. 5. Leg Lifting, Single and Double. Lying on the bed, or better still, on the carpeted floor, with hands clasped behind neck, lift the well-stretched legs way up, slowly letting them down again, either alternately or both together, Repeat six to twenty times. TREATMENT OF COMPLICATIONS 435 6. Trunk Bending Sidewise. Standing-with hands clasped behind neck, bend body way over to the left, then back, in upright posi- tion and then over to the right. Stop a second at the extreme bend positions. Repeat five to ten times. 7. Trunk Bending Forward and Bade /card. Standing with arms straight up in the air, bend forward as far as can be done without bending the knees, then upward and slightly backward. Repeat six to ten times. 8. Trunk Falling Backward. Sitting on a stool, with hands on hips, legs outstretched and feet braced under some heavy object (for instance, a radiator), let the upper body slowly fall backward, then raise to sitting position and repeat four to eight times. Mineral oil, salts, and enemata can be used when active treatment is necessary, but when daily measures are required one should attempt to depend upon the diet. If the patient's tolerance allows the use of an orange in addition to 5 per cent, vegetables, frequently nothing more is necessary. Great pains, however, should be taken to impress upon the patients the necessity of preparing coarse vegetables in a simple manner. It is perfectly possible to cook cabbage, cauliflower, turnips, parsnips, radishes, cucumbers, and onions so as to be unirritating to the digestive tract and yet preserve their laxative qualities. Patients whose tolerance allows potatoes are encouraged to eat them with the skins on. Diarrhea should be carefully avoided. Drugs which I have found most satisfactory have been cascara sagrada, compound rhubarb pills, aloin, strychnin and belladonna tablets, or aloin alone. The possibility of prolonging acid poisoning by the use of castor oil must be borne in mind. Mineral oil may be advantageously employed. Once a positive ferric chloride reaction followed its use without any other known cause. The oil may have been adulterated, for one could not expect the mineral oil to be burned in the body. The various aperient waters may be employed, but loose movements should be avoided. Sodium or magnesium citrate in doses of one or two teaspoonfuls helps to relieve the constipation, but the alkali which it furnishes may prolong the presence of a diacetic acid reaction. Bran bread is often helpful (see pp. 5!)5 and 531 ). Diarrhea is a most serious complication. Rest in bed, fasting and preservation of body warmth should be carried out from beginning to end of the attack. Hot water, solutions of hot, weak tea, cracked cocoa, and brandy may be given. When the indication is plain the bowels should be cleared out with castor oil, or what I find works far better, half a Seidlitz powder. Avoid strong 430 TREATMENT cachartics, for diabetic patients with diarrhea are usually too feeble to withstand their action. Should the cause of the diarrhea below down in the intestinal tract, eneinata should be used. If any drug is employed, it should be opium. The return to the diabetic diet is rendered easy by the use of sugar-free milk, lean meats, eggs, gelatin, and vegetables which have been most carefully prepared and sieved. These carefully prepared, tender vegetables are frequently better borne than a diet containing considerable quantities of albuminous and fatty food. If the carbohydrate tolerance allows, oatmeal gruel in divided portions is most desirable. Calcium carbonate in doses of 1 or '2 grams has been employed to offset the diarrhea which occasionally occurred when sodium bicar- bonate was freely given. X. Neuritis. Neuritis yields as obstinately to treatment in dia- betes as when it occurs associated with other diseases. Unfortunately it does not always disappear with the removal of the glycosuria. Possibly this may be accounted for by an increased quantity of sugar persisting in the blood long after it has disappeared from the urine. The most Frequent type of neuritis which occurs in diabetes is sciatica. From the help which modern orthopedic treatment has afforded my patients, I am inclined to consider the development of sciatica as having little causal connection with the disease, except insofar as such a complication may appear in any individual who lias lost weight and whose muscles afford only lax support to the large joints. Hydrotherapeutic measures avail much, but fixation of the back, particularly of the sacro-iliac joints, brings most relief, as my orthopedic friends have been able to demonstrate upon intractable cases. In some cases of this type symptoms persist despite such treatment, but sometimes these can be explained as habit pains. Moist flannel applications changed every one or two minutes almost, invariably give* comfort. These are appliod by placing a dry flannel upon the skin, for the sake of protection, with the moist, hot, freshly wrung-out flannel laid over it and the pro- truding ends of the dry flannel covered over to retain the heat and meanwhile a third flannel is kept in hot water for future use. Dry heat is less efficacious. Applications of this type continued for ten minutes twice a day are of the greatest assistance in lessening pain. Mental diversion, brought about by change in surroundings, often produces good results. Similarly, the improvement in the general condition which follows rigorous dietetic and physical treat- ment is of great advantage. Such cases are particularly helped by a brief stay in a hospital, or by having a well-trained diabetic nurse in the home. The type of neuritis almost, invariably encountered has occurred in the lower extremities. When diminished sensation, pain and TREATMENT OF COMPLICATIONS 437 tenderness exist, there is always doubt whether the condition is a true neuritis or simply the result of poor circulation. It is certainly true that a pure type of neuritis in other parts of the body in diabetes is seldom seen. For this reason I am inclined to believe that the poor bloodvessels of the lower extremities are the chief offenders. Alcohol is also an important etiological factor. 9. Eyesight. Ocular complications in diabetes are frequent. Von Xoorden reports that of 279 patients, 58.3 per cent, had trouble with the eyes attributable to the diabetes, and of those patients over fifty years of age, SO per cent. He summarizes the results of the examination of the eyes of 477 diabetics, taken in succession. Defects were found in 279 instances; in 259 of these cases there was no etiological cause other than the diabetes. A summary of the data is given in the following table: TABLE ISO. OCULAR COMPLICATIOX.S i\ DIABETES (vox Xo. of Cases. Retinitis ................. 81 Retrobulbar neuritis ............. 23 Atrophy of the optic nerve ........... IS Cataract ................. 62 Iritis .................. 2 Amblyopia without organic change ......... 33 Diabetic myopia .............. 21 Other conditions .............. 39 Fortunately complications often vanish after the patient has been free from sugar and acidosis for a few weeks. If treatment will accomplish so much when complications have appeared, how much it must accomplish in prevention of complications. One cannot too frequently impress upon the patient that treatment is instituted not simply to keep sugar out of the urine, but to save him from a great variety of serious incidental diseases. Conversely, with the institution of progressive treatment the eyesight often temporarily fails, and I have more than once had patients in the course of a few days become unable to read, and once even to recog- nize individuals. The explanation of this condition is undoubtedly due to changes in the water balance of the body and the disappear- ance of sugar. These incidentally involve the lens, causing refractive changes. I have always been able to reassure such patients with the statement that the return of their former eyesight and even an improvement in the same would eventually take place. It is remarkable to how great an extent the eyesight may return in a long-standing but neglected case of diabetes. Case XV). LSI was unable to read his name when he entered the hospital, but after three weeks' stay he was able to do so. Case Xo. 924 showed marked 1 Von Xoorden: Lor. rit., p. 217. 43S TRKATMKXT failure of eyesight on October 2S, 11)15. Her eyes were examined at this time by Dr. F. M. Spalding, who reported a shrinking and flattening of the lenses. With a convex lens ( + 12) in front of each eye the vision later came up to normal. Comparatively few of my .patients have had cataracts. The number is so small as to be surprising, and it is probable that I have overlooked the presence of cataract. One can easily understand that Graefe found cataracts in one-fourth of all his diabetics, but Frerichs, in 400 patients, discovered 10 cataracts, and Seegen also reports cataracts present in 4 per cent, of his series. One of my cases, Xo. 507, aged seventy-one years, onset doubtful, after preparatory treatment became sugar-free, and the cataract was successfully removed by Dr. Edward ,R. Williams. It is inter- esting that the haziness of the lenses in beginning cataract may diminish with constitutional treatment and improvement of the diabetes. One instance of paralysis of the abducens muscle 1 distinctly recall, though I cannot identify the record. Case Xo. S0(>, cited p. 27S, showed a retrobulbar neuritis, but it was probably due to syphilis and not to diabetes. M. SURGERY AND DIABETES. 1. Elements Predisposing to Surgical Failure. Diabetic patients stand in need of the surgeon far more than normal individuals or those suffering from most other chronic diseases. The frequent occurrence of carbuncles and gangrene is the chief reason. In the past the mortality following surgical interference, whether for the relief of diabetic complications already alluded to or for surgery in general, .was relatively high, so that the rule has been general to avoid surgery in the presence of diabetes whenever possible. In the last few years the mortality following surgical operations has decreased, until now there is less need for conservatism. But when- ever delay is not dangerous, the rule still holds to defer surgical interference. It is surprising how many apparently severe conditions clear up when the diabetes has been appropriately treated. If surgery is indicated, however, diabetes is no excuse for its non- performance. On March 17. 1017, a stone was removed from the common duct of an old man, Case Xo. 007 (Table 100) by Jones, and a large 1 abscess of the upper lobe of the right lung of Case Xo. ll'.'iO ('Table 101 ) drained by Scudder following a resection of portions of three ribs. Chute has removed the prostate from 4 diabetic patients without a fatality, Hubbard has performed two Cesarean sections and extensive repair operations successfully, and Lund, as SURGERY AND DIABETES 439 long ago as February, 1908, removed a large fibroid 1 without any complications, and has since obviated the necessity of as many amputations for gangrene as he has successfully performed such upon others of my diabetics. Before undertaking an operation upon a diabetic patient the surgeon should thoroughly understand the dangers with which the patient has to contend and the elements which favor surgical success. The dangers are four: acid intoxication, slow healing of wounds, exhaustion and lack of exercise. (a) Acid Intoxication. Diabetic coma has been elsewhere dis- cussed, and its prevention will be taken up under the medical treatment of surgical cases in what follows. (6) Slow Healing of Wounds. Statements regarding the slow heal- ing of wounds of diabetic patients are proverbial, but often untrue. Allen did not find this a necessary condition in his operations upon animals, and through experimental work he showed that the presence of an excess of sugar in the blood was not deleterious. Xo work that I recall, however, has been done upon the effect of an excess of j3-oxybutyric acid and the other acid bodies in the blood. Such studies are needed. It is possible that the supposed slow healing of wounds in diabetic patients is a tradition handed down from preaseptic days or has arisen from the fact that operations in septic conditions have been too conservative, or finally, that the frequent employment of surgery in old and debilitated diabetics, in whom healing even without diabetes would be slow, has helped to make this impression. The more I see of diabetic surgery the less differ- ence I observe in it from surgery of the non-diabetic. (c) Exhaustion. Too often in my experience blame has been attached to the surgeon when the patient wasted away after an operation. More frequently the blame should rest upon the shoul- ders of the medical attendant. With better medical treatment this element, which has played so large a figure in the past in the surgical mortality of diabetic patients, should decrease. (<7) Lack of Exercise. Diabetic patients, otherwise properly treated, thrive upon exercise; deprive them of it and they go down hill. Pulley weights should be attached to the bed; they should be encouraged to move about, to exercise any muscles not incapaci- tated; a masseur may be of great help, especially if he combines massage with resistant motions. 2. Elements favoring surgical success are: (a) Good medical care before and after the operation. (b) The method of anesthesia. (c) Employment of asepsis rather than antisepsis. ((/) Avoidance of trauma. 1 Joslin: Am. Jour. Mcd. Sc., 1913, cxlv, p. 474. 440 TREATMENT (tt) Good Medical Care. If the diabetic patient is free from sugar and acid lie requires no special preparation for operation, but if these products are present they should be removed according to the plan of treatment already described for any case of diabetes (see p. M05). A positive ferric chloride reaction is far more to be dreaded than a posi- tive sugar test both before as well as after an operation. Conse- quently, if the case is urgent, even in the presence of sugar, operation may be undertaken with comparatively little apprehension. If acid is also present a serious problem is presented, which must be solved for each individual case. At present writing I should consider the safest general rule to follow the routine method of treatment outlined for cases presenting acidosis and described on p. MS7. During the day prior to the operation, the freest possible adminis- tration of liquids should be favored. These should include broths on account of the salts therein contained, coffee, tea, water, and if necessary salt solution or tap water by rectum. I would specifically advise against the use of alcohol at such a time, because of the possibility of its upsetting the stomach. This is of all things to be avoided. If alkalis have not been previously given, they should not be begun, but if the patient has been taking sodium bicarbonate in doses under MO grams in twenty-four hours omit it; otherwise decrease the alkali MO grams a day. Liquids of the above character should be given freely up to as near the hour of operation as possible, so that the body will have a full water content with which to excrete acids, which may form as a result of the operation. I can conceive of a surgical emergency in which the patient has been actually fasted for a few days and yet acidosis may have appeared or increased. In this event the adminis- tration of MO grams of carbohydrate in the form of the juice of two or three oranges or 45 grams of dry oatmeal or flour, made into a water gruel, might be of great help. The dietetic treatment for the first twenty-four hours after the operation is fasting save for the addition of those' liquids which preceded the operation, but if the patient is able to eat, the same diet but in lessened quantities and without fat is allowed. One of the great advantages in not giving the patient sodium bicarbonate before the operation is that it need not be repeated after the opera- tion. A doctor's order to force liquids before and after operation might bring disaster by upsetting the stomach, but this can be avoided if salt solution and tap water are administered to a con- siderable extent by rectum and liquids are given cautiously by the mouth and at the first sign of need intravenously. It is a safe rule to administer 7~>0 to 1000 c.c. of liquid every six hours before operation to a patient who is excreting sugar and acid and ")()() c.c. every six hours to one who prior to the operation has been made su< r ar- and acid-free. SURGERY AND DIABETES 441 The next day after the operation, if the urine is sugar-free, the patient is gradually returned to his former diet, minus fat, being at the same time careful not to exceed his known tolerance for carbohydrate, and in the subsequent days, if sugar and acid con- tinue absent, the patient's customary dietetic habits are resumed. On the other hand, if sugar appears after the operation when it was not present before, it may be disregarded for several days because it may disappear of itself. (See Tables Nos. 190, 191, 192. Cases Xos. 34X, 097, 1239). Should this not occur, routine treatment of the diabetes should be instituted. Patients with severe diabetes or elderly patients should be sub- jected to very slight dietetic changes preceding operation. Unsuc- cessful results in the past have been caused in large part by these unfortunate individuals having been suddenly placed upon a strict protein-fat diet, to which they were not accustomed. In consequence the digestion was upset quite as much as would have been the case with a normal individual, and they developed acidosis as well. The fasting, incidental to a routine surgical preparation, predisposes to acidosis in a non-diabetic, and may still more in a mild or moder- ately severe diabetic patient unless previously accustomed to it. 1 Formerly severe diabetic patients were filled with sodium bicarbon- ate; this markedly influenced the fluid content of the body and often resulted in the release of so much acid as to overwhelm the kidneys, with resulting suppression of urine. (/;) Anesthesia. Every effort should be made to shorten the period of anesthesia no matter what the anesthetic, and to avoid apprehension and excitement on the part of the patient. Chloro- form is absolutely contra-indicated. Ether has many times been used with success, but it is often dangerous, and 1 believe the state- ment made upon p. 70 will bear repetition namely, that ether anesthesia is a burden which a light case of diabetes may easily bear, which may change a moderate to a severe case, and to a severe case prove fatal. Nitrous oxide or nitrous oxide and oxygen appear to be the best general anesthetics, provided they are administered by an anesthetist trained in their use. Postoperative acidosis in children is said less likely to follow nitrous oxide anesthesia than ether, and if it does occur is less apt to be severe. Spinal anesthesia has been satisfactory in several of my cases, though often undesirable because of the nervous tension to which it subjected the patient. Local anesthesia with novocain is the safest mode of anesthesia, but it should be delicately employed for fear of trauma to the susceptible tissues. In illustration of the above I would cite the charts of four cases: 1 Bradncr and Reimunn: Jour. Am. Mod. Assn., 1015, p. 1947. 442 TREATMENT Case Xo. 34S, male; age at onset forty-two years; operation for removal of the prostate at forty-nine, seven years later; spinal anesthesia appeared contra-indicated and light etherization was employed. It will be seen that a positive carbohydrate balance of 20 grams before the operation changed to a negative carbohydrate balance of 33 grams after operation, and that the acidosis was extreme after the operation, if the fasting condition of the patient is considered. The patient fortunately made an uninterrupted recovery and is now well and active. (See Table 190 and also page 434.) TABLE 190. CHART OF CASE Xo. 348. MODERATELY SEVEKK DIAHETES. REMOVAL OF PROSTATE. Urine. Diet. Carbo Sodium Total Carbo- Pro- hydrat e bicar- Diaeetic Nils suKar, hydrate, tein, Fat, balanr e, bonate, Date. acid. K'ns Kins. Kins. Kins. Kins. Kins. Kins. 1910 July 21 . . . 160 1915 July 9 . . . 15 +45 11-12 . . 20 30 50 + 20 July 11 , operation removal of prostate; I )r. A. L. C'hute; anesthetic, ether. July 12-13 + + 37 '.) 13-14 + + 33 -33 14-15 . . + + + 3.3 11 ()' -11 15- Hi . + + + 3.2 30 5 -31 s A uit. 1- 2 . . TABLE 191. CHART CASE Xo. 097 . MILD DIABETES. GALL-STONES. (.See p. 443.) Carbo- Carl).,- Sodium Diaeetic 1 ririe, total siiKar, hydrat e in diet, n\ (II ate balance, bicar- bonate, Date. acid. K'ns. gms. Kins. Kins. 1914 Feb. 2 . . . . . >.S r ; ) 17-18 ... . . 18-19 . . . + February IS, operation for gall-stone 105 + 105 ,8 35 8 - 27 12 Dr. D. F. Jones; anesthetic, ether. 105 60 - 45 18 110 34 75 101 So - 20 X 71 110 + 10 55 175 + 120 19 110 + 90 120 + 115 11 120 + 110 120 + 120 95 + 95 I 35 +31 0.x 1 no + l) () Mar. 17, operation for removal of trail-stone from common duct; Dr. D. F. Jones; anesthetic, ^as and oxygen, with 15 c.c. ether. Mar. 17 is ... + 1.9 15 20 + f> is 19 . . . 1.9 13 11 2 19 20 ... 2.2 25 11 + 10 25 20 . . 71 + 71 SURGERY A\D DIABETES 443 ( ase No. 097, male; age at onset forty-nine years; age at opera- tion for gall-stones, fifty-five; anesthetic-, ether. The tolerance on the day before operation was approximately lOo grains; on the day of the operation the quantity of sugar in the urine was 3o grams, and carbohydrate in the diet S grams, and a minus balance con- tinued for three days more. The patient completely recovered and remained well and active until the end of 191(1, when he had a recurrence of symptoms. A stone was then removed from the common duct by I). F. Jones. Anesthetic, gas and oxygen and 45 c.c. ether administered by Freeman Allen. Table 191, p. 442. TABLE 192. CHART OF CASK 1239. ABSCESS OF Luxe;. PAGE 444. Am- Total Diet in grams. Blood Blood Alveolar Date, 1017. Diaeetie nioma, arid. total gins. sugar in urine. Carbo- hydrate Pro- . tein. Fat. C'Oi>, sugar, air ('( )?, Alco- mm. Hg. per cent. mm. Hg. hoi. Feb. 17 Spec + + + Feb. 17-18 + + + 21 37 18-19 + + + 31 45 's '6 .. 0.17 19-20 H h 19 25 5 (I 20-2 1 + + + + 2.4 4 10 5 21-22 + +f+ 2.7 47.7 0.14 ;|;j 22-23 -r + + 2.8 2 10 5 23-24 + + + 2.6 2 24 . . . . 37 24-25 H h 3 GO 15 25-26 + + + ''1 8 92 12 30 . . . . 38 26-27 l.Q 7 68 24 15 27-28 6 5 72 41 30 28- 1 1.7 10 70 42 30 Mar. 1- 2 + + 1.4 2 15 70 61 30 2- 3 10 70 00 30 3- 4 4- 4/- 10 70 81 30 54.0 0.14 4- 5 1.3 10 81 82 30 5- + + 1.3 10 81 88 30 0- 7 + + + 1 15 84 88 30 7- 8 + + 1.2 3 20 84 82 30 . . . . 30 S- 9 + 1.0 15 56 21 30 9-10 4- 10 51 1 30 10-11 sl. + 24 77 40 30 11-12 1.0 10 00 43 30 12-13 -j- 3 24 82 01 30 13-14 + + 1.4 4 20 70 00 30 47.7 0.18 14-15 + + + 1.5 4 25 47 30 30 . . . . 35 15-16 + + + 1.6 8 37 46 12 30 16-17 v. si. + 1.1 7 40 OS 13 30 . . . . 38 17-18 4- 0.7 20 75 33 12 . . 35 Operation: Dr. C. I.. Scudder . An sthetir : Gas. and oxygen. 18-19 + 1.0 23 50 58 20 30 .. .. { :]!> 19-20 1.5 25 44 58 40 30 20-21 1 .5 9 34 56 40 21 21-22 1.3 7 34 56 59 30 22-23 + 6 34 63 75 30 23-24 1.1 3 34 56 09 30 24-25 . 2 34 63 85 30 25-26 6 39 63 85 30 26-27 1.0 2 29 63 85 30 27-28 4^ 2 29 06 03 30 28-20 0.8 2 20 66 03 30 29-30 + 2 31 68 01) 30 30-31 4^ 37 60 88 30 April 2- 3 . '. 40 74 03 May Sinus healed. Weight at entrance, 133j pounds; weight March 10, 131 \ pounds; no alkalis given 444 TREATMENT An abscess in the upper lobe of the right lung was successfully drained by (\ L. Seudder. The patient, Case No. 12.'W, first pre- sented symptoms of diabetes in September, 191(>, at the age of forty-six years, and three months later the disease was diagnosed. During January, 1917, pulmonary symptoms developed and in February, when he first came under my observation, after consul- tation with his physician and F. T. Lord, it was decided to treat the diabetes actively and the abscess of the lung expectantly for three weeks, and then operate if the pulmonary condition did not improve. From Table 192 it will be seen that the acidosis (diacetic acid and ammonia) was severe, though this was not revealed by the (X) 2 tension of the alveolar air or by the blood, presumably on account of the pulmonary complication. The acidosis was gradually, but not entirely, overcome. When it was seen that improvement did not take place the patient was operated upon by Dr. (\ L. Seudder under gas-and-oxygen anesthesia through a rectangular incision. A large abscess of the right upper lobe was evacuated after removal of two inches of the fourth, fifth, and sixth ribs. The cavity was large, receiving into its interior one large gauze sponge. Prior to the operation fat was decreased in the diet and the carbohydrate in- creased. The highest tolerance for carbohydrate reached in the three weeks prior to the operation was 19 grams, but now, May 14, it amounts to over S5 grams and the wound has closed. TAHLE 10:!. CHAUT CASK Xo. 173. MILD DIABETES. ONSET AT SIXTY-FIVE YEARS; OPERATION AT SEVENTY-THKEE VEAKS. LAPAK JTOMY. PACE 44,5. Urino, Total car- lota! bohydrates Sodium Oarbo- Diacctic suar, in did, Incur- hydrate Date. acid. KHuns. Knm'*- bonate. balance. 1!)07 Nov. 4 ........ 47 . . I'.IOS Feb. 10 ........ 170 170 1913 X<,v. 14 ........ 55 . . 55 1015 Feb. is ........ o o 100 o 100 Oct. 25 ........ 100 100 25 2(1 . ..... . . 2s 2< ....... IT 47 20 m ....... o o <;5 o 05 1910 May 1 ........ (I . . 1017 Jan. 1C, ........ (1.2 r ,') .. See also tables 191 and l'J5. SURGERY AND DIABETES 445 Case No. 173, female; age at onset sixty-five years, age at operation seventy-three; blood-pressure, 210. The operation con- sisted of a laparotomy for a complete procidentia, and perineal repair, which had seriously incapacitated the patient. Tolerance for carbohydrate before operation was approximately 100 grams; anesthetic, nitrous oxide, oxygen and 5 per cent, ether. It will be seen that neither sugar nor acid appeared after the operation. Successful convalescence until the nineteenth day, when a cerebral hemorrhage occurred. Fortunately recovery was nearly complete. (See Table 193.) (c) The introduction of aseptic methods accounts for many im- proved results in diabetic surgery. Minor surgical ailments, which were formerly treated with strong solutions of corrosive sublimate and hydrogen peroxide later required major operations. Weak anti- septic solutions only should be employed. (Compare the treatment described under Gangrene, p. 426). (<7) The avoidance - of trauma during operations is universally advocated, but this is especially to be borne in mind in diabetic surgery. The greatest care should be taken in handling the tissues. Skilful, delicate surgery has happy results in diabetes. Guard against a too free use of infiltration anesthesia. 3. Results. Omitting operations for gangrene and carbuncle, dis- cussed elsewhere, surgeons have performed the following 34 major operations upon my series of cases, with seven deaths : TABLE 194. SURGICAL OPERATIONS UPON* DIABETIC PATIENTS. Total number Results. Operation. of cases. Successful. Fatal. Gall-stones 4 4 Appendicitis 5 2 3 Cancer: breast 2 2 bladder 3 1 2 1 uterus 1 1 Prostate 5 5 Fibroid 5 4 1 Extensive perineal repairs 3 3 Exploratory laparotomy 1 1 Amputation of leg for septic knee ... 1 1 Mastoid 2 2 Abscess lung 1 1 Drainage bladder 1 1 Total 34 27 7 In Table 105 I have grouped 27 cases who were operated upon and under my care shortly before or after operation. There were but five deaths, two had cancer, one gall-stones, one pernicious anemia 1 One death was due to metastases and the other to pulmonary embolism. 440 TREATMENT and another a severe anemia. All patients who died were given ether, but the number of cases is too small to warrant far-reaching conclusions, for there were inanv who took ether and lived. TABLE 19.5. Sumac MY CARE AL OPERATION'S UPON DlABKTIC P.\TIKNTS UNDER SHORTLY HKFORK OH .A FT Kit OPERATION. Results. Operation. Total cases. Anesthetic. cases ' Ether <>97 Gas and oxygen, with Gull-stones .... 4 45 c.c. ether (J97 1 Ether Ether 'Ether 2'.) Appendicitis .... 3 Ether 'J4S Ether ( 'ancer of uterus . . 1 Ether 799 Cancer of breast. 1 Gas and oxygen and ether IONS Cancer of bladder . 1 Ether 49S Cancer of bladder . 1 Ether . . % . . . 57(i Cancer of bladder . 1 Ether . . *. . . Ether 348 Spinal anest hesia 559 Prostate . 5 Gas and oxygen 599 Spinal anesthesia 833 Spinal anesthesia 10S(> Drainage of bladder . 1 Novocain .... 1042 Ether 120 Fibroid . 3 Ether 127 Nitrous oxide Ether 60 Extensive perineal repairs 3 Nitrous oxide, oxygen and ether sequence . 173 Ether '{()S Exploratory laparotomy . 1 Ether :5:5:{ Abscess of lung 1 Gas and oxygen 12:;<) Mastoid 1 Ether 503 Fatal cases. G39 3 855 4 939 5 7216 Appendicitis. It will be seen that three patients with appendicitis died. (Table 194, p. 445.) The details follow: Case No. 5;5o was taken with acute appendicitis twenty-four hours away from his home, refused to go to a hospital and when he reached the surgeon, who operated upon him at once, "there was a quart of pus in the peritoneal cavity and thrombosis of many of the mesenteric veins." Case No. 54:2 was taken very ill with acute appendicitis upon a fishing trip and operated upon two days later, but died at the end of a week in coma. Case No. S55 is already reported under Cases Unsuccessfully Treated by Fasting, p. o50. 1 Two years after fir>t operation, stone removed from common duct. 2 Complicated with severe pernicious anemia. 3 Gall-stones, coma. Today an avoidable death. 4 Seen only once, four days after operation. '' Died of pulmonary embolism. 6 Complicated with severe anemia. SURGERY AXD DIABETES 447 Removal of Uterine Fibroids. Case No. 721 died after operation for removal of fibroid of uterus; age at onset forty-five years; age at operation, forty-nine; extreme pallor in fact operation was forced because of repeated hemorrhages. On entrance to hospital April 21, 1914, sugar was 3 per cent., diacetic acid present. At that time I did not employ fasting treatment, but the patient became sugar-free on May 7, with a tolerance for 15 grains carbo- hydrate. Shortly after she grew worse, acidosis reappeared. The initial condition of the patient, however, demanded operation. This was performed (anesthetic, nitrous oxide) and she succumbed forty-eight hours after operation. Today, with fasting treatment, transfusion and the use of gas and oxygen, it is possible that the patient might have been saved, but her condition was extremely grave apart from the diabetes. Case Xo. 127 is interesting as being the only one of my diabetic series in which sugar disappeared after the fibroid was removed. Age at onset, thirty -nine years; age at time of operation, forty-seven; following the operation, which was in the year 1900, there was marked suppuration of so extensive a character that drainage was established through the abdomen, the vagina and rectum. The patient at length recovered and nine years after operation was in good health and I thought the sugar had disappeared permanently, but on April 25, 191(>, which was ten years after the operation I was fortunate enough to secure a specimen of this patient's urine and it contained 104 grams. May, 1917. Urine, 0.7 per cent, sugar. Blood sugar, 0.22 per cent. Non-protein nitrogen, 34 mg. per 100 c.c. Removal of Prostate. The 5 cases operated upon recovered, though 1 died later of cancer of the stomach. Case Xo. 34S caused considerable apprehension after the operation because of the development of acidosis, undoubtedly due to the use of ether as an anesthetic. This case is reported in detail on page 339. Case Xo. 833, with onset at sixty-eight, was a mild diabetic and sugar-free at the time of the operation. His weight had fallen from 249 pounds to 1(>5 pounds. Spinal anesthesia was used. Case Xo. 599, with onset of diabetes at the age of sixty-two years, was operated upon two years later, April, 1915. Anesthetic, gas and oxygen. Case X"o. 559, with onset of diabetes at seventy-three years, was operated upon at seventy-four years, January 14, 1913. Spinal anesthesia was employed. X'one of these cases have shown the remarkable increase in tolerance for carbohydrate following the operation which has sometimes been observed by others and I remember to have seen reported. Strouse 1 states that of 38 operations upon diabetics in the last fifteen years at the Michael Reese Hospital, there have been 31.3 ' Strouse: Chicago Clinics, April 19 16, p. 37. 448 TREATMENT per cent, deaths, but there were no deaths among 8 patients who were properly prepared for operation. By properly prepared he evidently means made sugar- and acid-free and for a few days before operation given alkalis. Karewski 1 reported 225 operations with a mortality of 21 per cent. Fischer 2 had a total mortality of 4S.S per cent, with SO diabetic patients and 72.7 per cent, in the 11 operations for diabetic gangrene. Karewski 3 reports that 11.$ per cent, of 68 diabetic patients died of coma after operations on aseptic tissues, and 21.7 per cent, of 69 cases after operations on infected tissues. Berkman 4 reports 2 deaths among 26 cases at the Mayo Clinic during 1915. N. PREGNANCY AND DIABETES. A small quantity of sugar in the urine during pregnancy is so common an occurrence as to attract comparatively little attention, and the frequency with which it is met probably depends upon how often it is sought. According to Guelmuyden 8 glycosuria develops in 10 to 12 per cent, of all pregnancies and some have encountered it in 40 per cent. As a rule in such cases the sugar permanently disappears soon after confinement, although it may recur with succeeding preg- nancies, and ultimately a severe form of diabetes may develop. Such an eventuality is cited by Foster: ". . . there had been a pronounced glycosuria during two pregnancies which had entirely subsided after the birth of each child so that there was no sugar in the urine, and no dietetic restriction was employed. During the third pregnancy there was observed not only glycosuria, but also a moderate increase of thirst and following the termination of this pregnancy the sugar excretion and the symptoms persisted." Case Xo. o()9 showed sugar in 1X97 during pregnancy, but following confinement, with resulting dead baby, it disappeared, but returned in nine years in the form of moderate to severe diabetes. The association of this case with gall-stones and with involuntary fasting is described on p. 295. With our present knowledge it is quite possible that such an outcome could be prevented by active treatment of the glycosuria from the very start. Ten cases of pregnancy in which the patients showed small quantities of sugar have occurred in my series. In 1 Karewski: ('ited by St rouse. '-' Fischer: Drutsrh. Ztsrhr. f. Therapie, 1014, rxxxi, p. 40. :i Karewski: Deutsch. med. Wrhnsrlir., 1914, xl, p. S. 4 Berkman: .Journal-Lancet, 1910, xxxvi, p. .'->09. s Guelmuyden: Jour. Am. Med. Assn., 1914, Ixiii, p. 1798. e Foster: Diabetes. Mellitus, Lippiucott, 1915, p. 99. PRKGXANCY AND DIABETES 449 2 of these cases (Cases Xos. G98 and 318) the sugar was found before conception took place, but in all it was moderate throughout. Among these 10 cases there were fourteen pregnancies and eleven children, and both mothers and children are well. It is not strange that 4 of the cases occurred in physicians' families. (For a complete report of 7 of these cases, see Boston Med. and 8urg. Jour., 1915, clxxiii, p. 841.) Case No. 309 is described above and Cases Xos. 791 and 1017 were primipanv and delivered of healthy children. 1. Cases of Pregnancy Showing Large Quantities of Sugar. The gloomy outlook for pregnant women showing large quantities of sugar became somewhat modified when it was recognized that these cases could be divided into two groups, based on the appear- ance of sugar before or after pregnancy began. If pregnancy occurred in diabetic patients the outlook was considered far more serious than when diabetes first appeared during pregnancy. A somewhat more hopeful view of the whole situation was taken by Eshner, 1 and by Stengel, 2 and still more recently by Xeuman, 3 who reported G successful cases. It is not surprising that so little progress was made in the treat- ment of pregnant women with diabetes, because progress in the treatment of uncomplicated diabetes has been slow. However, the introduction of the therapeutic methods of Naunyn and von Noorden started an improvement, and a considerable number of pregnant women with sugar in the urine have now been sub- jected to the same kind of rational treatment which is employed in ordinary cases of diabetes. But it is only fair to say that the general practitioner still has a horror at the discovery of a con- siderable quantity of sugar in the urine during pregnancy. For this reason I shall record in some detail Case No. S12, in which diabetes of moderate severity occurring during pregnancy was successfully treated, but for the detailed records of G other cases would refer to the Boston Mcd. and Surg. Jour., 1915, clxxiii, p. 841. Another case of diabetes (Case Xo. 729), of eight years' duration, in which the tolerance for carbohydrate during the first seven months of pregnancy rose from 15 to 75 grams, is here reported for the first time, and Case Xo. 1070, with severe diabetes, alveolar CO 2 , 20 mm. Hg, is described as well. Before entering upon a dis- cussion of these cases, a summary of 5 other cases, not hitherto reported, will be recorded. Case Xo. 102 had four pregnancies and two healthy children during the course of her twenty-two years of diabetes, and Case X\>. 1018 had four babies and one miscarriage in ten years and is now in good condition. Case X'o. 729 (see p. 431) : Eshnor: Am. Jour. Mod. Sc., 1907, cxxxiv, p. 375. 2 Stengel: Pcnna. Med. Jour., 1907-S. ii, p. 900. 3 Xeunv.m: /tschr. f. -irtzl. Fortbild, 1913, x, p. 307. 29 450 T RE ATM EXT had two pregnancies with healthy children and two miscarriages during the course of her diabetes and finally succumbed to ether anesthesia. Cases Xos. 4o9 and 401 died within one year of the onset of the diabetes, though death was unconnected with the abortions which I understand were performed. Case Xo. S12, a physician's wife, first seen December 29, 1914. The examination of the urine showed no albumin, sugar 0.4 per cent., no diacetic acid. Xo diabetic- heredity existed. The patient had had measles, whooping-cough, and a rather peculiar case of typhoid. Married at twenty-one; one year later a stillborn child. In .June, 191 \, she became pregnant, and in Dec-ember, 1914, during the sixtltmonth of pregnancy, sugar was found in the urine. It was impossible to determine when it first appeared. The patient felt well except for slight dyspnea; polydipsia and moderate polyuria existed, but the appetite was decreased rather than increased; the highest weight was 179, and the present weight 1()9 pounds; normal reflexes; heart normal; pulse, 100; blood- pressure, l.'H); considerable edema of the lower extremities, espe- cially the right leg, and an abdomen consistent with a pregnancy of six months. The case appeared to me most favorable to attempt to carry through to term. The patient had previously been fat. The polyuria was not great, although the percentage of sugar was high. During the week intervening between the first visit and the entrance of the patient to the hospital no alteration in the diet was made (Table 19(1, p. 45o), and from it will be seen that the percentage of sugar underwent no change. Following entrance the carbohydrate in the diet was gradually decreased from approximately loO grams to 45 grams, and the patient was allowed a very moderate quantity of protein and fat- Sugar then disappeared, but the diacetic acid rose to three plus and the ammonia, which at entrance was !..'> grams in twenty-four hours, rose 1 to M grams, but with S grams of sodium bicarbonate (I would not use it today), a slight increase in the carbohydrate and decrease in fat, the ammonia fell to normal, where- it remained until just after confinement. The tolerance for carbohydrate rose to 100 grams. The protein was kept in the neighborhood of from 75 to 100 grams a little- over 1 gram per kilogram body weight; the fat was only such as enabled the- patient, to retain her weight. During the last three months of pregnancy the patient was in as good health as any normal pregnant, woman. The blood sugar on February !>, 1915, after dinner, amounted to 0.25 per cent., and the urine showed a trace of sugar and diacetic acid. The carbohydrate ingested was 110 grams. On February IS, 1915, fasting, but with the same quantity of carbohydrate in PREGXAXCY A XI) DIABETES 451 the diet the blood sugar was ().!(> per cent. The urine showed a trace of sugar. The method of delivery was carefully considered by Dr. -I. ('. Hubbard, who had charge of the obstetric side of the case, and it was decided that a Cesarean section presented the least danger to mother and child. This was performed March 24, 1915. Light etherization was employed for the operation. It had been previously agreed that if acidosis developed prior to the ter- mination of labor, local anesthesia should be used. Had gas and oxygen or local anesthesia been employed the acidosis which followed confinement would quite likely have been avoided. (See Case No. 1070, p. 457.) The blood-pressure, which was 130 on December 29, 1914, had remained essentially the same, and was 125 on the day before delivery. The urine for the twenty-four hours preceding the operation was as follows: 1710 c.c.; specific gravity 1012; acid; slightest possible trace of albumin; nitrogen 11.3 grams; ammonia 0.9 grams; no sugar. The diet for the same day contained about 120 grams of carbohydrate. During the twenty-four hours following delivery a trace of sugar appeared in the urine, but the ammonia remained at 0.8 gram and the nitrogen at 8.8 grams. Diacetic acid was present, and the sediment showed numerous brown granular and a few hyalin casts. Upon the second day there was 0.2 per cent, of sugar by the Benedict test, but the urine was 0.0 per cent, levorota- tory on account of the acidosis, which was proven by the presence of 2.2 grams of ammonia and a marked diacetic acid reaction. The sediment showed an occasional hyalin cast. Upon the third day the urine only showed a trace of sugar, but was 0.8 per cent, levorota- tory, the ammonia was 1.8 grams, and the nitrogen 8.7 grams. Upon the fourth day it was essentially the same; there were no casts. Thereafter the sugar disappeared, and up to the present time, April, 1917, has never returned in a quantitative amount. Upon the day of delivery and the two following days the patient was practically fasted on account of the return of sugar and diacetic acid. The diet was then gradually increased, beginning first with a small amount of carbohydrate and protein, and the fat was limited to approximately 50 grams. Within a week the diet had been brought back nearly to what, it was prior to delivery, and from this time on convalescence was uneventful, and finally the patient was discharged upon a diet containing 82 grams carbohydrate. The baby was normal in every way, and the weight at birth was eight pounds, reached thirteen pounds on July 30, 1915, and the child has remained well since. It is frequently maintained that diabetic- patients cannot nurse their babies. Nursing was easily maintained 452 T RE ATM EXT in this case until into June, when the milk rather abruptly ceased. Mother and child are in good condition April 3, 1017. TAHLE 190. CASE No. 812. PREGNANCY AND DIABETES. CE.SAREAN SECTION. 29.191 1 915 o 5- (1 (1- 7 7- S 8- 9 9-10 10-11 11-12 12-13 13-14 14-15 1(5-17 17-18 20-21 22- 23 24-25 25 27 27-29 29 39 30-31 31- 1 1- 2 2- 3 5- (i Mar. Diaeetic acid. Tra Urine. Total o.4';; (i.-i- ; 31 40 20 Tr. (I Carbo- hydrate Carbo- hydrate balance, Blood sunar. Weight. Mother, 1 ( >!i~i-rvc> decrease of ferric chloride reaction with elimination of sodium bicarbonate, -Delivery by (Vsarean section. .1. ('. Ilubbanl. Anesthetist, Freeman Allen. Anesthetic, et her. C )lerve appearance and persistence of acidosis. alter operation. In a subsequent operation upon ('a.-c No. 1(170, p 157, novocaiti was employed. Albumin at no t ime exceeded the slightest pa-.-ible trace, and casts were never present in the sediment except just after delivery. PREGNANCY AND DIABETES 453 Thirteen cases of severe diabetes coexistent with pregnancy have been seen, but nine of these occurred in former years when treat- ment was less satisfactory. Yet under these conditions occasionally a patient came through confinement successfully. An unfortunate complication has often been the appearance of albumin and casts shortly before delivery. This took place in Case No. 307, seen in consultation with Dr. J. G. W. Knowlton in 1910. This patient survived confinement despite the appearance of marked anasarca and 0.3 per cent, of albumin, but the baby was lost. If such a result could be attained six years ago, it is obvious that with the improvement in treatment of diabetes and nephritis, still better results might now be obtained. It is possible that an early Cesarean section might have saved the child. A patient may go through the pregnancy without notable symp- toms and be delivered of a healthy child. Such a case was Case No. IOC), seen in consultation with Dr. F. W. Taylor. 1 Multiple pregnancies may take place in the presence of consider- able quantities of sugar, and yet recovery be quite satisfactory. This happened with Case No. 854, who was seen by me in the third month of her third pregnancy, and following a successful confine- ment she reported on November 17, 1915: " I have got a big baby boy, born October 11, 1915. I am strong and well." The unfavorable course of most pregnant women has undoubt- edly been due to neglect of proper treatment. This took place in Case No. 60S, who, at the age of twenty-eight years, despite her knowledge of the existing diabetes, married and promptly became pregnant. As her father also had diabetes, there was a prospect of the disease running a mild course, and this was favored by her being placed upon dietetic treatment by her family physician, who had had considerable experience with diabetes. However, she gave up treatment and put herself under the care of a clairvoyant, and I saw her for the first time at the beginning of coma. A similar instance is that of Case No. (104, also an hereditary case of diabetes in a fat woman. When first seen, upon a moderate diet, the urine was free from sugar. However, in the absence of her family physician, she neglected treatment, and when the case next came to my attention, during the sixth month of pregnancy, 3.X per cent, of sugar was present in the urine and a slight trace of diacetic acid, and soon albumin appeared. In the following month, on account of urgent symptoms, the pregnancy was ter- minated, and the patient died a few hours later in coma, probably uremic. Ether was used. 2 1 Taylor: Boston Mod. and Surtr. Jour., 1S99, rxl, p. 205. 2 Today I should consider ether contra-indicated. 454 TREATMENT The advisability of an abortion is always raised when sugar in considerable quantity is discovered in pregnant women. In one of the milder eases of my group this was performed, not so much on account of the sugar, but on account of albumin and mental symptoms. The same procedure was carried out in one of the severe cases, that of Case No. (171, aged twenty-five years, first seen November 25, 1913. The patient was a Jewish woman, who probably developed diabetes in April, 1912, although unrecognized by herself. She was married in June, 1912, and the disease was discovered in August, 1912. At the first visit the urine contained 7.2 per cent, of sugar. She was advised not to become pregnant, but pregnancy did take place in January, 1914. The urine did not become sugar-free, and an abortion was successfully performed by Dr. L. V. Friedman in March, 1914. J later learned that she again became pregnant and in October, 1914, committed suicide. A review of the 23 cases of diabetes associated with pregnancy coming under my observation gives hope that in the future such patients will take a far more favorable course than has been generally believed. It is true that nine of the women with severe diabetes have died, but in each instance some harmful cause was present which today would be susceptible of treatment. The favorable outcome of Cases Xos. XI 2, 430 (p. 454) and 1070 (p. 457) encourages one to believe that even if the patient had a somewhat severer type of diabetes, it could be successfully overcome. Several expedients not required in the treatment of Case Xo. X12 could be employed, notably: (1) fasting; (2) the selection of gas and oxygen in prefer- ence to ether (of course chloroform would not be employed) or even local anesthesia; (3) the early performance of a Cesarean section at the first sign of unfavorable renal symptoms. 2. Conclusions upon Pregnancy and Diabetes. From a study of these cases it would seem that the secret of success in the treatment of pregnant women with sugar in the urine could be summarized as follows: 1 . The patients should remain under constant supervision through- out the course of the pregnancy and for months and years after con- finement, because it. is not uncommon for the sugar to return. 2. Treatment should follow exactly the same methods which are employed in the treatment of the usual case of diabetes, although especial pains should be taken to prevent the appearance of acidosis by restriction of fat in the diet. 3. Fven when sugar appears to a slight extent in pregnant women, it should be carefully watched and controlled by diet. 4. The advantages of a Cesarean section should be borne in mind. 5. Ether anesthesia is not as safe as gas and oxygen. If ether PKEGNAXCY A XI) DIABETES 455 should be used, as brief an anesthesia and as little ether as possible should be employed. Local anesthesia has proven satisfactory, G. Many statements occurring in the literature of pregnancy and diabetes must be revised. Pregnancy in diabetes does not demand immediate abortion, even if acidosis is present. (See Case. Xo. 1070.) If pregnant diabetic cases are suitably managed, they will very likely abort less frequently. It cannot yet be accepted as proven that pregnancy aggravates a diabetes. It is quite possible that the reason for a patient with diabetes becoming worse during pregnancy is simply due to the ingestion of an unusual quantity of food. Reliable data upon the carbohydrate tolerance of diabetic patients before and after pregnancy are not available, although Maase 1 records a single such instance. Nursing is not contra-indi- cated following the confinement; the diversion which it affords the patient may offset the extra demands thrown upon the metabolism. Too few data have been accumulated regarding the blood sugar of pregnant women to warrant conclusions. The same statement may be made about the alleged excessive weight of children of diabetic mothers. The next few years may show that pregnancy may take place in diabetic patients far more readily than has been sup- posed. It is certainly true that with the improvement in the treat- ment of diabetic patients, diabetic women will be less likely to avoid pregnancy. In fact, the above prediction has come true, but before reporting a recent case I would call attention totlie experiments of Carlson and Ginsberg, 2 which are interesting rather than encourag- ing. Whereas total extirpation of the pancreas in non-pregnant dogs results promptly in the onset of pancreatic diabetes, complete pancreatectomy in pregnant bitches near term is not followed by hyperglycemia and glycosuria as long as the fetuses are alive and the placental connections are not severed. At the onset of labor blood sugar begins to rise and characteristic pancreatic diabetes is established on the completion of the delivery. 3. Notable Increase in Tolerance for Carbohydrate during Preg- nancy in a Case of Diabetes of Seven Years' Duration. Case Xo. 43(> first came under my observation August 1, 1911, with onset of diabetes at the age of twenty-two, in May, 1909, when S per cent, of sugar was found. The family history is remarkable in that the mother died of diabetes at the age of fifty-five, having probably had the disease ten years, and five uncles and aunts, of whom one is Case Xo. 759, p. 489, have had the disease. The past history was nega- tive except that the onset of the disease took place coincidentally with the death of the patient's parents. The patient was overweight, and at the first visit weighed 152 pounds, which was pounds less 1 Masse: Char. Ann., 1910, xxxv, p. :. 2 Carlson and Ginsberg: Am. Jour. Physiol., 1914, xxxvi, p. 217. TREAT ME XT than thr highest weight. Physical examination was negative. The course of the disease in the patient is shown by the annexed chart. (Table 1<)7.) T.VHU-: 107. CASE Xo. 4:>C>. ILLTSTHATIOX OF RISK ix TOLKKAXCK FOR CAKHOIIYDKATK IHKIM; PKKCXAXCV. Date. Ainmo- Vol., Diacetic nia. c.c. acid. total grains. I'rine Carho- Nuked ,., MWir, Carho- hydrate weight of l!l "'"' total hvdrute. balance, patient, Kraiiw. srams. pounds! " ( ' r <><1 ' lt - 1909 May s r ; 1911 July 31 1250 45 152! 1912 Jan. 9 1020 + 22 151! June 11 900 + 2 1913 Mar. 22 MO + 2 HO; Dec. 23 1200 + + 21 191 4 Jan. 9 1050 + + 11 23 1140 + + + + Mar. 2s 750 + + Sept. is 90i) + + 11 20 + 9 Dec. 5 1200 + + + 2.52 1915 April 12 900 ++ 1.8 15 +15 Prctr nancy li ran Au trust, 191 5. Autr. 27 030 + Sept. 20 + + Oct. 13 1230 ++ 2.3 133! Nov. 4 900 + + + 1.0 +10 133! 21 900 -r + + 2.2 45 +45 0. m Dec. 15 900 75 +75 1910 Jan. (i 1110 Slitrht -r 1.5 00 +00 12 992 1 . 1 2 0.20 April 4 1440 + 1.9 Trace 55 +55 152! 0.30 Successfully online,! of 9! Il>. hoy. April 25 1200 ++ 1.3 May 25 1050 + 2.0 50 +50 June 22 1530 50 +50 1917 Jan. 12 900 + SI. + Mar. 2s 7MI ++ 1.4 20 2O 30 It will be seen from this chart that su^ar was constantly present in the urine from May, 1! )()'.>, until -January '2'.\, 101 1, when a pro- longed, energetic attempt was made to remove it. However, it quickly returned; it again disappeared, although with the presence of considerable acidosis, on December ;">, 1 ( .)14, this time with the help of fasting treatment. Thereafter it remained absent, and the, patient acquired a tolerance for 15 grams of carbohydrate PREGXAXCY AXD DIABETES 457 and lived strictly within this limit, even though considerable acidosis existed. The patient became pregnant, the last catamenia being August 4, 1915, and it was observed that the tolerance for carbohydrate gradually increased. From the chart it will be seen that it finally rose to at least GO to 75 grams. During the pregnancy the diet was halved on one day each week. Confinement took place unevent- fully on April 30, 1910, and the patient was delivered of a 9^-pound boy. May, 1917, finds both mother and child doing well. The blood sugar on January 13, 191(>, was 0.2 per cent., and on January 27, 1910, 0.19 per cent., when the blood-pressure was 122 systolic and 7 y. o - z y ^ E < 1916 June 22 3100 + + 13.3 1.9 93 101 24 500 121 (I 23 2500 + -j- 05 S9 22 444 122 21 20 24 2900 -f 41 1 89 22 444 . 21 25 1500 SI. .. 1.0 12 05 17 . . 328 124 20 2000 i SI. . . 1.7 4 50 10 240 29 27 2000 ! SI. 30 5 140 125 31 2S 1SOO 10 5 GO 120 0.13 28 July 14 2300 04 70 105 1481 122 . . 0.13 Aug. 31 , . . . : I Delive ry by Ces : area|nsec' :ion lo cal anes the si ! a. Sept, 1 700 22 30 382 . ; 1 2 900 + . . 51 19 2s 532 . . . . . 28 3 800 + -f- . . 1 47 34 37 057 . . . 29 4 000 + 47 34 37 057 . 32 5 800 . . 70 44 i 50 960 . 33 25 2570 Mi s2 180 2280 1 J9 1917 Feb. 2S 800 _l 1 I 22 Mar. 1 , 700 + + + + . . i 12 , . . . 0.33 29 2 1000 + ; 0^0^ . 3 i 1900 000 . o 19 1400 .. .. 1 18 52 101 1189 . . . 0.24 In Table 198 is shown the course of Case Xo. 1070 who in the sixth month of pregnancy entered the New iMigland Deaconess Hospital. She had severe acidosis, for the COo tension in the alve- olar air was 20 mm. Ilg and the CO 2 in the blood 21 in terms of mm. tension Hg. Under the care of my assistant, Dr. Ilornor, she 4f)X TREATMENT became acid- and sugar-free, went home and returned at the begin- ning of labor. A Cesarean section was successfully performed by J. ('. Ilubbard under local anesthesia. The mother and child remain in good condition at this writing, May, 1917, but the tolerance of the mother for carbohydrate has decreased to a few grains just as it has to a lesser degree in Case No. 430. O. DIABETES IN CHILDREN. It is unfair to base a prognosis upon the course of diabetes in a child by the unfavorable records of the past. There are but few cures of diabetes, but the few which are on record have most often occurred in children, and there is always the possibility that this is the exceptional case which may recover. While the parents are told the usual outcome of such cases, they should be given hope until the outcome of the disease in a given child is plainly evident. Whereas formerly the prognosis for children under ten years of age was measured in months, today it is rare for a child to live less than one year. Among my cases there have been .">! who developed the disease within the first decade, and of these 30 TREATMENT The diet of children with diabetes from the age of two upward requires surprisingly few modifications of that of an adult. Fasting is borne perfectly well for a few days, and with comparatively little complaint. Even the complication of a slight temperature and a croupy cough (not diphtheria) does not interfere with the fasting treatment. Such symptoms were present during the first few days of hospital treatment of Case No. 9oX, whose dietetic and urinary charts arc recorded on page 4S5. Following the period of fasting the addition of 50 grams of 5 per cent, vegetables with broths goes a long way toward relieving hunger, and subsequent weekly strict diet, days cause no trouble if broths and 300 grams of 5 per cent, vegetables can be allowed. Especial pains should be taken in the preparation of the vegetables, for children eat fast and occasionally diarrhea results. Strange to say, more frequently the complaint is of constipation. Children's specialists, accustomed to dealing with children of delicate digestions, sometimes err in making the diet too simple. In my experience with diabetic children two or three years of age and over, they easily bear the diet of adults if reasonable care is taken. The quantity of protein required by the child is proportionately more than that of an adult. So far as I can learn from the published data, .') grams of protein per kilogram body weight is the amount which the normal child of about four years requires. As age advances this gradually decreases to 1 gram or 1.5 grams for the adult. The following table shows the nitrogenous urinary excretion of four healthy children selected for their health and size from among the playmates of some of my diabetic children. TAHLE 200. THE XITKOCJEN AND SALT IN THE TUIXES OF FOUR HEALTHY CHILDREN. Nf. Weight. VJ. Sp . gr . NaCl, 1910 Oct. IN 19 ; <". I'. ' 2 yrs. 10 inns. ' Us.} :5I5 11,. 19 20 20 21 11 :54 11>. M oz. 20 21 A. ('. 4 vrs. 2 inns 1016 21 2."> I). J., Jr. yrs. mos. 10^ -11 11>. l.'i <>z. s70 KIM son 101:5 DIABETES IN CHILDREN 461 I am quite open-minded upon the question of the proper quantity of protein for the child, but at present I try to give 3 grams per kilogram body weight. Possibly one should give much less. The quantity of protein in broth must be considered in planning the diet of children. One specimen of broth showed 36 grams to the liter. The quantity of carbohydrate is determined by the same methods as in adults. Tolerance tests are demoralizing, because the child cannot reason out why so much food of the character he enjoys should be good for him one day and not the next. It is better policy to keep the carbohydrate low rather than near the toleration limit. The Eskimo children have little carbohydrate and thrive, and so ought diabetic children with practice. New tests in treatment should be worked out upon adults; a colt needs a steady rein. In giving carbohydrate to a child, do not forget that the addition of 5 grams carbohydrate to the diet of a child weighing 15 kilograms is proportionately equal to the addition of 20 grams to that of an adult. Fat has been the stumbling-stone in the pathway of the treatment of diabetic children in the past. Children require many calories and seldom can get along on less than 50 per kilogram body weight, but they are very susceptible to the development of acidosis and when the acid danger lurks, beware! The addition of an egg con- taining only 6 grams of fat brought out a marked acidosis in Case No. 938 (see p. 485) on October 28. A few days later, when the child had become accustomed to fat, 32 grams were taken with impunity. No longer is it true that all of my diabetic children have died of acidosis, for one succumbed to inanition; but with 37 out of 38 deaths due to one cause there is no question as to what the enemy is. And as acidosis is due to fat not being oxidized, we must strive in every way to favor its oxidation by increasing the tolerance for carbohydrate. The restricted diet of children may possibly conceal dangers due to the absence of salts of one kind or another. To avoid such a contingency, raw vegetables are prescribed freely and I always like to give a little cream for the sake of the calcium. Even if che tolerance is above 50 grams, I believe it wiser to keep the diet at or below that figure. The prognosis in children is so grave that no chances should be taken. Furthermore, it is a great disappointment to all to cut down the allowance of carbohy- drate, and therefore one should try to avoid it. It is better not to give children saccharin. Strict dieting must be carried out for such a long period that it is kinder to the child and easier for him to give up the taste of sweets. For this reason, various diabetic breads are not to be encouraged. Meat, eggs, vegetables, butter, 402 TREATMENT cream, nuts, cheese, and possibly fruits of the 5 and 10 per cent, groups should form the permanent diet. Temporarily, I often employ sugar-free milk, but eventually, as tolerance grows, sub- stitute cream for it. Exercise has also been beneficial even when it has been violent. Case No. 7S5, who took no exercise and made no exertion, improved very slowly as compared with Cases Nos. 92") and !)2)>, who played tennis and walked miles. All the children appear to do better when taking exercise. I shall take pains in the future to provide massage for those confined to bed. \Ye must never forget that one-half the glycogen in the body is stored in the muscles, and that undoubtedly fully as much of the carbohydrate ingested is there burned. In addition to the following illustrative cases, other cases in children are described on pages 4S4 and 485. TABLE 201. CHART OF A CHILD, CASE No. 1209, AGED 8 YRS., 6 Mos., SHOW- IXC; H()\V HE CoNQUEKKD IllS TENDENCY TO EAT CANDY. Diet in grains Urine. Svifjar. Date, 'Z 1917. a M '' PCI ct. To- tal "3 rt K111S. .Ian. 1 Spec. '.).<) 1- 2 300 4 . ( 12 2- 3 .')()() . I 3 3- 4 (500 Tr. 0.1 1 4 .') 1000 31 31' 5- (i 325 o o.: 1 (i- 7 500 : o 7- X 550 (t.l 11 X- 9 300 (I 9-10 150 10-11 775 1.7 13' 11-12 <;oo (1 12- 13 ,soo lii-17 X( )( ) 17- is soo 0. (1 IS- 19 1 10(1 1) 22 27- 1240 1) 27 -2.S 1300 o o 1) Case No. 9)>X illustrates the preliminary treatment of a child two years old, and is given in detail m Section \ 1 (see p. -IS")). Cases Nos. ( .>2:> (see p. -1X4) and ( .>2."> (see below) illustrate fasting treatment. To the latter a fe\v grains of carbohydrate were allowed instead of complete fasting, but with the former this was not indi- cated because he became sugar-free so quickly. rpon the first three days at the hospital the urine of Case No. !)2.") contained respectively o.(> per cent., ().(> per cent sugar, and 1 Ate candy. - Average for 6 days. DIABETES IN CHILDREN 4f>3 then became levorotatory. On the fourth day it was sugar-free. The diet for the first day consisted of one orange and 50 grams 5 per cent, vegetables; for the second day one orange and 150 grams 5 per cent, vegetables; the third day was fasting except for broth; the fourth day to the broth were added 30 grams 5 per cent, vege- tables. After that the diet was increased by approximately 5 grams carbohydrate a day up to 39 grams. TABLE 202. CHART OF A CHILD AGED 2 Yns., 6 Mos. CASE No. 1199. Urine. Diet in grams. Dietary prescriptions in grams. Sugar. = ft d Date, 3 2 "5 E E hi 5 C. 1917. T3 2 6 o 3 Per To- a - cc z [3 o u bt 3 E o tal - -^ c s. ^ s- CH r* 0) ~ ct. ^ c M 8,3 J5 .% D3 w: be + c ^ ^j Q 2* _,nis. _ 2 a _?; c3 P o t. < c- S-* C O o ^ CO w u P3 Dec. 4 Spec. 6.3 15 5 780 5 . 3 41 8 13 Spec. 2.4 1.2 35 32 Carbohydr En tra nee ate to only j part hospital. lye lim inat ed. 15 330 1 16 510 Tr. 15 11 108 300 0.5 30 17 480 Tr. 25 28 3 229 32 300115 0.5 !K) 00 30 18 19 20 630 540 290 30 in 41 42 43 17 24 21 437 544 572 si 300 , 15 300 : 30 300 30 1.0 1.0 1.0 90 60 60 60 60 60 2 2 2 30 21 420 53 45 30 662 300 30 1.0 00 60 2 60 30 22 480 000 60 47 40 788 300 : 30 1.0 60 00 2 90 60 Case Xo. 814 shows the course for over two years. TABLE 203. -CASE No. 814. AGE AT OXSET, TWELVE YEARS. Date. Volume, c.c. Diacetic acid. Ammo- nia, grams. Total sugar, grams. Diet in Grams NaliCOs Blood sugar, per pent. Garb. Protein. F; lt grams. 1915 Jan. 3 11,500 920 4 1 4,000 6 . 1,200 + + + + 1.8 12 + 20 7 2,882 + 24 19 1,530 50 GO 90 Mar. m 0.17 May 21 0.15 Nov. 10 0.6 0.13 IS 35 1917 Jan. lo 52 62 OS 0.18 Mav L'G __"_ 03 Gl 80 Case Xo. 295 is reported in detail in Publication 13(> of the Car- negie Institution of Washington, I). C., 1910, page 120; and also 1 Fat -protein diet. 4b'4 TREATMENT Publication 17(>, 1912, page 21. The case is inserted here because of the exceptionally long duration, which may \w explained by the presence of a diabetic heredity. Case No. 295; male; born February 2, lSS(i; married; chauf- feur; came under observation October 25, 1909; onset of diabetes at the age of fourteen years in 1900; death in coma in May, 1912. Fain ih/ History. Yather died of diabetes, complicated by appendicitis, at age of fifty-one years, having been ill with the disease for two or three years. One brother died of diabetes at the age of eleven years. The mother has gall-stones. Two brothers and one sister well. The patient was married September, 1907, and has one child in fairly good health, five months old. Paxt History. A frail boy; scarlet fever, measles, chicken-pox, whooping-cough. October, 1907, typhoid fever. I'rctit'iit Ill/u'M. The date of onset is not accurately known, but during 1X99 the patient was thirsty and arose several times a night to void urine. In May, 1902, sugar was found in the urine. Dr. Oscar (). Roberts, who examined the patient at that date, said that the indications were that sugar had been present for some time previous. Sugar was absent from the urine for a brief period in 1901. Since the onset of the disease the patient has worked steadily, only occasionally losing some time, as during an attack of typhoid fever, and on account of a furuncle on the heel. lie began his occupation as chauffeur in April, 1909. The greatest quantity of urine noted in twenty-four hours was 12 quarts, and upon October 2o, 1909, the quantity measured amounted to 10 quarts. The patient has a mania for ice-cream sodas. lp to October 1 the patient felt, in fairly good health, but was then especially upset because he could not get the diet to which he had been accustomed, being forced to live, while upon a visit, chiefly upon sugar and starch. lie now "eats any time." No headache. He is recovering from a cough, but has no sputum. Phij^'n-dl Examination.- The greatest weight, dressed, was (>O.X kilograms; on October 25, 1909, it, was 19.5 kilograms. Height, 17C> cm. Typical gaunt, flushed, dry appearance. Reflexes normal; eyelids red; teeth in good condition; tip of tongue red and back of tongue slightly dry; lungs and heart normal; pulse, 12S, blood- pressure, 100. "Acetone" odor to breath, and whole room filled with it- An attack of influenza in the spring of 1910 reduced the strength of the patient materially, yet he recovered sufficiently to go into the poultry business. He did not limit his diet, and it included, among other articles, 2 quarts of milk, (> oranges, and 10 eggs daily. In I'Ybruary, 1911, he returned to the hospital on account of diar- rhea, which had existed for some time. I uder hospital care the DIABETES IN CHILDREN 4G5 number of stools decreased from 12 to 5 daily, and his condition improved. During 1911 the patient's condition did not change. He agreeably passed the early winter in Florida, but in February, 1912, as the diarrhea had returned, again entered the hospital for a few days. Except for emaciation and weakness and the presence of numerous furuncles, his condition had changed but little. Death took place in coma several weeks after his return home in May, 1912. Case No. 304; female; age at onset thirteen years; came under observation at the age of twenty-one years on November 24, 1909, and died in coma on January 4, 1910, following restricted diet and unusual exertion. Family History. Several of the family died of diabetes, the grandmother at thirty-five, the mother at forty-one, and a brother at twenty-five; another brother has had the disease for six years. One brother well, one brother died of pneumonia at twenty-eight, a sister of diphtheria at four. Past History. The only illness observed was measles. Cata- menia began at the age of thirteen and continued until the end. Symptoms of diabetes also appeared when the patient was thirteen. When first seen she appeared in far better condition than one would expect. She was nervous, the eyes were prominent and her hair had been falling out, but the thyroid was normal. The pulse was 92, blood-pressure 130; there were scars of tubercular glands in the neck, but the physical examination otherwise was negative. The diet was restricted, and at first the patient did well, but during the Christmas vacation she grew very tired one night at a dance, returned to Boston for school the next day, and became exhausted with the weight of a heavy bag which she carried a long distance and in a few hours was in coma. The data regarding four children under observation are recorded in Table 204. TABLE 204. DIETARY CHARTS OF FOUR CHILDREN- PRESENT OR REPRESENTED AT THE TUESDAY MORNING CLINIC AT THE NEW ENGLAND DEACONESS HOSPITAL, FEBRUARY 29, 1916. Age. Diet. ,,, . i , C.'ilor- Duration, April 1, V\ eifzht, - Case No. At onset, Feb., 1010, Carbo- hydrate, Protein, Fat, Calor- kllos - kilo 1917, yrs. yrs. yrs. Kms. S94 1.3 3.6 30 30 54 702 10 50 3.4 997 . 7.7 S.O 33 59 75 1043 20 52 15 995 . . . 0.0 0.2 44 45 84 1112 19 59 1.3 950 . . . 2.1 2.7 24 41 70 944 12 79 1.8 The chart is recorded here because by chance the children or the mothers of three were present at the clinic February 29, 1916, and 30 Kit) TREATMENT a letter was received the previous day from Case No. SOI. The cases therefore reveal actual conditions. Inspection of the chart shows that Case No. 050 is receiving the highest number of calories undoubtedly too high, I should say, writing in the spring of 1917. A year ago she did so well I hesitated to make a change. Later she showed sugar, but under observation in the hospital became sugar-free. In the autumn the tolerance became almost nil, and I asked Dr. Allen and Dr. Fit/, at the Rockefeller Institute, to take her as a patient. She improved at the Hospital of the Rockefeller Institute and was discharged with a tolerance for 7 grams carbohydrate. This case has proven to be the most severe of the four children and it is noticeable that a year ago the diet of this child was very much higher in calories than that of any of the others even when the ages are taken into account. Case Xo. S04 is holding her weight, but should have more calories; probably 55 to 00 per kilo body weight per twenty-four hours would be borne well. Xo change need be made in the calories which Case Xo. 005 receives, for his case is a recent one and his weight is practically constant. Case Xo. 007 is also a recent one and I am not inclined to raise the calories at present. P. DIABETES IN OLD AGE. Diabetes frequently occurs in old age. The onset in 25, or 2 per cent., of my cases, took place above the age of seventy years. It is often claimed that such patients need no treatment, and are, in fact, better off without it. To this statement I cannot agree until figures supporting such a position are furnished. It is true that four of the twenty-five patients have outlived the normal expecta- tion of life of individuals of the same age at which they developed diabetes. This point of view of prognosis in diabetes is interesting and I shall follow the lead farther. (See Table 205.) The treatment of elderly patients must be undertaken with greal caution. I hesitate to interfere with the daily habits of any indi- vidual who has succeeded in living over three score years and ten. Hut if upon thorough study of the case it appears that the patient is losing ground or is hampered by diabetic symptoms or com- plications, there should be no hesitation in active treatment. Treatment of elderly patients should be begun most gradually, there must be no haste; supervision should be close and all sur- roundings made favorable. Hearing in mind the vulnerable kidneys of elderly patients, acidosis should be avoided under all circum- stances. Sodium bicarbonate should not be employed, because ol the marked variation in the water content of the body which follows its use. Dietetic changes should be so slowlv introduced DIABETES IN OLD AGE 407 as to disturb the equilibrium as little as possible. Do not con- fine the patient to the bed. All of these regulations are met by a preliminary withdrawal of fat from the diet. If the urine does not become sugar-free in consequence, so soon as the total quantity of sugar in the urine does not continue to show a steady daily dimin- ution, the protein and carbohydrate should be halved. Should this procedure fail, the protein is omitted and the carbohydrate reduced to 40 grams and halved the next day, and on the following day entirely omitted. Having determined the carbohydrate toler- ance, protein may be quickly added up to a gram per kilo body weight, and fat gradually given until the weight ceases to fall. TABLE 205. THE COURSE OF DIABKTES WITH ONSET ABOVE THE AGE OF SEVENTY YEARS. Case Xo. LIVING CASES. A 20 ;it onset. 629 71 687 75 817 72 826 73 N97 79 899 83 1057 71 1063 71 1093 77 Normal expectation of life. Duration to Dec. 1, 1916, years. 5.5 6 8 .0 6 5 .1 ,3 5 5 .0 8 1 ( .3 3 5 .6 3 4 .1 4 6 4.7 2 2 4.0 1 S 6.7 2.5 0.5 FATAL CASES. Case No. 90 . 118 . 227 . 234 . 252 . 267 . 2ss 393 . 410 . 4ls . 495 . 597 . 628 . AP at onset. 4.7 5 . 9 7.1 7.1 7.6 8.0 7.6 5.1 7.6 per cent., and it is more accurate to reckon it empirically as .'> per cent., or 1 gram to the ounce. Thus, !.">() grains (o ounces) of 5 per cent, vegetables contain about 5 grams carbohydrate. (6) Protein and Fat. Protein and fat. are the other two con- stituents of the diet. Protein, examples of which are white of 472 AIDS IX MAXAUKMKXT OF DIAIiKTIC CASES egg, Iran of meat or fish, aue formed in the body in severe cases. Fat does not directly lead to the formation of sugar, but if taken in excess may indirectly cause sugar to appear in the urine. I'ntil very recently diabetic patients were given too much fat. This is explained by the fact, that only within the last few years has it been possible to determine whether the fat taken by diabetic patients was assimilated. Xew methods now show that it. promptly accumulates in the blood, as is shown by the frontispiece, which is a picture of the blood of Case No. 7N(i. The approximate content of carbohydrate, protein and fat in the foods commonly eaten is given on the diet card, as shown in Table 211,]). 4X2. (c) Weights and Measures Employed in Computing the Diet. The best diet for a patient is most readily determined by testing the effects of weighed quantities of various foods. The patient should learn how to do this for himself and should find out how much carbohydrate, protein and fat lie is taking, and whether he can keep sugar-free upon it. A\ithin one week from the beginning of treatment he should be able to estimate roughly how much food he is eating without the frequent use of scales. J$y learning at once to judge the weights of food portions, much annoyance is obviated and the patient ceases to be conspicuous at table. In weighing foods and calculating the diet it is desirable to use the metric system, for in the end it will be found simpler. TAIH.K 207. WKICHTS AND MKASTHKS KMPLOYF.I) i\ ( 'OMITTIXC Tin: DlAUKTIC DlKT. = )}() grains (actually 2S.4 grams). = '.]() cubic centimeters (actually 2!U> cubic centimeters i or 2 tablespoonfuls. = 1 pound, or 4.~>l grams. = 1 quart, or 9K> cubic centimeters. = I kilogram, or 2.2 pounds In estimating carbohydrate, protein and fat in the diet, or sugar in the urine for clinical work, unless otherwise directed by the physician, enough accuracy is obtained by considering oO grams or .">() cubic cent i meters as equivalent to one ounce dry or liquid measure. The quantity of protein, fat, and carbohydrate is recorded in Table I'll for each i!l) grams (one ounce) of the articles of food usually eaten. Thus .'!() grams of oatmeal (weight before it is cooked) contain 20 grams carbohydrate, 5 grains protein and 2 grams fat. WHAT EVERY DIABETIC PATIENT SHOULD KNOW 473 (//) The Quantity of Food Required. The amount of food neces- sary for a patient is determined by his weight. The weight of a diabetie patient should always be less than his former greatest weight, Ixx-ause thus he can be assured that he is not overeating. In this resjx'et it is better to emulate the Indian than the Eskimo. The individual 10 per cent, and even 20 per cent, below weight may not l>e a delight to our eyes, but if over thirty-five years of age and in this condition he is much more acceptable to the insur- ance company. It is often desirable for a patient to lose weight, but this should bo undertaken only under the doctor's direction. Frequently it is only by losing weight that a patient regains the power to tolerate carbohydrate. (<>) Calories. The proper quantity of food can also be deter- mined by estimating the calories in the diet. A calorie represents the heat which is necessary to raise 1 kilogram (2.2 pounds) of water 1 C., or 4 pounds of water 1 F. For each kilogram body weight per twenty-four hours it has been found that an individual requires at rest 25 calories, and at light work 30 calories. The heat which is liberated in the body from the combustion of 1 gram of protein or carbohydrate produces 4 calories, and from 1 gram of fat 9 calories. From these figures it is easy to estimate approximately the calories in the diet, and to compare them with the number of calories needed. It is not possible to calculate the calories in the diet with exactness unless frequent analyses are made of the food eaten. (/) The Normal and Diabetic Diets Compared. Four-sevenths of the calories of the diet in health are made up of carbohydrate and two-sevenths of fat and one-seventh protein, but in diabetes the diet is composed almost exclusively of the latter two foods. This is not discouraging, for until recently the Eskimo's diet con- tained only about one-seventh carbohydrate. It takes time and exj>erience to learn to live successfully upon a diabetic diet and it is only with time that the body adjusts itself to a diet with so marked a reduction of carbohydrate and so marked an increase in fat. It is indeed wonderful that it is possible for the body to do so at all. ((/) Observe the Method of Making the Urine Sugar-free. The patient is expected to become sugar-free during the early part of treatment. He should observe and record how this is done, so that if sugar reappears in the urine he can follow the same method by himself and again become sugar-free. For this reason it is almost desirable for sugar to return before a patient leaves the hospital. If the return of sugar is promptly detected, fasting for twenty-four hours will almost invariably eliminate it. The diet may then be resumed, except that the quantity of carbo- hydrate should be reduced, and it may be necessary to reduce the 474 AID* J.\ MA.\A(;KMK\T OF DIABETIC CAXKS total quantity of all foods. Even if sugar does not reappear, on one day each week the diet should he made more strict than upon the other days. (/;) Distribution and Exchange of Carbohydrates. The carbo- hydrate, in the diet should be divided between the throe meals. Kven if the K) per cent., 1 "> per cent., and 20 per cent, vegetables are allowed, vegetables from the ."> per cent, group should be taken as well. I snally it is allowable to substitute for a given quantity of f> per cent, vegetables one-half as much from the 10 per cent, group, one-quarter as much from the !."> per cent., or one- sixth as much from the 20 per cent. Exchange vegetables for fruit only under advice. So-called diabetic foods often contain consider- able quantities of carbohydrate, and usually contain so much protein and fat that, they should not be taken by the patient without due allowance for the same. They should not be taken under any circumstances unless their composition is known. (/) Special Dietetic Rules and Hints. Eat too little rather than too much. With a return to normal weight sugar may appear. All food must be eaten slowly, and the coarser the food the more thoroughly it should be masticated. If in doubt about a food, let it alone until you have found out whether it is allowed. Do not yield to the temptation of friends to break the diet, for if this is done the plan of treatment is upset, a week's time may be lost and several pounds of weight sacrificed. Be especially careful to note the eil'ect of any increase in carbo- hydrate. The same rules hold for protein. The quantity of fat is generally regulated by the patient's weight. Remember it is always possible to get articles of food which arc included in a strict diabetic diet for a few meals, such as eggs, meat, butter, oil and even 5 per cent, vegetables, fresh, raw or canned. Quiet outdoor work agrees with diabetic patients. One of my cases who has done 1 exceptionally well has a diabetic garden and thus provides liberally for bis table both summer and winter. In ease of illness curtail the fat in the diet and if acid poisoning is shown by the ferric chloride reaction, omit fat entirely. 7. Care of Teeth. Clean the teeth after each meal and have them cleaned by a dentist, every one or two months. The teeth should always be kept in good condition. If they are to be extracted, take gas and oxygen, but no ether. Novocain injected cautiously acts admirably. S. Care of Skin. The skin must bo kept unusually clean. Take a tub bath daily, but avoid prolonged cold baths. Short cold baths are often desirable. Protect the skin from injuries. Caution chiropodists and manicurists not. to draw a drop of blood. If any infection occurs, see a physician at once. WHAT KVKRY DIABETIC PATIKXT XIIOt'LI) KX()\V 475 9. Treatment of Constipation. The bowels should move daily. Ise an enema if necessary. Never purge the bowels, but simple laxatives, such as one-fifth grain aloin, fluidextract cascara sagrada 10 to MO drops, or compound rhubarb pills, are allowable. Bran muffins made with agar agar 1 (see page 5M1) and coarse vegetables or fruit for breakfast may prove sufficient. If diarrhea occurs go to bed. keep warm, and drink hot water. ](). Exercise. Exercise freely for short periods six times a day. It is desirable to walk after meals and to arrange to get exercise for the upper part of the body as well as the lower. If the accus- tomed exercise is not taken, eat less. 1 1 . Best. Rest is essential. A tired child is put to bed and wakens refreshed; one of the most noted surgeons in our country is not ashamed to lie down for fifteen minutes after his luncheon; the best treatment for a failing heart in heart disease is to put its owner in bed for a week. Diabetic patients should rest often, should never get tired and should avoid athletic contests. The diet is designed to give a rest to the pancreas. Forget you have diabetes and do not talk about it before others. This is one reason for not using saccharine and the other is to avoid the perpetuation of a sweet taste. ^Year warm clothes instead of staying by the radiator or in an overheated room. ^Mental diversion is desirable, but anxiety is harmful. 12. Sleep. Sleep nine hours or more and get another hour of rest i>y day. Short periods of complete relaxation yield maximal returns. 1M. Examination of Urine. To collect the twenty-four-hour quantity of urine, discard that voided at 7.00 A.M., and then save in a cool place all urine passed thereafter up to and including that obtained at 7.00 A.M. the next morning. The urine should be tested daily before breakfast. Usually all that voided the preceding twenty-four hours should be saved and mixed. If this is impracticable, save as much as possible and test the mixed specimen. Enough liquids should be drunk to keep the total quantity of urine at 1200 to 1500 c.c. (10 to 50 ounces in the twenty-four hours). If the quantity of urine is larger and sugar and acid are absent, it may be that too much salt has been used. 14. Test for Sugar. Benedict's solution is used for testing the urine for sugar as follows: To about 5 c.c. (one large teaspoonful) of the solution add S drops of urine; the test may then be continued in either of the two following ways: (1) Boil the mixture of the solution and urine for three minutes and set aside to cool to the temperature of the room; or (2) Place the tube containing the mixture of the solution and urine 1 Sco-wccd make.- a snti.-fnrt<>rv substitute. 47(1 AIDS IX MANAGEMENT OF DIAHETIC CASES in bubbling, boiling water, where it must remain, with the water actually boiling, for five minutes. In either ease, if the solution remains clear, the urine being tested is sugar-free; if a heavy greenish precipitate forms it usually means that there is a trace of sugar; the appearance of a yellow sediment indicates the presence of a few tenths per cent, of sugar in the urine, and a red sediment more 1 . 1."). Visit to the Doctor. In consulting a physician take a one- half pint specimen from the twenty-four-hour quantity of urine which has been mixed and measured. Record the twenty-four- hour quantity on the bottle with your name. Enclose a list of all food eaten at the three meals during the preceding day. Arrange this list so that the names of all the different foods eaten are in a vertical row with the total quantities of each opposite. (See Tab'e 20S, p. 477.) A careless report of the diet is exasperating to a doctor. Keep a note-book in which to record all questions which may arise. Allow space for the answers. Gradually the note-book will become valuable for reference. B. DIRECTIONS FOR NURSES IN CHARGE OF DIABETIC PATIENTS. Diabetes offers a new career. When trained in the care of diabetic patients, nurses are often of more value to the patient than is the doctor. By living in the patient's home they can better adapt the diet to existing conditions and their opportunity for instruction of the patient and family is large. It is often desirable to carry out treatment from beginning to end with the aid of a nurse in the patient's home instead of sending him to the hospital, and it is usually more economical for the patient, provided the nurse is thoroughly trained in the disease. The requirements of a diabetic nurse are many she must be honest, accurate, intelligent, a good companion, cheerful and keen as to physical exercise. Her knowledge of the diabetic diet should be thorough. She should be able to examine a urine for specific gravity, reaction, albumin, sugar both qualitatively and quantita- tively either by the Benedict or fermentation test, and test for diacetic acid; understand ho\v to quantitatively estimate acidosis by one method such as estimating the ammonia in the' urine, the alveolar ('()..> with the I'Yidericia apparatus, or the Marriott method. She should have observed at least -0 cases of diabetes, including a case of threatened or actual coma. The qualifications are many but the character of the nursing is attractive. It is true that the supply of such nurses is seldom equal to the demand. Several of my nurses who have been with patients during preliminary treat- ment have later gone back to the same patients for a few days at a DIRECTIONS FOR NURSES IN CHARGE OF DIABETICS 477 time, in order to instruct them in the newer ideas of treatment which have arisen in the interim. A physician who carries through the preliminary treatment of a case of diabetes with one of these nurses is really taking a course in postgraduate medicine, and more and more physicians are asking for these nurses, so that they can treat their own diabetic patients without sending them to hospitals or specialists. Such a nurse pays for herself by the saving of laboratory and medical fees. TABLE 208. METHOD OF REPORTING DIET TO THE PHYSICIAN.' NAME on CASE No.... DATE Food. Breakfast. Dinner. Supper. Total grams. Carbo- hydrate. Protein. Fat. 5% veg .... 10% vcg . . . 100 200 150 450 15 8 Eggs .... 2 2 12 ]2 Meat, cooked Chicken 60 GO 16 10 Fish GO GO 12 Bacon 15 15 30 5 15 Butter 10 10 10 30 25 Cream, 20% Cream, 40% Oatmeal .... Etc 30 15 30 30 90 15 3 's 3 3 18 1 Totals . . Calories 26 4 59 4 81 9 Total calories = 10G9 104 236 729 The daily duties of a diabetic nurse are summarized below: 1. The nurse should familiarize herself with the printed matter which is given the patient. 2. The patient should be encouraged to write down all questions in a note-book and the nurse should keep a note-book of her own. 3. The treatment of diabetes is constantly improving as it is in any disease where there is not an absolute cure, but the general plan of treatment will not depart far from the directions given below. The nurse should understand the principles of treatment, for then she can take better care of the patient in the hospital and what is more important train the patient how to live after leaving the hospital. It is just as much a part of the nurse's duty to explain to the patient the details of the daily diet, with the values of carbo- hydrate, protein and fat allowed, as it is to serve the trays. Food given to a patient lasts him for a few hours; the explanation of food values lasts for life. 1 Diet cards. Thomas Groom & Co., Boston. 47S AIDS IX MA\A(!KMK\r OF DIABETIC 4. Collection of I'rlnc. Collect all urine and preserve in a bottle large enough to hold the twenty-four-hour quantity. This may be graduated to ">() cubic centimeters. The urine should be voided at 7 o'clock each morning and the twenty-four-hour quantity meas- ured at that time. Save at least 120 c.c. (4 ounces) as a specimen. Record date and (ju ntity in cubic centimeters both upon the specimen bottle and chart as follows: November 1(5-17, 1C>2() c.c. One ounce may be considered !>() c.c. (actually 2!).(> c.c.). Caution patients to save the urine while at stool. Preserve the urine during collection in a cool place. The vessel and the urinal are to be thoroughly cleaned and scalded each day. I rine may be kept from decomposition by placing in the bottle at the beginning of the day one teaspoonful of one of the following solutions: Toluol, 10 percent, alcoholic solution of thymol, chloroform. 1 Formalin (10 per cent, formaldehyde) may be used (5 drops to 1000 c.c.) if the urine is not to be tested for ammonia. The nurse should examine the urine daily for sugar and diacetic acid and once a week for albumin. ">. Wcitjhi of Patient. Weigh the patient daily at 7 A.M., after the urine has been voided. Note weight of clothes. Record total weight and weight of clothes. (i. Nurses should realize that the diabetic diet makes strenuous demands upon the digestive organs. A healthy individual could be easily upset by it; if a diabetic patient is so upset, serious consequences may result. (Jive the patient too little ratherthan too much food, and caution the patient to eat slowly. Plan in advance the day's diet, so that suitable quantities of carbohydrate, protein and fat shall be distributed between the three meals. (a) Record the character and quantity of all food actually eaten. Remember that food may be served and yet not eaten. (h) Allow slightly more carbohydrate at breakfast than at the other two meals, and rather more protein at noon. (<} If cream is allowed, the amount for the day should be measured out before breakfast and kept in one receptacle until used. (d) Five per cent, vegetables should always be on hand and care taken to secure perfect eggs. (e) In the preparation of vegetables bo sure they are thoroughly cleaned. Cut away all discolored and decayed portions. Only the tender portions of green vegetables should be served raw. The coarser parts may be cooked in soups. Vegetables should he deliciously cooked, served in an appetixmg manner, preferably as a puree, and if allowed, prepared with butter or cream. A portion of the carbohydrates may be removed by changing the water used in cooking, and if the water of ."> per cent, vegetables -erv;it ive, il iiiu>t !>< expelled by heat before DIRECTIONS FOR NURSES IN CHARGE OF DIABETICS 479 is changed three times, this constitutes "thrice-cooked vegetables" (see p. 533), which contain practically no carbohydrate. Two vegetables, of which at least one is cooked, should be served at noon and night, and one vegetable at breakfast as well, if the diet is without bread, oatmeal, potato, or fruit. (/) The composition of the meals depends upon the diet ordered, but usually they are arranged best as follows: Breakfast. Coffee, cream, bacon, an egg. If fruit is allowed a portion of the twenty-four-hour quantity can be taken. Dinner. Clear soup, n eat or fish, vegetables. Sn])})er. Tea or cracked cocoa, and a small portion of a cooked vegetable, salad, meat or fish. Never serve cold meat without some hot food. If fat meat is served, one must allow for it in reckoning up the diet. 7. Bowels. The bowels should move daily; give an enema if necessary. No cathartics should be given while fasting, unless prescribed for the particular case; at other times extract cascara sagrada, 1 to 5 grains once or twice daily; aloin, grain -i, or a compound rhubarb pill; less frequently Hunjadi, Apenta water, a Seidlitz powder or mineral oil may be employed as needed; but it should be remembered that the sodium bicarbonate in a Seidlitz powder may lead to the appearance of diacetic acid in the urine if the patient has a low tolerance for carbohydrate. 8. Exercise. Exercise should be moderate at first, later con- siderable, and should always be taken after meals. Caution patients not to get overtired, but encourage them to exercise vigorously, steadily increasing the amount of exercise to a point that would put a healthy individual into splendid physical con- dition. Patients must learn to know that restriction of exercise means restriction of diet. 9. Direrbion. Diversion should be furnished for patients. Encourage reading aloud, games and other activities which will enable the patient to enjoy life with others. Seek to prevent worry. Allow no conversation about the disease or the diet after the noon meal, and banish any thought of invalidism. Remember Eskimos formerly lived on a diet containing only about 50 grams carbohydrate per day. 10. If massage, is given, use cocoa butter or some other prepara- tion rich in fat, alcohol seldom or not at all because of its drying effect. 11. Instantly report loss of appetite, nausea, vomiting, restless- ness, unusual fatigue, excitement, vertigo, headache, drowsiness, listlessness, discomfort, painful or deep breathing, because such symptoms are frequently significant of diabetic coma. Neglect to report the above symptoms is inexcusable. Note the rules for the treatment of threatening coma on page 391. 4SO AIDS IN MANAGEMENT OF DIABETIC CASES ('. TABLE 209. DIAKKTIC HISTORY CHART FOR TIIK I'SE OF INSTITUTIONS AND OF PHYSICIAN'S ESPECIALLY INTERESTED IN DIABETES.' NAME AUK M S RKF. by Du. No. DATK Yrs. F M Mo.s. W ADDRESS ADDRESS OCCUPATION- RACK HEREDITY. Diab. incll. Obesity Gout Tuberculosis F. M. Children ]5ro. Sist. Other relatives HABITS. CJIsOII . Tobac. C't'a or Von. Exercise. Sleep. PAST HISTORY. ONSKT. J)atc: Acute. Gradual. SYMPTOMS. SUUAR FIRST DISCOVERED. CAUSE OF EXAM. ETIOLOUY. a. Obesity. b. Heredity. c. Dietary excesses. d. Strenuous life. e. Xervous: (1) organic; (2) functional. f. Infections. g. Arteriosclerosis, h. Syphilis. i. Trauma. j. Pancreas, abd. pain; fatty stools, k. Thyroid, size exophth. tremor nervousness palp, vom. diarrhea perspiration. 1. Hypophysis, aeromegaly adolesencc {iiuitisiii dial), insi]). in. Liver. n. Gout. o. Kidney. PREVIOUS TREATMENT. RESULTS SYMPTOMS SINCE ONSET. a. Weight : Highest At. onset First visit Height Date b. Loss of strength c. Polyuria d. Polydipsia e. Polyphagia f. Pains g. Extremities h. Cramps i. Bowels OTHER SYMPTOMS. a. Digestive c. Renal b. Circ. d. Resp. e. Nervous PRESENT COMPLAINT. PHYS. EXAM. General appearance Skin and mucous ineinb. Thyroid Pupils and muscles Eye-grounds Glands: C. A. I. Hearing Breath : acetone Al. air: ('()> Teeth Pulse Arteries Bl. pr. (recumbent) Syst . Dias. Apparatus Heart Apex, size sounds Lungs Abdomen Liver Spleen Kidneys Genitals Hernia, Edema Knee-jerks DIRECTIONS FOR NURSES IN CHARGE OF DIABETICS 481 gjj -uiui ).J 'JIT! JB[OOA[V juoo poo[$.[ spunod a gui pj^oj 'uiuouuuy piji: ju.vjnq.^xo-d o 31 4S2 1//AS L\ MAXAC-EMEXT OF DIABETIC CASES ' W o> W 03 5 P DIRECTIONS FOR NURSES IN CHARGE OF DIABETICS 483 .3 2 ~> "' ~* O ** CO iO 1C -^ C C O O "O CO O fl 3 e C >*.* . 484 AIDS IN MANAGEMENT OF DIABETIC CASES E. ACTUAL DIETS EMPLOYED TO RENDER PATIENTS SUGAR- AND ACID-FREE. 1. Case No. 92o, male, onset September 1915, at twelve years, nine months. TABLE 212. CASE Xo. 92:?. AGED TWELVE YEARS, NINE MONTHS. ONSET SKI-TKMHEH, 1015. Urine. Diet in scrams. Sugar. ^ a -^ c 2 g "5 Date. c a ~~ ~? ^ "> 2. C .i 'Z J? d J? ~ "* x CJ ; C *7* 'C Q ^ ~? 3 1 .3 6 . 1 "5 o ->; J2 " " J2 g > 3 ! ! U u 5 " A 1915. Oct. 9 2600 6.6 172 13 2600 7.7 200 14 Protein - fat diet with out my advice Entrance to hospital ev eningof October 15 i 15 Even- + + + + . S < - ing 15 16 200" + + + 000 79 16 17 1200 ++ 5 3 32 + 5 80 17-18 1150 + 10 13 , 5 137 +10 78 18-19 1150 + 13 28 16 30S +13 77 19 20 750 IS 29 16 332 +18 76 20 21 1675 21 31 31 487 +21 76 21 22 1525 24 33 43 605 +24 76 22 23 1900 38 43 66 91 S +3S 76 23 24 1950 39 ' 57 ' 67 987 +39 76 24 25 1700 39 59 80 1112 +39 76 25 26 1950 10 5 60 +10 76 26 27 1475 39 59 SO 1 1 12 +39 75 27 28 1600 (I 40 60 94 1246 +40 76 28 29 1400 17 62 95 1291 , +47 76 1916. Nov. S 1000 0.2 2 IOC 1917. May UP ... 55 65 87 1263 +55 Observe absence of aeidosis October !) and Y.\ before treatment was be'im, its appearance immediately after a fat-protein diet during one day prescribed without my knowledge and its gradual disappearance so soon as patient commenced fasting. If the fat- protein diet had been continued, would the ease have proved so benign? (See Tables 212 and 215.) '_'. Case Xo. U.'iS, male, onset September, 191"), age two years, four months; came under observation October 2">, 1915; weight in 1 In eleven hours to 7 A.M., October 16, n<> alkali iven. - Blood sutrar, 0.08 per cent. ACTUAL DIETS EMPLOYED 485 August, 1915, 32 pounds; no diabetic heredity. and 210.) (See Tables 213 TABLE 213. CASE No. 938. AGED Two YEARS, FOUR MONTHS. ONSET SEPTEMBER, 1915. Crii ie. Diet in grams. 03 -3 O Date. Sugar. Tjj "cj * ',mill]0 \ 5 . 1 s s I 3 >=; c- o o .a o T. t-t -H"3 f 1 a o s. s. o 3 1915. Oct. 25 25-26' 26-27 175 400 7.6 32 28 28 + 00 27-28 275 SI + 3 4 2 46 + 3 27 28-29 250 + + 0.6 1 10 12 8 160 + 9 26 29-30 300 000 7 3 40 + 7 27 30-31 350 000 12 6 72 + 12 27 Oct.31-Xov. 1 150 000 15 17 5 173 + 15 27 1- 2 300 000 15 20 1 1 239 + 15 27 2- 3 150 000 16 21 17 301 + 16 25 3 - 4 200 000 17 22 23 363 + 17 26 4- 5 225 : 18 23 29 425 + 18 26 5- 6 300 000 18 27 35 495 + 18 26 7- 8 150 00 19 28 41 557 + 19 26 8- 9 200 000 10 16 21 293 + 10 26 1917. Fob. 15-10- 1000 010 10 32 43 555 + 10 24 Xo alkali given. TABLE 214. ABBREVIATIONS EMPLOYED ix RECORDING FOOD UPON* THE DIABETIC CHART. asp. 1m. bts. b. br. sp. bt. cab. car. caul. ccl. ch. cl. cof. c. j. c.c. cr. c.c. cue. egpt. fh. g- asparagus bacon beets broth Brussels sprouts butter cabbage carrots cauliflower celery chicken clams coffee coffee jelly cracked cocoa cream cubic centimeters cucumbers egg plant fish grams gr. fr. let. mt. mshr. oat. oin. or. ol. r. o. oys. pot . rad. spin. sq. s. b. s. f. m. torn. tur. veg. wh. w. or c. grapefruit lettuce meat mushrooms oatmeal onion orange olives olives (ripe) oysters potato radishes spinach squash string beans sugar-free milk tomato turnip vegetable white with (cum.) 1 Blood sugar, 0.17 por cent. 2 Blood sugar, 0.08 per cent. 480 AIDS IX MANAGEMENT OF DIABETIC CASES X o H 2 c _; -w - ^ /} j, a-a H_ H ^ cJO -Co a S"So v? sS o^ oo C^f K^3_3 oj (H C < ^ 5 ' ~ i c -H C - .. 5 ) o '.-: CO CO "o "a - - o o lO "1 " ^ H O hn co c? m "3 "3 M ! -i; o_c-oo t- ^ H X ir' | " ?"l --." " "- -3 - S ) ?l -'-t T-H ^ ^H 1 ~^ ^~* r- '""' tl tC' ~* t ~_. ; o ,; o M g MV; 6 1 o -C^ , 0_ or: CJ J o " c H ^ a a o "^ 42 o o - . CO 4- . tib"~ c ^ <=5 C-J - ACTUAL DIETS EMPLOYED 487 1 G on G O U3 1? o o 6" r- PO Tn to m . . O O . * t- rH . ..Bo |i> . - . . 00^ ^H'* ^ a tc s o c O i < O g S fe S ^ . s o " ^s 48S AIDS IX MAXAdEMEXT OF DIABETIC CASKS P s e <- <->^ it ??5 If 7171 71 ACTUAL DIETS EMPLOYED 489 3. Case.Xo. 759, female, onset of diabetes at age of forty years. Marked diabetic heredity. Present observations made at age of fifty-five. Just prior to first observation the patient ate a fat- protein diet in order to make a favorable impression. Note the acidosis, its disappearance when fat was omitted, the decrease in the glyeosnria, although the carbohydrate in the diet remained essentially unchanged, and that the patient became free from acidosis and sugar within twenty-four hours of entrance to the hospital. Notwithstanding that the patient had suffered with heart disease for years and had had attacks of angina pectoris, she gave up her automobile and eventually walked three and a half miles daily before leaving the hospital. Died July 23, 1916. Angina Pectoris. ' (See Tables 217 and 218.) TABLE 217. CASE Xo. 759. WOMAN AGED FIFTY-FIVE YEARS; DURATION OF DIABETES FIFTEEN YEARS. (See page 341.) Urine. Diet in grains. "3 I .= o u e "3 ~ C "5 o c 5 .2 o ; -*^ i~ >> H ~ -3 "5 '- 3 'u -3"! - ~z 2 [c 5 .2 C ~ a 7- C 's. y. r ~ H 1915. Sept. 20 27 1500 + Ac. + + + + 47 little much much 4 164J 27-28 1200 35 45 little or none 28-29 72 14 + 30 little - or none 29-30 108 " 24 40 little - or none Entran -c to Xe\v Ent: land Done oness Hosp ital. Oct. 1 1100 tr. . . 156? 1- 2 1000 + 25 2.5 110 +20 ; 155J 2- 3 1550 25 2.5 110 +20 1351- 3- 4 1200 25 2.5 110 +20 ! 1551 4- 5 1250 25 25 . 17 253 +25 154^ 5- 6 1000 30 30 . 22 402 +30 , 153^ 0- 7 1500 + 32 30 . 18 410 +32 1521 7- 8 1550 33 31.0 30 520 +33 1511 8- 9 1500 " + 39 50 . 2S 008 +39 152 J 9-10 1500 41 52.0 40 780 +41 152^ 10-11 157 5 " 42 53 . 52 848 +42 ! 1521 11-12 1200 + 48 54 . 52 870 +48 1525 12-13 1200 + 4S 54 . 49 849 +4s 152 1 13-14 80 + 55 50 . 58 900 +55 152| 14-15 120 55 50.0 58 900 +55 1531 20 90 i 101) AIDS IN MAXAdEMKXT OF DIABETIC CASKS - ACTUAL DIETti EMPLOYED 491 t t> MM TABLE 219. CASK Xo. LSI, 1 ACKD .5o YEARS. DURATION OF DIABETES, 11 YKAKS. MY FIRST CASE OF INTKRMITTKNT FASTING. Trine. Diet in grams. ^ _ * c - c ^ " -r u a t^ c." i|ffi 1 c ~ S t; >i >>Ss "o-S.2 Date. c. - c j3 ~Ji ? ~6 si g ^ ~ J 'E _c o o o a "3 li-'^ i "o o"^ ~ .2 _o .t; ~ OS 5 03 c "3 r; ~~ 3 o. a > CJ S ~ y, < i ;_) O PL, Fn < O U * 1915. Sept. 19-20 1800 23 + + + + 11.3 3.7 74 109 20-21 2000 24 + + + . 34 3 . 36 30 5 140 - 6 107 21-22 2000 22 + + + . . ! 32 30 5 140 - 2 108 22- 23 2000 . . + + + 2 . 5 24 20 5 100 - 4 106 23-24 2000 + + + + .. 3.0 16 3 2 20 -13 107 24- 25 1SOO 23 + + + + 0.34 .. 2.8 11 -11 108 25 26 1800 26 _(- + + + 8.3 3.4 13 91 103 26-27 1SOO 27 + + + . . 2.6 20 140 103 27-28 1300 31 + + + .. 2.2 21 40 21 20 264 + 19 104 2s 29 1300 32 + + 31 40 21 20 264 + 9 103 29 30 1300 .. + + 6.5 1.9 40 40 21 20 264 103 Oct. 30- 1 1600 + + + 7.1 2.0 ! 6 20 140 - 6 106 1- 2 1400 27 + + + 20 636 101 2- 3 1700 + + + 9.3 1.9 20 20 50 24 20 636 103 3- 4 1800 i .. + + + 9 .3 1 . 5 29 20 50 24 20 636 - 9 104 4- 5 2250 . . + + + 1.7 29 20 50 24 20 618 - 9 106 5- (i 2150 .. + + + 1.4 4 20 140 - 4 105 6- 7 1400 . . + + + 1.0 3 20 140 - 3 105 7- 8 1000 .. + + + 1.4 4 10 50 24 20 596 + 6 105 8- 9 1400 .. + + + 1.5 10 10 50 24 20 596 105 9-10 1600 4-4- 13 10 50 24 20 596 ;.j 104 10-11 1400 + + + 8 10 50 24 20 596 + 2 104 11-12 1400 + + + 20 140 _ 9 104 12 13 1400 + + 4 10 50 24 20 596 + 6 104 13-14 1300 + + 10 50 24 20 596 + 10 105 14 15 1200 + + . . 2 10 50 24 20 596 + 8 105 15-16 1700 + + 10 5-1 28 20 64s + 10 105 16 17 1SOO + + 7 10 54 2s 20 6 is + 3 105 17 is 2200 + + 4 10 54 2s 20 6 is + 6 104 18-19 1600 + + 1 . 5 20 140 104 19 20 2500 . . i . . 60 10 20 120 i 20 21 1 S50 28 2s 20 504 105 21 22 1450 .. slight + 32 32 20 556 105 22 23 1500 34 34 20 5S2 105 23 21 2050 34 34 20 5S2 105 24-25 1500 + 34 34 20 5X2 105 25 26 1-100 20 1-10 106 26 27 1500 13 36 20 636 106 30- 31 1900 10 43 36 20 676 106 31- 1 2000 . . 10 51 11 10 6S3 105 Nov 1 2 SOO . . 5 35 106 2 3 1600 . . 2 44 40 20 684 + 2 106 3 4 2150 36 35 10 529 105 4 5 1SOO 44 40 10 606 105 5 6 1700 ' . . 16 45 10 659 105 6 7 1SOO I) 0. 19 . . . . o 46 45 10 659 105 7 s 1700 : . . 46 45 10 059 107 S 9 1500 . . 10 70 106 21 22 141)0 . . 1 43 53 10 723 + i 108 ( 402 ) 1 See jintic 493 and Table 220. HOSPITAL TREATMENT 493 4. Case No. LSI , male, onset in January, 1904, at forty-four years of age. February 5, 1908, (U> per cent, sugar. Became sugar-free in March, thereafter sugar usually present up to entrance to hos- pital September 19, 1915. Patient feeble and unable to read at entrance. Upon discharge could read a little and able to walk several miles a day. Died April, 191 (i. Coma. The reaction of the urine was acid throughout and the albumin never exceeded a very slight trace. No alkali was given save 4 grams of sodium bicarbonate September 20, 21, and 22. On fast days coft'ee, tea and broths were given freely. (See Tables 219 and 220.) TABLE 220. SUMMARY OF DIKT OF CASE No. 181. (See Table 219.) 1915. Sept, 20, 21 5 per cent. veg. 300; or. 2. 22 5 per cent. vcg. 300; or. 1. 23 5 per cent. veg. 100. 24, 25, 26 Fasting, save alcohol, as recorded on chart. 27, 28, 29 5 per cent. veg. 300; or. 2; lean mt. GO. 30 Fasting. Oct. 1 Fasting. 2, 3, 4 5 per cent. veg. 300; or. 1; lean nit, 120; eggs 2. 5, Fasting, save alcohol. 7, 8, 0, 10 As on Oct. 2, 3, 4, except, orange. 1 1 Fasting. 12, 13, 14 As on Oct. 2, 3, 4, except, orange. 15, 16, 17 Add sugar-free milk 60. 18 Fasting. 19 5 per cent. veg. 600; or. 2; gr. fr. 1. 20 5 per cent, veg. 300 (washed); mt. 30; eggs 3; bacon 10. 21 Idem., plus sugar-free milk 60. 22, 23, 24 Idem., plus sugar-free milk 90. 25 5 per cent. veg. 300 (washed). 26, 27, 28 5 per cent, veg. 300 (washed); mt. 30; eggs 3; cl. 6; bacon 10; Akoll 1; sugar-free milk 90. 29 5 per cent, veg. (not washed). 31 Idem., except mt. 60; whisky 20. Nov. 1 Fasting except 5 per cent, veg. 150 (washed); whisky 10. 2 5 per cent, veg. 300; mt, 60; eggs 3; bacon 10; cl. 6; Akoll 1; sugar-free milk 90; whisky 40. 3 Idem., except mt, 30; whisky 20. 4 Idem., except mt. 60. 5, 6, 7 Idem., except bacon 20. 8 Fasting, except 5 per cent, veg. 150 (washed). 9, 10 As on November 4. 11, 12, 13, 14 Idem., except mt, 30. 15 Fasting. 16, 17 As on November 11. 18, 19 Idem., plus Hepco 1. 20 Idem., plus 40 per cent. cr. 15. 21 Idem., plus 40 per cent, cr. 30. F. HOSPITAL TREATMENT. More than once I have cleared up the mysterious downward course of a case of diabetes by sending the patient to the hospital. 404 AIDS IX MAXAGEMEXT OF DIABETIC CASES It is surprising ho\v benign severe cases of diabetes become when under the physician's eye. On the other hand, great caution is necessary in the treatment of severe cases of diabetes in the first few days following their entrance to an institution. Habits of life and of diet are broken, and the patient is under some excite- ment. Formerly coma was by no means an uncommon occurrence, because of the radical elimination of carbohydrate and the change to an excessive protein-fat diet. Xow all this is altered. Never- theless, watchful care over each patient should be exercised. Xo matter how mild the case, the patient should be daily seen by the physician until his exact condition is understood. If acidosis is present, two, three, or four visits should be made during the twenty-four hours. I$y this means I believe I have more than once averted a threatening coma. I have' gradually acquired so much confidence in the methods of treatment outlined on pages oOf) and />S7 that I find hospital patients require far less attention. The patient's routine should be disturbed as little as possible. Entrance to a hospital by no means eliminates his going outdoors. It is quite as harmful to rest too much as too little. The utmost simplicity of diet should be maintained. A complete physical examination of the patient should be made just prior to his discharge. The first specimen of urine voided by the patient on entrance to the hospital gives little indication of his true condition. It is the rule to find that diabetic patients have made violent alterations in their diet in the fc\v days preceding entrance. They have either broken all dietetic rules which they may have previously followed with considerable care, fearing that their diet is to be curtailed, or, in order to make a favorable impression upon the physician, may have lived upon a far more rigid diet than that to which they have been accustomed. This was well illustrated by case Xo. 7."><). This patient had been free from acidosis for a considerable length of time, but, prior to her first visit to me, a diabetic relative had suggested restriction of diet, and this resulted in the appearance of a four-plus diacetic acid reaction. That this was but temporary was shown by its prompt disappearance when fat was withdrawn from the diet. Table 21 7. I lospital treatment is desirable for the untreated case of diabetes, because it gives him a liberal dietetic education. I Ic really is attend- ing school and acquiring knowledge which will be of value to him for the rest of his life. He learns from association with other patients, observes their mistakes and discloses his own. He gains confidence because he can observe the condition of patients as they leave the hospital and can compare it with his own upon entrance and \vit h that of others who conic later. lie learns what the mean- ing of suu'iM' in the urine is and how to test his urine. He has HOSPITAL TREATMENT 495 illustrated before him methods which will enable other patients and himself to become sugar-free, and, what is more important, becomes thoroughly familiar with what to do when sugar returns. The diet, which formerly was obscure, becomes simple, and he appreciates the reasons for its restrictions. Hospital treatment is desirable because it affords the usually tired diabetic patient freedom from worry, and even if his diet were unchanged, such a respite would be helpful. During the last year, and particularly since the newer ideas about the desirability of exercise have been introduced, the spirit pervading the hospital has changed. The prevailing attitude is one of cooperation and mutual helpfulness. Each patient is expected to teach someone else, and from each patient, as he comes in, the older patients expect to learn something which will benefit their particular case. It is advantageous to have some diabetic patients who are really very ill. I quite appreciate the help in the management of patients which 1 have received from having such come for treatment, having been told by others that they were past hope. Hospital treatment is actually dangerous for the diabetic patient if the nurses are not conversant with the details and execution of the diabetic diet. This is now obviated to a considerable extent because physicians give more precise directions. Formerly it was about as hazardous to the patient to have a nurse without diabetic training supervise his diet as it would be to be etherized by a nurse without experience in the use of anesthetics. Excellent nurses make grave errors in the treatment of diabetic patients. It is sur- prising how often nurses who have been trained in the foremost hospitals err in saving and measuring the urine and in carrying out dietetic orders. Of course, the physician is responsible for the nurses' mistakes in large share. Nevertheless, it is notable how poorly a diabetic patient gets on with some nurses and how well with others. After having had the same head nurses for months in charge of the diabetic patients at the New England Deaconess Hospital and Corey Hill Hospital, I wonder how the patients man- aged to get along as well as they did under the old regimen. Coma ceases to exist as an emergency when the patient is under the charge of a trained diabetic nurse. The labor entailed in educating a diabetic nurse in a hospital where diabetes is rarely treated is far more than that required to teach the patient. The patient has the diabetes not the nurse and is consequently the one most interested. Furthermore, the patient is aided in the treatment by his own feelings and can communicate these to the physician. In a hospital he often neglects to do this because he trusts to the nurse and often believes that whatever is done is all right and thus neglects to discuss questions which come up. 490 AIDS L\ MANAGEMENT OF DIABETIC CASES Hospital treatment fails of its purpose unless the patient is dis- charged to the care of his physician at home with a report of his condition while at the hospital and with recommendations for future care. This entails much labor, which at times may not he appre- ciated by either patient or physician, but it is a duty which the hospital doctor should never neglect. It has been my custom on Tuesdays and Fridays at nine o'clock, to give a clinic at the hospital for the benefit of the patients, their physicians, nurses and relatives. Every effort should be made to interest the family physician in the care of the case, and the directions given the patient should be in such form that a busy practitioner can readily acquaint himself not only with what has been done, but how he should proceed. The management of diabetic patients in many large hospitals is an example of inefficiency. Instead of having the diabetic patients grouped together where the diets could be easily supervised, they are often scattered throughout many wards. If the patients were assembled, labor in the kitchen, labor in nursing, and labor on the part of house officers and attending physicians would be saved. Furthermore, the patients would get far better treatment if they could be under the charge of a head nurse thoroughly skilled in their care, instead of having a nurse who had had only a few diabetic patients in her two or three years of training. Two anesthetists in the Mayo Clinic at Rochester have held their posi- tions for twelve and thirteen years respectively. Hospital superin- tendents and physicians, in their search for efficiency, should profit from this example. Each hospital nurse during her course of hos- pital training should eat at a metabolism table for one week, weigh all her food and calculate the carbohydrate, protein, fat and calories in her diet. A diabetic patient is likely to progress far less rapidly on the road to health in a ward with twenty to forty other patients, which is attended by three to five nurses with varying periods of duty and no special nurse 1 responsible for any special case, than under the efficient supervision of a head nurse trained in the management of diabetics. Furthermore, association of one diabetic patient with another diabetic patient is of mutual advantage. In one large hospital where I was asked to prepare a diabetic patient for a laparotomy, nine different nurses, all without adequate diabetic training, cared for the patient during the twenty-four hours preceding, during and after the surgical operation, and this was notwithstanding the fact that the administrative head of the hospital meant to cooperate with me. A little rearrangement of hospital routine would save hospitals in the aggregate many thousands of dollars. ('. II. Mayo' has emphasized the importance of efficiency in the 1 Modern Hospital, I'.ll-l, iii, p. L'l.">. HOSPITAL TREATMENT 497 hospital treatment of cases. The stay of cases of appendicitis in the Mayo Clinic has been reduced to about one week; so short a stay is not desirable in diabetes, but by full and constant instruction of patients and their relatives, the hospital period could be greatly shortened and patients will be more inclined to return after a few months to learn new methods of treatment and to assure themselves that they are following a proper course. The annoyances from diabetic patients breaking the diet in open wards are now practically abolished. All patients are expected to have the urine free from sugar. It does not take long for the patient to learn that although he may "cheat the doctor, he cannot cheat the disease." When one of my sugar-free diabetic patients carelessly ate figs, fasting followed automatically until his urine again became sugar-free. Furthermore, as I had taken the trouble to get him sugar-free the first time, I told him it was unnecessary for me to do so a second time, and as his eyesight did not allow him to do his own Benedict test, a friendly diabetic neighbor performed this for him. Thymol is probably the best antiseptic to use in preserving urinary specimens, but its expense makes it unsatisfactory. Chloro- form Avorks well, but the chloroform must be expelled by heat before testing for sugar, as it gives a positive reduction test with Benedict's and Fehling's methods. Frequently no antiseptic is required if the specimens are simply kept in a cool place, and particularly is this true if the bottles are sterilized. This is easily done with steam by the use of a wash boiler in which only a little water is needed. Next to chloroform as a preservative is toluol. Sufficient should be used to make a layer on top of the specimen. Formalin (40 per cent, formaldehyde) may be employed (5 drops to 1000 c.c.), but if used it makes determinations of ammonia untrustworthy. All specimen bottles should be tightly corked. Cheap bottles for collecting the urine are easily obtained. It is desirable that they be wide-mouthed so that they can be readily cleansed. The attractive percolator bottles graduated to 50 c.c. and of 2000 and 4000 c.c. capacity cost about one dollar apiece if bought by the dozen. Candy bottles of similar size can be obtained from drug stores, and these have the advantage also of glass stop- pers. Such bottles can be graduated by the use of diamond ink. The collection of urine is simplified if bottles of this type are employed. Labor expended in measuring the urines of ten diabetic- patients with a 500 c.c. graduate is considerable. This is wholly obviated by the employment of the large graduated bottles, by which the accurate pouring of small quantities of urine is avoided. The method of multiple examinations of urine for sugar has been described on page 177. By this means ten urines are examined in 32 1<)S AIDS IX MANAGEMENT OF DIABETIC CASKS about the same length of time as it formerly took to examine a single specimen. The weight of the patient recorded in terms of the naked weight should he taken daily at the same morning hour. The diabetic diet in a hospital is simple. Emphasis should be placed upon the point that the patient shall be taught the value of each portion of food he receives. It may be easier for the hospital dietitian to have the physician specify the grams of carbohydrate, protein and fat which the patient is to have, but in so doing the doctor fails to treat his patients unless the nurses explain the quantity of carbohydrate, protein and fat in each different article of food. From the first day of entrance to the hospital the patient should begin to learn his diet, and unless his food is given to him in forms which he can easily duplicate upon leaving the institu- tion, his hospital stay is a failure. Almost anybody can get a dia- betic patient sugar-free, but the education of the patient to care for himself upon leaving the hospital constitutes more than 90 per cent, of the treatment. Simplification in weighing food is accomplished by the use of scales 1 of 500 grams capacity with a movable dial which can be set at xero when an empty dish is placed upon them. Food can then be placed in the dish and the weight instantly read off. Such scales reduce the labor of weighing fully 90 per cent, as compared with scales where weights must be added to a pan or lever. Dietetic prescriptions should be given in simple form. These are illustrated by the tables on pages :]()(, 308, :>09, 4S(>, 4SS, 490, and 5:55. In the kitchen the diet for each patient can be written by the nurse with pencil on cheap paper blocks twenty by thirty inches in sixe. These are visible at some distance, so that the nurses actu- ally in charge of patients can fill the orders by glancing at the chart while about their work. Furthermore, it is easy to keep such slips for reference and by this means improvements in the diet can be made. A diabetic patient, upon entrance to a hospital, should be made to understand that he is taking a course in diabetes. For successful graduation in the course it is understood that he shall he able: ( 1 ) To demonstrate ho\v to test a urine for sugar. (2) To serve himself with approximate accuracy, without scales, 75 grams of a 5 per cent, vegetable. C!) To record a summary of his diet for the previous day. (4) To explain the quantity of carbohydrate which it contains. (5) To state his diet on his weekly fast day. ((}) To describe what he is to do if sugar returns in the urine. 1 John Chatillon & Sons, New York. II OS PI TA L TREA TMEN T 499 THE DIET OF TEN PATIENTS FOR A DAY AT THE NEW ENGLAND DEACONESS HOSPITAL, JANUARY 23, 1916, DISPLAYED IN THE KITCHEN FOR THE GUIDANCE OF NURSES. BREAKFAST, JANUARY 23, 1916. 1 Name. Bacon. Eggs Bal. . 2 Cleve. '. 45 2 Hoy. . . 1 John . Keif . . '. 20 1 Leff. . . McN. 20 2 Drap. . 20 1 Tier . . 30 2 Vegetables, 5 per cent. Butter. 200 100 50 100 100 100 10 10 10 10 10 10 Cream. 10 30 15 10 30 30 10 Orange. 0.1 DINNER, JANUARY 23, 1916. Name. Meat. Bal. . . 120 Cleve. Drap. 60 Hoy. . . 60 John . Keif. . ! 60 Leff. . McN. 60 Tier. . . 60 Eggs. Vegetables. o'o- 200 100 100 50 50 50 200 150 75 75 75 Butter. 10 10 10 10 10 Cream. 30 45 30 30 45 10 Orange. 0.5 0.5 0.4 Extra. Bran Bran Lister Bran Bran Bran Bran Extra. Lister Bran Bran Bran Strawberries, 75 g. Lister SUPPER, JANUARY 23, 1916. Name. Bal. . Cleve. Drap. Hoy. . John . Keif. . Leff. . McN. Tier. Meat. 6 oysters 90 60 6 oysters 60 60 60 Eggs. 1 1 Vegetables. 5 per cent. Butter. Cream. Orange. 200 100 200 250 150 150 150 150 10 5 10 10 10 10 10 15 45 45 15 15 15 10 Extra. Lister Lister Lister Lister Lister Several years ago the hospital class system for diabetic patients was employed by Dr. Mosenthal at the Vanderbilt Clinic in Xew York City; shortly afterward it was adopted by my former assistants, H. W. Goodall at the Boston Dispensary, and by F. G. Brigham at the Massachusetts General Hospital, where his diabetic clinic now includes 275 patients. Goodall 2 has found the results most striking. He attributes much of his success to the painstaking work of his secretary-chemist, who has had immediate supervision of the cases. 1 The weights are in grams. 2 Goodall: Boston Med. and Surg. Jour., 1916, clxxv, p. 156. AIDS I.\ MANAGEMENT OF DIABETIC CASES A special feature of the work has been the instruction of the members of the household responsible for the diet. Mosenthal 1 has described the system he inaugurated and pointed out how it may serve to familiarize physicians with the treatment of diabetes, aid in the instruction of students and improve the treatment of patients. lie calls attention to the necessity of limiting the number of patients in such a clinic in the same way hospitals limit the number of admis- sions to the wards. The usefulness of this method of instruction of patients is obvious. Each patient acquires a knowledge of the disease by coming in contact with many other cases, and he is sure to disseminate his knowledge. G. AMBULATORY TREATMENT. Most diabetic patients in the past have been treated in physi- cians' offices, and most diabetic patients in the future will con- tinue to be treated in the same manner. It is perfectly possible to carry out good treatment under these circumstsnces, but it involves great patience, much time in the education of the patients, and thus necessarily expense. It is advantageous in that the habits of the patients are little changed, the whole responsibility is placed upon the patient, and from the start the diabetic treatment is made to fit into the routine of life. The danger of this method is chiefly that the patient is not seen frequently enough at the very begin- ning of treatment, and with the alteration of diet there is a possibility of acidosis. With modern treatment, especially with the omission of fat from the diet for the first days, this danger is greatly lessened, and I see no reason why treatment of this character should not be maintained if physicians and patients so desire. (See Case No. 12:>7, p. o()(i, as an illustration of how simple the dietetic part of such treatment may be made.) The same routine as that carried out at the hospital should be taught the patient. In general, he should be made conversant with the rules embodied in A of this section. Much time will be saved if the patient is accurately taught at the start how to save the urine, to measure the food, and to record it neatly and in a systematic manner so that the physician can quickly detect mistakes in the diet. Under no condition allow the patient to tell what he has eaten, but insist on his presenting a dietary record. The urine should be examined and reported upon while the patient is in the office 1 , for unless this is done the patient is apt to feel, and he has a right to do, so that he is not getting satisfactory and prompt treatment. Intil a patient has been free from sugar and acid for 1 Mosenthal: Medical Record, 1015, Ixxxviii, p. 589. AMBULATORY TREATMENT 501 o g S - o u a tf a S3 El a S3 H 5j - o B u p ^ o U 2 O J3SUO JO JISIA }S[ ojnssdjd-poojy asinj 'A\ 'IV 'S xa S '"-'<> A i.2^ 'qjuoj\: . :t ri "~ C =2 'J^A fa < a ^ S5 a S ^H O a PH T 502 AIDS IN MANAGEMENT OF DIABETIC CASES at least a week, lie should report daily to the physician in person or by telephone. Most of my cases in the past have been treated as ambulatory patients. This was far more difficult under the older methods of treatment than today. With a fairly intelligent patient in a single visit the history can be taken, physical examination made, the prin- ciples of treatment explained and instruction given in the use of the Benedict test. Under favorable conditions it is then possible for a patient to attend to some of his routine work, become sugar-free and present himself for a second visit after a few days. Whereas it is seldom possible to accomplish as much at a single visit, it does show the contrast in the expense of treatment for a patient under the old and new regimen. SECTION VTI. FOODS AND THEIR COMPOSITION. A. DIABETIC FOODS. THE improvement in the treatment of diabetes owes mtieh to the recent dissemination of knowledge regarding the composition of foods. To the United States Government we are indebted for an excellent monograph by Atwater and Bryant entitled "The Chemical Composition of American Food Materials, Bulletin No. 28, Revised Edition," which was first issued in 1906. This can be purchased by sending ten cents in coin to the Superintendent of Documents, Washington, I). ( 1 . From this I have abstracted such analyses as are especially useful in computing the diets of both normal and diabetic individuals and have computed the calories per 100 grams instead of recording the same per pound. The State of Connecticut, from its Agricultural Experiment Station in New Haven, has also published excellent special reports on diabetic foods, and from year to year adds analyses of new preparations. So far these have been distributed upon request. Nearly all of the analyses in these reports will be found to have been incorporated in my tables, because they are of so much practical value. For ready consultation I have inserted lists of foods often used by diabetic patients, with their carbohydrate content, which also appear in the Connecticut reports, although some of these analyses exist in somewhat different form in the Government tables. To these latter lists have been added the values of protein and fat as well. Whereas the analyses of many so-called diabetic foods are recorded, no special food is recommended. Each physician must decide the merits of any particular food for himself. The medical profession is under great obligation to the Connec- ticut Agricultural Experiment Station for having provided for work of this character, and I regret very deeply that the apprecia- tion of physicians and diabetic patients cannot be shown in some tangible way. I am also personally indebted to Professor John (503) 504 FOODS AND THEIR COMPOSITION Phillips Street, who carried out these investigations, for furnishing me much helpful information on many topics relating to diabetic foods and for constructive criticism in the practical treatment of my cases. ll'haf /.s 1 a Diabetic Food? The definition of a "diabetic" food formulated by the Joint Committee on Definitions and Standards is as follows: "Diabetic Food. Although most foods may be suitable under certain conditions for the use of persons suffering from diabetes, the term 'diabetic' as applied to food indicates a considerable lessening of the carbohydrates found in ordinary products of the same class, and this belief is fostered by many manufacturers on their labels and in their advertising literature. A 'diabetic' food contains not more than half as much glycogenic carbohydrates as the normal food of the same class. Any statement on the label which gives the impression that any single food in unlimited quan- tity is suitable for the diabetic patient is false and misleading." Professor Street very properly lays down the following detailed requirements which a diabetic food should fulfil: 1. It should contain very much less carbohydrate than found in a normal food of the same class certainly not over half as much. 2. The label should bear a correct statement of the percentages of protein, fat, and carbohydrate present. 0. The amounts of the different carbohydrates present should be declared on the label, /. e., starch, sucrose, levulose, lactose, etc. 1. The processes of manufacture should be so standardized that uniformity of composition, within reasonable limits, will be main- tained from year to year. f). Xo statement should be placed on the label which would give the impression that any food in unlimited quantity is suitable for a diabetic; patient. ('). In the advertisements of these foods emphasis should be put on the carbohydrate content rather than on the amount of protein present. The narrow confines of the diabetic diet have greatly stimulated the manufacture of so-called diabetic foods. These are often serviceable, but are to be employed with discretion. Their use should be discouraged at the beginning of treatment. The patient should never become dependent upon special diabetic foods, for they are often unobtainable, always make him conspicuous, and when he acquires a disgust for foods of this class it is all the harder to abide by the original diet, \\heu the patient buys one of these foods, unfortunately he is often given a list of other diabetic foods and a new diabetic diet list, and confusion in the diet often results. The patients under my care who have done best either never use DIABETIC FOODS 505 special diabetic foods or only a few varieties, such as Akoll Biscuits, Barker's Gluten Flour, Casoid Flour, Ilepco Flour, Lister Flour, No. 1 Proto Putt's, and Sugar-free Milk. The user of diabetic foods should also remember that such foods may serve two distinct purposes: (1) as an aid in determining the diabetic's tolerance for carbohydrates, and (2) when such a tolerance is determined, as a source of supply of food containing reduced amounts of carbohydrates and suited to the individual patient's use. Diabetic foods of the first class should be as near carbohydrate-free as possible; foods of the second class may con- tain considerable amounts of carbohydrates and still be useful for the diabetic. It is apparent, however, that it is the province of the physician, not of the patient, to determine this tolerance; and, furthermore, it is evident from the analyses that the mere calling a product a "diabetic food" by no means establishes its right to such a name or its usefulness to the diabetic. Foods containing 60 or 70 per cent, of carbohydrates are no more "diabetic foods" than potatoes, rice, or oatmeal, and are no more entitled to that name. 1. Substitutes for Bread. Many of the preparations upon the market contain as great, or even a greater, quantity of carbohydrates than ordinary bread ; a few contain less ; but the percentage of carbo- hydrate may vary from time to time. Patients, and sometimes physicians, forget that substitutes for bread must be prescribed only in definite amounts. A diabetic bread should never be prescribed without a knowledge of its content of carbohydrate, protein, and fat. The bread of one of the largest bakeries in Boston, upon analysis, showed 55 per cent, carbohydrate. Bread made without milk or sugar, but with water and butter, contains 45 to 50 per cent, carbo- hydrates. Such a bread is undoubtedly superior to many different bread substitutes upon the market. The percentage of carbohydrate in toast is greater than in plain bread because it contains less water. Some of the coarser kinds of bread, such as rye bread, graham bread, black bread and pumpernickel, contain somewhat less carbo- hydrate. Never give bread substitutes early in treatment. Teach patients to live without them. There are four types of bread substitutes. (a) Bran Bread. Bran is being more and more employed in the diet of diabetic patients. This is neither more nor less than the use of cellulose, and this is supposed to have no effect upon the metab- olism. Unfortunately, the availability of the protein, fat and carbohydrate of wheat bran to the diabetic patient has not been determined, although there are plenty of data upon its digestibility by ruminant animals. Bread made of bran alone is not very palat- able, though with the fat of bacon or butter it is liked better. It furnishes bulk and acts favorably upon constipation. If made with .")()( i FOODS AND THEIR COMPOSITION eggs and butter tlie flavor is improved. Tt sliould be remembered that bran often contains a considerable quantity of starch. For this reason bran biscuits often prove to be a delusion and a snare and I dread to see them on a patient's tray. In large hospitals where diabetic patients are constantly being treated the danger is less, for the bran is bought by the same person and at the same place, but in private practice this is different. In purchasing bran go to a feed store and ask for coarse bran for cattle and not for bran for the table. The. various preparations of bran, bran breads, and cookies sold under trade names often contain carbohydrate other than bran, hence the reason for their palatable taste; beware of them! They may contain over 00 per cent, carbohydrate, of which less than 10 per cent, is real bran. Mild diabetics get into little trouble with bran, but the serious ones often suffer. The starch may be washed out with water by tying the bran in cheesecloth and fastening the same on a faucet. It should be thoroughly mixed and kneaded from time to time to be sure the water reaches all portions and should be washed until the water comes away clear. This may require an hour. 1 (b) Gluten Breads. '\ hese breads are made by removing the sugar- forming material from the flour. It is surprising how thoroughly this can be done. I have often found the percentage of carbo- hydrate in one such flour to be negligible. The large quantity of protein in small bulk which they contain is objectionable. ((} Light Breads. French bread cut in thin slices is often useful, because it is bulky, gives the appearance of a large quantity, and carries much butter. Manufacturers have taken advantage of this idea, and many light breads are on the market. These breads often contain about the same quantity of carbohydrate as ordinary bread, though a few contain considerably less. Their virtue often consists solely in their bulk, which allows a surface on which to spread butter. I seldom advise breads. It is better for the patient to forget the taste. (d) Various other substances have been used for Hour in the manu- facture of bread. Thus, aleuronat meal has been employed, and with it have been mixed various vegetable products. A group of casein breads is upon the market in the form of casoid flour and Lister's Diabetic Flour, and to some diabetics these are valuable. Soya bean is also extensively used, and probably deserves a still wider introduction into the diabetic diet. The carbohydrate in it is 'Four preliminary analyses of washed bran showed the following percentages of starch: (Mi, l.S, '2.7, ~>.'2 per cent. Two preliminary analyses showed pentosan '_".). s. " !>:>. .">. The wide variations in the percentages of starch will account for the occa- sional occurrence of suszar in the urine following the use of bran cakes. I hope these invest Stations will be continued in the laboratory from which 1 obtained these analyses,. DIABETIC FOODS 507 unassimilable. It is used in the manufacture of Ilepco Flour. Agar agar may be used to dilute the flour or to add to bran and also to re- lieve the constipation of thediabetic, which is frequently troublesome. 1 2. Substitutes for Milk. A few tablespoonfuls of cream are a great comfort to a diabetic patient. Except in cases with a very low tolerance a gill (120 c.c.) of 20 per cent, cream can generally be allowed, and if it is desirable to give more fat without increasing carbohydrate and protein, a gill of 40 per cent, cream is also well borne. Formerly patients took half a pint of 40 per cent, cream readily. With severe cases it is seldom possible to allow more than GO to 90 c.c. of 20 per cent, cream, for the balance of the fat which can be safely employed can more advantageously be taken in meat, butter, oil, and cheese. On the other hand, fat having been removed the chief value of the milk to the diabetic patient is lost. The percentage of sugar in sour milk is not much less than in fresh milk. Recently, sugar-free milks 2 have been put upon the market on a large scale, and many of my patients, particularly children, have found them of distinct advantage. These preparations of diabetic milk will keep from one to three weeks, and are conse- quently of great value to patients when traveling. As a rule they are concentrated one-half. Consequently they should be diluted before being used. They are so valuable for diabetic patients that I always encourage their use in small quantities at first, so that the patient can become accustomed to the artificial taste and can deter- mine the form in which the milk is most agreeable to him. This is often as equal parts of milk and Celestins Vichy. Williamson 3 suggested the following rule for the manufacture of artificial milk: "To about a pint of water, placed in a large drink- ing pot or tall vessel, three or four tablespoonfuls of fresh cream are added and well mixed. The mixture is allowed to stand from twelve to twenty-four hours, when most of the fatty matter of the cream floats to the top; it can be skimmed off with a teaspoon easily, and upon examination it will be found practically free from sugar. This fatty matter thus separated is placed in a glass. The white of an egg is added to it and the mixture well stirred. Then dilute with water until a liquid is obtained which has the exact color and consistency of ordinary milk. If a little salt and a trace of saccharin be added, a palatable drink, practically free from milk sugar, is produced, which has almost the same taste as milk, and which contains a large amount of fatty material. With very little practice the right proportions can be easily guessed, and of 1 See page 531. 2 D. Whiting & Sons, Boston. 3 Williamson: Diabetes Mellitus and its Treatment, The Marmillan Company, 1898, p. 334. 508 FOODS AND THEIR COMPOSITION course much larger quantities can be employed, in order to prepare a considerable amount of the drink at one time, than those men- tioned above." Rennet may be made from milk, but unless the curd is carefully washed it will contain 2 to 2.5 per cent, lactose. When the rennet is made from cream the lactose is materially diminished. Kefir contains approximately 2.4 per cent, milk sugar. Von Noorden says this milk has been of great help in the treatment of diabetes in children. Lawrence Litchfield, of Pittsburgh, gives whipped cream to his patients made according to the following rule: Add two ounces of 40 per cent, cream to a pint of cold w r ater in a Mason jar and have it shaken vigorously until the cream is thoroughly "whipped." Sometimes a trace of saccharin is added, usually not. "My patients like to eat this with a spoon, but, of course, it can be used in any way that is desired. It contains only a trace of sugar." The fermented milks contain about half as much carbohydrate as ordinary milk. B. COMPOSITION OF AMERICAN FOOD MATERIALS. 1 In Table 222 I have abstracted from the Government Publication of similar name the analyses of typical foods. These data, taken in connection with Table 223, will enable patient and physician to make substitutions in the simple diet scheme shown in Tables 120 and 124. TABLE 222. Protein, per cent. Fat, per cent. Total Caloric carbo- value hydrates, per 100 per cent. Kranis. Bocf, cooked: Roast 22 .3 28.6 356 Round steak, fnt removed .... 27 .6 7.7 1S5 Beef, canned: I )ricd beef '-to 2 5.4 211 Beef, corned and pickled: ( 'orned heef , all analyses 1.5 .(> 20 . 2 307 Mutton, cooked: Mutton, leg roast 25 22 . 312 Pork, pickled, .salted and smoked: Ham, smoked, lean 19 8 20.8 274 Bacon, smoked, all analyses .... 10 5 04.8 045 Sausage, A : Frankfort 19 18.6 1 . 1 25S Pork 13 44 . 2 1 . 1 468 Deerfoot Farm, cooked, analysis furnished by the manufacturers 19 . 9:5 54.21 0.34 Poultry and name, fresh: Chicken, broilers 21 . 5 2.5 111 Fowls ........ 19 ,3 10.3 230 Cioose, young 16 ,3 30 . 2 403 Turkey 21 1 22 . 9 299 'Bulletin No. 28, U. S. Dcpt. Agriculture, 1906. This can be obtained by send- ing 10 cents in coin to the Superintendent of Documents, Washington, D. (,'. COMPOSITION OF AMERICAN FOOD MATERIALS 509 TABLE 222. (Continued). Total Caloric carbo- value Protein, Fat, hydrates, per 100 per cent. per cent. per cent. grams. Beef liver . . 21.0 4.5 1.7 133 Chicken liver, as purchased . . . 22.4 4.2 2.4 141 Goose liver, as purchased . . 16.6 15.9 3.7 231 Mutton liver, as purchased . . . 23 . 1 9.0 5.0 199 Pork liver, as purchased .... . . 21.3 4.5 1.4 135 Turkey liver, as purchased . . . 22.9 5.2 0.6 144 Veal liver, as purchased .... . . 19.0 5.3 127 Fish, fresh: Cod, sections . . 16.7 0.3 72 Flounder, whole . . 14.2 0.6 04 Haddock, entrails removed . . 17.2 0.3 74 Halibut, steaks or sections . . 18.6 5.2 124 Mackerel, whole . . 18.7 7.1 142 Salmon, whole . . 22.0 12.8 209 Shad, whole . . 18.8 9.5 105 Trout (brook), whole .... . . 19.2 2.1 98 Fish, preserved and canned: Cod, salt, "boneless" .... . . 27.3 0.3 108 Herring, smoked . . 36 . 9 15.8 298 Sardines, canned . . 23.0 19.7 277 Shellfish, etc., fresh: Clams long in shell .... . . 8 6 1.0 2.0 53 Crabs, hardshell, whole . . 10 . 6 2.0 1.2 91 Lobster, whole . . 16.4 1.8 0.4 86 Mussels, in shell . . 8.7 1.1 4.1 63 Oysters, in shell . . 6.2 1.2 3.7 52 Scallops, as purchased . . 14.8 0.1 3.4 70 Terrapin . . 21.2 3.5 120 Turtle, green, whole .... . . 19.8 0.5 86 Gelatin 1 . . 91.4 0.1 375 Eggs, edible portion: 2 Hens', uncooked . . 13.4 10.5 158 Hens', boiled . . 13.2 12.0 168 Hens', boiled whites .... . . 12.3 0.2 55 Hens', boiled yolks .... . . 15.7 33.3 376 Dairy products, etc. : Butter . . 1.0 85.0 793 Cheese, American, pale . . 28 . 8 35.9 0.3 3 452 Cheese, American, red . . 29.6 38.3 476 Cheese, cottage . . 20.9 1.0 4.3 112 Cheese, Dutch 17.7 316 Cheese, full cream . . 25.9 33.7 2.4 429 Cheese, Limburger .... . . 23 . 29.4 0.4 369 Cheese, Neufchatel .... . . 18.7 27.4 1.5 337 Cheese pineapple , . 29 . 9 38.9 2.6 494 Cheese, Roquefort . . 22.6 29.5 1.8 374 Cheese, skimmed milk . . 31.5 16.4 2.2 290 Cheese, Swiss . . 27 . 6 34.9 1.3 442 Koumiss . . 2.8 2.1 5.4 53 1 I \inderstand many of the brands of commercial gelatin contain from S3 to S7 per cent, gelatin, 11 to 14 per cent, of moisture and 1 to 2 per cent, of ash. E. P. J. 2 One egg contains approximately protein G grams and fat 6 grams, of which one- half the protein and all the fat are in the yolk. E. P. J. 3 Estimated. E. P. J. 510 FOODS AND THEIR COMPOSITION TABLE 222. (Continued) . Total Caloric carbo- value Protein, Fat, hydrates. per 100 Dairy products, etc.: per cent. per eont. per cent. grams. Milk, condensed .sweetened .... 8.8 8.3 54 . 1 334 Milk, condensed, unsweetened, "evapo- rated cream " . . 9.6 9.3 Ho 1 TO Milk, .skimmed . . 3.4 0.3 . 5.1 1 /^ 37 Buttermilk . 3.0 5 1 Si o/> Milk, whole 3.3 4 ^t . o r fl oo o Cream 3.5 40 o . u t n / 2 *-JWA Flours, meals, etc.: o . u OtM) Barley meal and flour . 10.5 2.2 79 Q ^A1 Buckwheat flour 0.4 1.2 t t . o 77.9 OOl 356 Cornmeal, unbolted .... 8.4 4.7 74 '^Wl Hominy 8.3 0.6 79 Ool OftO Oatmeal .... 16.1 7 2 P,7 '-. oOo 1 no Rolled oats .... 10.7 7.3 \j t . O 66 2 *tuy J.H7 Rice 8.0 0.3 79.0 T:U / 359 Rice, boiled 2.8 0.1 24.4 112 Rye flour 0.8 0.9 78.7 359 Wheat flour, California fine .... 7.9 1.4 76.4 358 Wheat flour, entire wheat 13 8 1 .9 71 ^CiQ Wheat flour, patent roller process, high t i . j ouy grade (average of all analyses of high and medium grades and grade not indicated) . 11.4 1.0 75.1 363 Wheat preparations: Macaroni . ....... 13.4 0.9 74.1 366 Macaroni, cooked ...... 3.0 1.5 15.8 91 Bread, crackers, pastry, etc.: Bread, brown 5.4 1.8 47.1 231 Bread, corn (Johnnycake) 7.9 4.7 46 . 3 265 Bread, rye 9.0 0.0 53 . 2 260 Graham bread ... ... 8.9 1.8 52 . 1 266 Rolls, French 8.5 2.5 55.7 286 Rolls, all analyses 8.9 4.1 56 . 7 307 Toasted bread 11.5 1.6 61.2 312 White bread, home-made 9.1 1.6 53 . 3 270 White bread, miscellaneous 9.3 1.2 52 . 7 266 Zwieback 9.8 9 . 9 73 . 5 433 Crackers: Boston (split) crackers 11.0 8.5 71 .1 415 I'needa biscuit 1 10.1 8.8 70.0 399 Graham crackers 10.0 9.4 73.8 430 Pilot bread 11.1 5.0 74.2 396 Saltines 10.6 12.7 ON. 5 441 Doughnuts 0.7 21.0 53 . 1 440 Pie, apple .... 3.1 9.8 42.8 279 Custard 4.2 . 3 26 . 1 183 Squash .... 4.4 8.4 21 .7 185 Pickles, condiments, etc.: ( Hives, green- 2.14 12.9 1.8 137 ( Hives, ripe 2.0 21.0 4.0 220 Peppers i paprica), green, dried 1 5 . 5 s . 5 63.0 400 Banana:-, yellow . . 1.3 (>.(i 22.0 101 1 Analysis from Conn. Kxp. Sta. Report, 1914, p. 230. One biscuit weighs 7 grams and contains about ~> grams carbohydrate, 0.7 gram protein and 0.5 gram fat. 2 Univ. California College Agriculture, 1916. Personal communication. FOODS AND THEIR CARBOHYDRATE CONTENT TABLK 222. (Continued). 511 Total carbo- Caloric value Protein. Fat, hydrates, per 100 per cent. PIT cent. per cent. grains. Nuts: Almonds 21.0 54 . 9 17.3 667 Brazil 17.0 66.8 7.0 364 Butternuts 27.9 61.2 3.5 95 C 'host nuts, fresh G.2 5.4 42 . 1 248 Coeoanuts 5.7 50 . G 27 . 9 607 Filberts . 15. G G5 . 3 13.0 724 Hickory nuts 15.4 67.4 11.4 736 Peanuts 25.8 38. G 24 .4 563 Pecans 11.0 71.2 13.3 760 Pino nuts* pignolias 33 9 49.4 6.9 626 Pistachios, first quality, shelled . 22.3 54.0 16.3 659 Walnuts, California IS. 4 G4.4 13.0 726 Walnuts, California, black . . . . 27. G 56 . 3 11.7 683 Walnuts, California, soft shell 16.6 63.4 16.1 723 Miscellaneous: Chocolate 12.9 48.7 30 . 3 629 Cocoa 21.6 28.9 37.7 510 Coroal coffee infusion (1 part boiled in 20 parts water) 0.2 1.4 7 Soups : Home-made : Beef 4.4 0.4 1.1 26 Bean 3.2 1.4 9.4 Go Chicken 10.5 0.8 2.4 61 Clam chowder 1.8 0.8 6.7 43 4.G 4.3 5 . 5 81 Canned: Bouillon 2.2 0.1 0.2 11 3 8 0.9 4.7 43 Chicken soup ....... 3 6 0.1 1 .5 22 Consomme . . 2.5 0.4 12 Julienne 2.7 0.5 13 Mock turtle 5.2 0.9 2.8 41 3 7 0. 1 5.7 40 Oxtail 4.0 1.3 4.3 46 Pea. soup ......... 3 G 0.7 7.G 52 Tomato soup 1.8 1.1 5.G 41 Vegetable 2.9 0.5 14 For analyses of vegetables see pages 200 and 513. C. FOODS AND THEIR CARBOHYDRATE CONTENT. (See Connecticut Agricultural Experiment Station Report for 1913, Part I.) The basis for Table 223, "Foods and Their Carbohydrate Con- tent," was originally printed in the Connecticut Agricultural Experi- ment Nation Report for 1913. That report, however, gave the values for carbohydrate alone. Patients and physicians have found the table most useful, but in order to increase its worth, through the help of Professor Street the values for protein and fat are now added. 512 FOODS AND THEIR COMPOSITION TABLE 223. Carho . Protein, Fat, hydrates, I'OOD. ANIMAL PRODUCTS. percent. percent. percent. All ordinary meats, poultry and game (except those specified) . . Brains, bone marrow, pig's feet, gelatin, sweetbreads, tongue, eggs, beef juice and true meat extract ... . . Camembert, Cheddar, Edam, Minister, Roquefort and Swiss cheese . . . Tripe, dried beef and beef kidneys . . Trace Bologna sausage (range 0.2 -3 . 1) IS. 7 17.0 0.0 Pork sausage (range . 0-8 . 6) 13. S 14.2 1.4 Beef liver 20.7 4.5 1.5 Neufehatel cheese 18.7 27.4 1.5 Kephir 3.1 2.0 1.0 Skimmed milk cheese .... (range 0.0-9.0) 31.5 10.4 2.0 Full cream cheese (range 1 . 1-4.0) 25.9 33.7 2.3 Chicken liver 22 . 4 4.2 2.4 Pineapple cheese 29.9 3S.9 2.0 Frankfort sausage (range . 0-0 . 0) 19.0 18.0 3.4 Cream 2.5 IS. 5 4.5 Sheep liver 23 . 1 9.0 5.0 Whey 1.0 0.3 5.0 Calf's foot jelly 4.3 0.0 17.0 FISH AND SHELL FISH. All common varieties except those specified ..... . . Canned shrimp and lobster . . Trace Crayfish 10.0 0.5 1.0 Shad roc 20 . 9 3 . 8 2.0 Scallops (range 1.1-6.0) 14.8 0.1 3.4 Clams, round 8.0 0.8 4.7 Sturgeon caviare 30.0 19.7 8.0 OILS AND FATS. Butter, lard, tallow, oleomargarine, cod-liver oil, olive oil and other edible oils (crisco, oleo. 1 E. P. J.) ... . . CEHEALS (WHOLE GRAINS). Buckwheat 10.1 2.5 01.0 Millet 10.3 3.0 OS.O Oats, hulled 10.2 7.0 09.0 Corn (maize) 9.1 4.1 09.0 Barley S.5 1.1 70.0 Wheat 10. > 1.7 71.0 Rye 10.2 1.7 72.0 Rice, hulled 8.0 2.0 77.0 Rice, polished 7.1 0.3 79.0 FLOUUS, MEALS, ETC. Soy-bean meal 42.5 19.9 34.0 Pea Hour 25.7 1.8 57.0 Acorn meal 7.3 4.9 04.0 Craham flour 13.3 2.2 70.0 Kafir-corn Hour 0.0 3.S 70.0 Prepared wheat flour 10.2 1.2 73.0 Self-raising buckwheat flour 9.2 1.2 73.0 Pup corn, popped 10.7 5.0 77.0 Cassava meal 1.3 1.2 Sl.O Polato starch 0.9 0.1 Sl.O Sago starch . . 2.2 0.0 Sl.O Tapioca (Arrowroot) 0.1 0.1 S4.0 1 If these two preparations arc estimated as butter the error will be slight. E. P. J. FOODS AND THEIR CARBOHYDRATE CONTENT 513 TABLE 223. (Continued). Protein, FOOD. PASTES. percent. Banana flour 3.9 Corn starch 1.2 Noodles 13.3 Vermicelli 10.9 Spaghetti 12.1 CEREAL BREAKFAST FOODS. Rolled oats 16.7 "Holland Rusk" 12.1 "Ralston Health Food" 11.9 "Quaker Wheat Berries" 13.8 "Wheat let" 12.8 "Force" 10.6 Cracked wheat 11.1 "Pettijohn's Breakfast Food" 9.1 "Malt Breakfast Food" 13.8 "Cream of Wheat " 11.5 "Triscuit" 11.0 "Grape Xuts" 11. "Zest" Farina " Wheatena" 9.0 11.0 11.3 A.STRY. 33.6 "Mapl-Flake" 11.0 "Shredded Wheat Biscuit" 1 8.3 Hominy 7.6 Puffed rice 6.7 Toasted corn flakes BREAD, CRACKERS AND P. Peanut bread Acorn bread Cassava bread Peanut zwieback Whole rye bread Mince pie (range 30.0-44.0) Apple pic Corn bread Brown bread Whole wheat bread Graham bread Doughnuts (range 45.0-63.0) Rye bread Wheat bread Rolls Toasted bread Cake (except fruit cake) . . (range 53 . 0-78 . 0) Jumbles (range 52.0-71 .0) Alfalfa bread Fruit cake Macaroons (range 57 .0-70.0) Crackers (range 63 .0-M .0) 1.0 0.0 0.8 2.0 0.4 7.3 5.1 1.7 1.9 1.6 1.1 1.7 2.0 1.5 0.9 1.4 0.6 1.2 1.4 2.8 1.4 0.6 0.2 0.4 12.8 8.0 . 6 12.3 9.8 4.7 1.8 0.9 1.8 21.0 0.6 1.3 4.1 1.6 9.0 13.5 1.3 10.9 15.2 8.8 Carbo- hydrates, per cent. 85.0 85.0 72.0 72.0 74.0 64 . 72.0 72.0 72.0 74.0 74.0 74.0 74.0 75.0 75.0 75.0 75.0 75.0 75.0 76.0 76.0 76.0 78.0 80.0 81.0 20.0 27.0 27.0 28 . 35 . 38.0 43.0 46.0 47.0 49 . 51.0 52.0 53 . 53 . 56.0 61.0 63.0 63 . 64 . 64 . 64 . 71.0 FRESH VEGETABLES. Lettuce 1.2 0.3 2.2 Cucumbers 0.8 0.2 2.3 Spinach 2.1 0.3 2.3 Asparagus 1.8 0.2 2.4 1 Weight of one biscuit, 30 grams, and it contains, approximately, carbohydrate 23 grams and protein 3 grams. 33 FOODS AM) THEIR COMPOSITION TAISLK 223. (Continued'. Oarho- FOOD. FKKSII VEOK TAKI.KS. Cont imied. 1'rotcin, i)er cent. Pat, per cent. liydnitos, per cent. Rhubarb 0.6 0.7 2 . 5 Kndive 1.0 0.0 2.0 Vegetable marrow 0.1 0.2 2.6 Sorrel 3.0 Sauerkraut .... 1.7 0.5 3.0 Beet greens, cooked . 2 2 3 . 4 3.2 ( 'elery 0.9 0.1 3.3 Tomatoes 0.9 0.4 3.3 Brussels sprouts . 1 .0 0. 1 3.4 Watercress .... 0.7 . "> 3.7 Sea-kale 1.4 0.0 3.S Okra 1.6 0.2 4.0 < 'auliflower .... 1.8 0.5 4.3 Kgg j)l:int .... 1.2 0.3 4.3 ">) 1 (i 3 4 7 Kadishes (range 2 '.7- 7.5) 1.3 0.1 5 . Leeks 1.0 0.4 6.0 Mushrooms' .... (range 2. 1S.O) 3 . :> 0.4 6.0 Pumpkins .... (range 3 0-14.0) 1.0 0.1 6.0 String beans . (range 3. o lo.O) 2.3 0.3 6.0 Turnips (range 2 3 1S.O) 1 .3 0.2 6.0 Celery root .... 6.3 Kohl-rabi ... (range 3 5-14.0) 2.0 0.1 7.0 Oyster ))lanl .... 1.2 0.1 7.0 Rutabagas .... (range 3 . 12.0) 1 .3 0.2 7.0 TruHles 0.1 0.5 7.0 Squash (range 3 . lo.O) 1 .1 0.5 S.O Heels (range (i. 10.0) 1 .(I 0.1 9.0 Carrots (range "> . 11.")) 1.1 0.4 9.0 Onions (range 4. 14.0) 1.6 0.3 9.0 Parsnips (range (>. 0-14.0) 1.6 0.5 11.0 Chicory 15.0 Peas . . . 7.0 0.5 15.0 Artichokes 2 .... 2.0 0.2 16.0 ^ ains 1(> Corn 3. 1 1.1 1 .).() Potatoes (range 13. 27.0) > ^ 1.1 20.0 Lima beans .... 7. 1 0.7 22.0 Sweet potatoes (range 10. 5 4-1.. r >) 1 .S 0.7 20 . Soy beans .... (range 10. 3 30.0) 2S.O DUIKD V i-;<;i- :TAHU:S. Beans 22 . ") 1.8 . r )f).0 ( 'ow peas 21 .1 1 .4 55.0 Peas 21 .(i 1 .0 5S . Lentils 2.") . 7 1 .0 59.0 Lima beans is. 1 1 .5 6.0 Beans, haricots verts Asparagus Brussels sprouts Okra Tomatoes String beans . .... 'The carbohydrate which theK AGENT AND BRAND. percent. percent. 1 Alfonso it Hipolito: Sancho Vinos do Jerez Amontillado Don Quixote (Wm. J. Sheehan Co., New Haven, Agents) 20.00 1.23 Brotherhood Wine Co., New York City: Sunnyside Claret 11. S7 0.10 Riesling 12.37 0.34 Yin-Crest Brut 12.24 1.00 California Wine Association, New York City: Riesling 11.31 /infandel 11.02 C'alwa Distributing Co., New York City: 2 "Calwa" Brand Greystone (Light Hock Type) 5 . . . 11. SI 0.19 "Calwa" Brand La Loma (Burgundy Type) 2 .... 11.27 0.14 "Calwa" Brand Vine Cliff (Riesling) 2 10.90 0.17 "Calwa" Brand Winehaven (Table Claret) 2 .... 11.40 0.14 II. T. Dewey & Sons Co., New York City: I ves Claret ... 12.53 0.24 Moselle Type S.37 0.14 Old Burgundy Type 11.14 0.27 Ruby Claret 13.03 0.27 Pedro Domecq's Manzanilla Sherry 3 20. so 0.32 Kmpire State Wine Co., Penn Yan, N. Y.: DryCatawba 12. SO 0.15 State Seal Champagne 12.39 1.51 Los Angeles Co., Boston, Mass: California Chasselas 12.12 2.97 California Chasselas 11. OS 2.99 California Cutedel 11. S7 0.79 California Cutedel 11.50 0.19 Monticello Wine Co., Charlottesville, Va.: Extra V. Claret 12. SO 0.25 Norton's Virginia 12.57 0.37 Virginia Claret 12.54 0.20 Virginia Hock 12.00 0.22 A. Pierlot it Co., Bouzy, iiheims: Champagne Vin Nature sans Sucre 11.97 0.30 Pleasant Valley Wine Co., Rheims, N. Y".: Claret 11.22 0.29 DryCatawba 12.02 O.ls Groat Western Kxtra Dry 12.33 4.30 M. Schreiber, Baden, Austria: Diatotischer Rothwoin (K. Loeb it Co.. New York, Agents) 11.21 0. 15 Diiitetisoher Weisswein (K. Loeb it Co., New York, Agents) 10. 4S 0. 1 I William J. Sheehan Co., New Haven, Agents: California Cabernet 11.49 0.31 California Hock 11.21 0.14 California Rie^ing 11 .15 0.14 California /infandel 11.32 0.10 frbana Wine Co., I'rbana, N. Y.: Gold Seal Brut 12.14 2.30 Gold Seal Absolutely Dry 12.05 0.54 Gold Sparkling Rod, Special Dry 11.20 2. SO Gold Sparkling Red, Absolutely Dry 1 1 . 9S 0.29 Wilholmi, Hock it Co., Frankfurt, Germany: 1910 Brauncbergor 11.10 1 Grains reducing sugars per 100 c.c. 2 Sold by M. /under it Sons, New Haven, Conn. 3 Sold bv Chris. Nander, Washington, 1). C. COMPOSITION OF SO-CALLED DIABETIC FOODS 519 D. COMPOSITION OF SO-CALLED DIABETIC FOODS. The composition of so-called diabetic foods has been more thoroughly studied at the Agricultural Experiment Station in Xe\v Haven than elsewhere. In using the tables of the Connecticut Agricultural Experiment Station it should be understood that the percentages in the protein column are uniformly calculated from the nitrogen found, using the conventional factor (5.25. \\ith pure wheat products the factor 5.7 gives more accurate results, and strictly speaking, the latter factor should be used for gluten flours and other gluten products. In baked products where the protein may be derived from other sources than wheat, such as soya beans, cotton seed, nuts, etc., it is impracticable to vary the factor with each particular food without causing endless confusion. Similarly, it has seemed to the analysts best to retain for the flours the old factor for the sake of uniformity. In the high-grade ground glutens, containing as much as 13.7 per cent, of nitrogen, the use of the proper factor would reduce the protein by about 7.5 per cent, and the nitrogen- free extract would be increased in the same proportion. The terms "nitrogen-free extract" and "carbohydrates" are used synonymously, but the explanation of nitrogen-free extract in the preceding paragraph should be borne in mind. The values given for starch, however, are absolute, being direct determinations and having no connection with the protein factor used. A "0" means no starch was found, but a blank space does not mean the same, indicating merely that starch was not tested for. TABLE 224. ANALYSES OF SO-CALLED DIABETIC FOODS. (For original analyses sec Connecticut Agricultural Experiment Station, Report for 1913, Part 1; Report for 1914, Part 5; Report for 1915, Part 5.) Date of analysis. Manufacturer and Brand. 1910 1912 FLOURS AN MEALS. Acme Mills Co., Portland, Ore. . Amthor & Co., Halle: Weizen-Protein Herman Barker, Somerville, Mas.s: 9 84 4 ] 1 1 1 9 4 77 4 4 8 71 4 Barker' s Gluten Food, "A" . SO 9 5 3 7 trace 1913 Barker' s Gluten Food, "B" . NO 1 1 6 7 2 3 7 1913 Barker' s Gluten Food, "C" . 84 1 G 8 6 3 4 1914 'Battle Creek Sanitarium Co , Battle Creek, Mich., 80 per cent Gluten Meal . . S4 o i 5 8 1907 Bischof & Co , London: Gluten Four 79 8 3 6 5 3(34 3 Gs 370 375 377 372 FOODS AND THEIR COMPOSITION TABU-: 224. (Continued). Date .j of Manufacturer and Brand. analysis. a c, 'Ss 2 C C-, 1.0 3.1 357 1913 Ceroo Co., Tappan, N. V.: Soy Bean Gruel Flour . . . 13. 1 21.4 24.9 trace 405 1013 Farwell tt Rhinos, Watertown, N. Y. (.lluton Flour 43. 1.2 40.0 3S.1 370 1013 Gluten Flour 40 ; i 1.1 42 . 9 32 . S 307 1013 Cresco Flour IS. 1.0 07.4 57.2 351 1013 Special Dietetic Food . . . 27. t > 2.S 50. 40.0 302 1010 Gericke, Potsdam: Aleuronat . . S3.- I 3 . 3 3.1 370 1013 ('.olden Rod Milling Co., Portland, Ore., Acme Special Flour . . . 15. > ; 1.4 71.4 57.9 301 1013 O. I?. Oilman, Boston, Mass.: Gluten Flour 47.; 2.0 40.4 31.4 369 1000 Karl Goldscheidcr, Carlsbad: Conalbin-Mehl, No. 1 . . 10 ! 0.4 7S. S 302 1010 Gumport, Merlin: I'ltramohl . . 30.. r > 44.0 9.4 585 100S Hazard's Wheat Protein .... 41. i > 1.2 49.1 374 1013 Health Food Co., New York: Almond Meal 50.; 14. S 17.9 trace 400 1014 Almond Meal 40.1 21.S 15.0 457 1911 G 15 X Cold Blast Flour, 25 I>o r cent, protein . . . 10. . 70 . OS . 9 307 1913 Proniren (Gridle-eake Flour) 37..' 1.2 . . 37 . 7 349 1013 Glutosac Gluten Flour . . 30. '. ) 2.3 47.5 30.9 370 1014 Gluten Flour No. 1 ... 75.1 0.9 12.S 7.1 362 1013 Protosar Gluten Flour . . 42.1 ' \ 1.7 46.4 30.3 372 1913 Protosoy Soy Flour . . . 42..' , 19.S 24.5 trace 440 1913 Pure Washed Gluten Flour . SO.; 1.0 20.5 7.0 3SO 1014 Gluten Flour 15.! 2.0 42.3 31.5 370 1014 Protosoy Soy Flour . . . 42.! 10.2 20.0 1.9 448 1014 Pure Washed Gluten . . . S5 . ( > 1.0 5.4 2.S 373 1S92-0 H. Ilundhausen, Hamm.: Aleuronat (pure) .... SO.' 0.5 4.0 305 1000 .Hreh Diabetic Food Co., Xew York: 1000 Diabetic Flour 14.: 2.2 71.9 00. O 1 305 1000 Diabetic Flour 12.1 1 . S 72 . 7 355 1013 Flour l 14. -1 2.3 72.0 00.9 370 1013 Patent Barley 11 . 1.0 SO. 2 07. S 3S1 1013 Patent Cotton Seed Flour . 40.- 12.7 21.3 0.0 396 1013 Patent Lentils Flour . . . 27.: 1.2 50. S 42.0 350 1013 Protein Flour 31 . 2.0 50.7 IS. 5 370 1913 Soja Bean Flour .... 42.: IS. 2 25. S 0.0 135 1913 Wheat and Barley Flour . . 11 . > 1.9 73.5 00. 2 2 358 1011 .Johnson Kducator Food Co., Boston, Mass.: 1 1 Ml ) 1(1 y the diastase method, without previous washing with water, ant calculated as starch. - Possibly in part due to the copper-reducing power of the agar-agar present. COMPOSITION OF SO-CALLED DIABETIC FOODS 521 TAHLK 224. (Continued). Date of analysis. Manufacturer and Brand. Go A 'o t- t- -So - o o & ft i & \t [. y -*- -^> ' **3 ~ 3 C fi O SR 83 ? ,8 : & > u iU "5^0 2 - O ; ft a a -3 -~ x O FLOURS AND MEALS. Continued. 1912 The Kellogg Food Co., Battle Creek, Mich.: 20 per cent. Gluten Meal . . 27.5 0.5 7 1.7 49.6 357 1913 40 per cent. Gluten Flour . . 43 . 7 0.9 4' 7.3 40 . 5 367 1912 80 per cent. Gluten . . . 81.3 0.9 6 . 2 365 1913 Kugene Loeb, New York: Gluten Cracker Meal . . . 27 . 8 7.7 5. 1.5 40.2 394 1913 Imported Gluten Flour . . 76.3 0.9 1 1.8 4.4 361 1913 Pure Gluten Flour . . . . 40 . 3 2.4 4( 5 . 3 39 . 6 368 1913 Whole Wheat Flour . . . 14.6 2.2 7( ).5 54.6 360 1913 Gluten Flour 43 . 9 1.1 4< 1.4 39.8 363 1915 Lister Bros., New York: Diabetic Flour 84 5 36 372 1913 Thos. Martindale & Co., Phila.: Special Gluten Flour . . . 40.3 1.5 4< ).l 41.4 371 1913 Mayflower Mills. Fort Wayne, Incl.: Bond's Diabetic Flour . . 40.2 1.3 4S 1.3 40 . 6 366 1913 Theo. Metcalf Co., Boston, Mass.: Soja Bean Meal, 18 per cent. starch 41.0 20.0 2. 5.0 .. 444 1913 Vegetable Gluten, 8.1 per cent. starch 80.4 1.5 ' ) . 8 5.9 374 1901-2 H. Niemoller, Giitersloh: Roborat . 82.3 3.7 '. 2.9 . . 374 1913 Pieser Livingston Co., Chicago: Gluten Flour 433 1 . 3 4( 5.2 38.4 370 1911 Pure Gluten Food Co., New York: Gum Gluten Flour . . . . 38.3 1.6 5( ).8 42.4 371 1900 Gum Gluten Ground . . . 50 . 1 1.9 3< ).6 38. 6 1 376 190G Hoyt's Gum Gluten . . . 31.8 1.6 52. O 1 358 1901 Plain Gluten Flour . . . 53 . 6 1.2 3- L 5 . . 363 1914 Hoyt's Gum Gluten Biscuit Crisps 52.7 0.5 3f >.0 I 31.2 368 1914 Breakfast Food . . . . 45.4 0.9 4( >.4 39.2 375 1914 Flour, 50 per cent. . . . 49.7 1.2 4] .5 37.1 375 1914 Flour, Ground .... 41.9 0.9 41- < . 1 42 . 6 369 1914 Granules 42.7 0.7 4^ i.8 41.9 372 1914 Noodles 40.5 1.2 4? ).l 41.8 369 1914 Self-raising Flour . . . 42 . 7 0.8 4c > . 39 . 357 1914 Special Flour . . . . 90 . 7 0.7 1 .7 2.2 376 1914 No. 1 Dainty Fluffs . . . 79.9 0.5 1] .3 10.7 370 1914 No. 2 Dainty Fluffs . . . 66.3 0.5 2^ t.9 21.9 369 1913 Rademann's Nahrmittelfabrik, Frankfurt: Diabetiker Mehl . . 37.9 0.8 5( ).7 46.8 362 1913 Sprague, Warner & Co., Chicago: Richelieu Gluten Flour . . 47.7 1.2 3< .7 31.6 368 1913 G. Van Abbott & Sons, London: Almond Flour 24 . 6 58 . 6 ', .9 0.0 657 1913 Gluten Flour 75.1 0.9 11 .6 12.4 359 1 Determined by the diastase method, without previous washing with water, and calculated as starch. D:it of analvs FOODS AND THEIR COMPOSITION TAHLK 224. (Continued). Manufacturer and ]5r;ui;, Texas: Allison's Cotton-seed Flour /JO . 4 11.2 1001 Troponwerke, Mulheim: Tropon . SS.5 0.8 04.6 63.5 54 . n 51.S 861 842 44S 872 8M 877 880 1018 1S02 1S02 1010 1010 1010 1010 1014 1014 1010 1010 1010 1010 1010 1010 1010 1010 1010 1S02 SOFT BHKADS. Ferguson Bakery, Boston, Mass.: Gluten Bread .... Frank (V: Co., Bockenheim: Protoin-HotfKonbrot . . Protein-Weizenbrot . . Fritz, Vienna: Aleuronatbrot Kleborhr Litonhrot Froinni & Co., Dresden: Contrliitinhrot Litonhrot Con^lnt in Drop CoiiKlutin- Gorirkc, Potsdam Doppcl-Por Dreifach-P Kinfach-Po Sifarhrot Karl Goldscheider, Carlsbad: Sinamylbrot Gumport, Berlin: Diahetiker-Dopiiel-Schwarz Diabctiker-Doppel-Weissbr Diubetiker-Einfaoh-Sc-hwarz niabotikor-Kinfach-Weissbr ritrahrot F. Giintlier, Frankfurt : Klel)orbr 2s. 2 17.8 2s8 2S4 (t . . . . 15 6 ,8 16 .6 250 Schwarz . 21 5 IS .6 8S 6 15 .4 jadcn: >t . . . . IS 3 47 o 85 X 14 .8 n )J s 50 s 1 1 86 1 20.2 85S wioba'-k 14. 8 21 3 (i 20.7 470 erhrot . 20 6 85 1 rterbrot . 80. f 6. 5 26 () 10.S 2S5 lerbrot . 17 S 1 S 48 2 2SO 87 3 5 . .3 15 12.8 257 1 Detcrmini'd l>y the diastase method, wilhoul previous washiiifj; with wati calouliited as starch. -Chiefly derived from Soja bean and therefore non-assimilalile and for p can be considered carbohydrate-free.- F,. P. .1. COMPOSITION OF SO-CALLED DIABETIC FOODS 523 TABLE 224. (Continued). Date of analysis. Manufacturer and Brand. r c, C 5 Carbohydrate, per cent. ~ C'alculated calories per 100 grams. 1906 SOFT BREADS. Continued. Health Food Co., New York: Glutosac Bread .... 27.4 2 7 36 . 1 29. 9' 278 1914 Glutosac Bread .... 27.2 2.1 31.1 22 2 1906 Health Food Co.: Protosac Bread 32 . 5 37 . 1.6 I 292 1914 Protosac Bread 29.8 1.8 35 . 2 27.7 276 1914 .1. Heinbockel & Co., Baltimore, Md.: Diabetic Bread for Diabetes . 8.6 1.5 52 . 1 40.4 256 1892-6 R. Hundhausen, Hamm.: Aleuronatbrot, low gluten 17.3 40.6 0.3 234 1906 Jireh Diabetic Food Co., New York: Whole Wheat Bread . . . 9.4 48.6 0.4 236 1913 Eugene Loeb, New York: P. & L. Genuine Gluten Bread 10.4 2.6 53 . 7 44.2 280 1914 P. & L. Genuine Glubetic Bread 38.8 4.1 25 . 7 19.2 294 1915 Lister Bros., Andover, Mass.: C^n^ciii 13 rend 36 6 18.4 o 322 1910 Rademann's Nahrmittelfabrik : Diabetiker-Grahambrot 9.8 5.1 49.4 45.6 283 1910 Schwarzbrot 14.5 2.5 15.5 45.8 283 1910 Diabetiker Weissbrot . 2.3.3 0.5 40.1 37 . 258 1910 "D-K" Brot (dry) . . 12.3 5S . 9 1910 Litoiibrot ..... 30 2 2.5 21 .0 17.5 230 1910 Seidl, Munchen: \leuronatbrot ..... 21.9 0.3 47.3 2> Z 3 ! J3 - 1 i I* cT 1 I Calculated calorics per 100 grains. HARD BREADS, ETC. Continued. 1913 Health Food Co., New York: Alpha Best Diabetic Wafer . 66 . 1 13.6 11.3 trace 432 1914 Alpha Best Diabetic Wafer . 67 . 1 8.4 11.7 1 .3 391 1913 Diabetic Biscuit . . . . 25 . 9.2 54.2 46 .5 , 400 1914 Diabetic Biscuit . . . . 35 . 9 8.8 46.5 39 .8 409 1913 Gluten Nuggets . . . . 30 . 2 12.8 48.3 38 .6 ! 429 1906 Glutona 22. 1 11.8 58.5 54 .9' 429 1906 Glutosac Butter Wafers . . 27 . 6 12.9 49.4 41 .2'! 424 1906 Glutosac Rusks . . . . 36 . 5 3.8 51.6 42 .o 1 387 1906 Wafers, Plain 29 4 9.6 49.9 41 . 6 1 404 1906 Zwieback 32. 5 6.9 49.3 40 .91 389 1913 No. 1 Proto Puffs . . . . 76 . 3 2.9 10.7 4 .3 374 1913 No. 2 Proto Puffs . . . . 56. 6 i 2.1 30.7 19 .0 368 1906 Protosac Rusks . . . . 40 . 9 2.0 48.7 43 .9 1 376 1913 Protosoy Diabetic Wafers . 43 . 1 24.9 21.2 4 .7 481 1906 Salvia Sticks 39 2 20.8 2.4 18 . 7 1 440 1914 Gluten Nuggets 31 7 14.3 45.7 34 .9 438 1914 Gluten Butter Wafers . . . 31 . 1 13.9 47.0 38 .9 438 1914 Gluten Rusks . . . . 39. 3 3.4 47.0 33 .6 376 1914 Gluten Wafers, Plain . . . 42 . 6 1.7 44.3 29 .6 ' 363 Gluten Zwieback . . . 36 . 4 7.7 46.6 32 .5 401 1914 Manana Gluten Breakfast Food 42 . 6 2.0 43.6 29 .9 363 1914 No. 1 Proto Puffs . . . . 72.3 2.8 13.0 9 .2 366 1914 No. 2 Proto Puffs . . . . 58 . 8 2.1 27 . 20 .7 362 1914 Protosac Rusks 39 7 3.0 46.7 35 .9 373 1914 Protosoy Diabetic Wafers . 37 . 1 23.5 29.3 14 4 477 1914 Salvia Almond Sticks . . . 22.3 29.9 41.0 28 3 523 1913 Heinz Food Co., Chicago: Gluten Biscuits . . . . 12.8 18.3 57.7 21 4 447 1914 Heudebert, Paris: Pain d'Aleurone pour Diabet- iques 76 . 1 1.5 9.2 4 2 354 1914 Pain do Gluten pour Diabet- iques 80. 7 0.8 6.5 3 4 356 1914 Pain de" Essential "en Biscottes 26.4 1.2 62.2 49 9 365 1892-6 R. Hundhauscn, Hamm.: Aleuronatzwieback, high gluten 66. 2 5.0 17.7 381 1892-6 Aleuronatzwieback, low gluten 22. 9 8.6 59.6 407 1894 Aleuronat-Biskuits . . . . 24 . 8 11.2 52.2 409 1891 Aleuronat-Kakes . . . . 20 . 1 9.4 64.8 424 1913 \leuronat-Kakes 54 5 27.4 . . tra no 492 1913 Huntley & Palmer's Akoll Biscuits . 54. 5 27.4 6.8 trace 492 1906 Jireh Diabetic Food Co., New York: Diabetic Biscuits . . . . 14. 8 3.7 72.3 65 .41 382 1906 Diabetic Rusks 14 . 6 5.0 67.7 374 1913 Dietetic Biscuits . . . . 13 . 2 7.4 70. S 49 6 403 1913 Dietetic Rusks 14 . 9 8.7 68.0 47 410 1906 Wheat Nuts 19.0 15.0 54.5 50 I 1 434 1 Determined by the diastase method, without previous washing with water, and calculated as starch. FOODS AND THEIR COMPOSITION TAHI,K 224. (Continued). C. ~ .Manufacturer and Brand. II.VHl) BUKADS, ETC. Coilt ilHUMl. 1 !)()(> Johnson Educator Food Co., Boston: Almond Biscuits .... 20 o s 8 54 . 3 50 . 1 412 L906 Diabetic Biscuits .... 25 3 7 5 59.0 54.0' 405 UK Hi Educator Crackers, Greseni Gluten 23 4 G 63 . 1 57 . 9 1 386 1013 Educator Gluten Bread Sticks 35 9 7 2 45. S 37 . 5 302 1911 Gluten Cookies 26 4 16 49 . S 37 . S 440 i DOC. Gluten Rusk, Gresoni Gluten . ) 1 3 68 . 1 63.3' 364 1 1)0(1 Gluten Wafers 30 3 4 61 .2 57.0' 370 1906 Glutinp, Greseni Gluten 21 9 b 07. 7 63. 1 1 366 1012 The KelloKK Food Co., Battle Creek, Mich.: A vena-Gluten Biscuit . 21 4 12 7 55 . 5 41.1 422 1913 Potato Gluten Biscuit . 41 5 5 48.0 30 . 5 363 1000 Pure Gluten Biscuit 4S 3 3 3 39 . 1 370 1013 Taro-Gluten Biscuit 31 3 5 57 . 7 4,S . 2 361 101:5 40 per cent. Gluten Biscuit 37 2 8 53 . 2 45.0 369 1012 SO per cent. Gluten Biscuit 82 4 9 4.4 4.7 355 1.S05 Kirche, Diisseldorf: Alcuronat-Kakes 17 13 7 61.8 430 1010 Klopfer Chernische Fahrik, Dresden: Glidinchrot 47 G '> 9 34 . 32 . S 350 1013 Eugene Loch, New York: Gluten Luff, Bread .... 27 9 9 2 54 . 2 44.1 111 101 1 Gluten Luft Bread .... 52 4 13 2 26.0 22 . 433 914 Chocolate Almond liars If) 3 41 31.8 .) . 7 561 914 Diabetic, Almond Macaroons . 46 5 37 7 s . o 0.6 558 9 1 4 Diabetic Dread Sticks 50 4 3 4 34 5 24 (i 371 914 Diabetic Chocolates II ")1 4 23 . 6.9 614 914 Diabetic Lady Finders 56 6 2S 3 (i.O I .s 505 914 Diabetic Sponge Cookies 54 7 30 1 5.0 1 .2 510 oi. v , Pure Gluten Food Co., New York: Gum Gluten Biscuit Crisps 12 9 7 4s. 5 30.3 372 1913 Rademai m's Nuhrmittelfabrik, Frankfurt: Diabetiker Biskuits . . . 20 G 10 G 44 . 5 25 . 9 473 1913 Hret/el 31 4 S 5 50.1 10.7 102 1913 Cakes 20 G 13 5 47 . 2 30. 1 120 lso3 I )iabctiker-( 'hokoladc-Hiskuits 11 9 27 21.9 11.8 516 Hademann's XiUirmit telfabrik, Frankfurt (( 'ontinued) : ioi"3 Diabetiker Dessert-Gel >:ick 22 2 12 4 27 . 5 5. 580 1913 I )iabetiker Makronen 23 2 Is () 20.li 3.0 607 1913 Diabetiker Staiificn .... 17 1 11 .2 20.5 21 .1 586 1010 Diabetikcr-Zwieback 2.) .2 11 .9 5 1 . 3 17.0 113 IS93 Krdmiss-Biskuits .... 31 . N 21 .5 30. 1 9.0 ISO 1913 Kiisestantien 17 . 5 33 7 46.4 38.0 521 S07 Schelle, ! Braunschweig: . \Ieuronat-Kakes .... 10 9 10 S 63 . 1 120 nit previous washing with water, and COMPOSITION OF SO-CALLED DIABETIC FOODS TABLE 224: (Continued). Date of analysis. Manufacturer and Brand. Protein, G rt Calculated calories per 100 grams. per cent I a Carbohydi per cent a G HARD BREADS, ETC. -Continued. 1910 Seidl, Munchen: Kleberzwieback 14 .8 7 .8 66 .6 396 1913 Roman Fill, Karlsbad: Carlsbad- Water Biscuits, "Sprudel" Brand . 10 .0 5 2 74 .8 55 .6 386 1913 G. Van Abbott & Sons, London: Caraway Biscuits for Diabetics 35 .6 37 . 5 15 .9 8 .6 544 1913 Diabetic Rusks for Diabetics . 70 .9 .8 16 .0 12 .6 355 1 Q1 'i 35 g 40 .7 13 .2 f, .9 562 L J i '> 1913 Gluten Biscottes or Rolls . 51 .6 2 .3 33 .0 29 .8 359 1913 Gluten Bread or Slices 54 .1 2 .2 30 .9 27 .4 361 1913 Gluten Butter Biscuits for Dia- betics 44 .1 33 .2 12 .7 9 .0 526 1913 Ginger Biscuits for Diabetics . 34 .6 39 .4 16 .7 10 .9 560 1913 Midolia Biscuits .... 17 .6 36 .4 31 .6 13 .4 524 1913 Walnut Biscuits for Diabetics 20 .9 57 2 12 .3 trace 648 1913 Waukesha Health Products Co., Waukesha, Wise.: Hepco Dodgers 41 .6 21 .3 20 .7 trace 441 BREAKFAST FOODS. 1913 Brusson Jeune, Villemur, France: Farine au Gluten .... 33 .9 .6 53 .8 48 .8 356 1910 Gluten Semolina .... 17 2 .5 71 .6 64 .9 360 1913 Farwell & Rhincs, Watertown, N. Y.: Barley Crystals .... 11 .5 1 .3 75 2 62 .7 359 1913 Cresco Grits 17 .8 1 .4 68 .6 54 .1 358 190S Hazard's Wheat Protein Breakfast Food 40 .1 1 .0 49 .7 368 1913 Health Food Co., New York: Mariana 37 .6 1 .9 46 .8 31 .0 355 1913 Jireh Diabetic Food Co., New York: Whole Wheat Farina . 12 9 2 3 74 .6 59 .5 371 1913 Frumenty 12 .3 1 7 77 .3 65 .4 374 1911 The Kellogg Food Co., Battle Creek, Mich.: Granola 13 9 8 76 .3 45 2 368 1911 Pure Gluten Food Co., New York: Gum Gluten Breakfast Food . 37 8 1 3 51 .8 37 .9 370 1911 Gum Gluten Granules . 45 5 1 6 43 .6 32 3 371 1901 Pure Gluten Breakfast Cereal 43 7 1 6 44 .4 367 Waukesha Health Products Co., Waukesha, Wise.: Hepco Grits. 1 MACARONI, NOODLES, ETC. 1910 Brusson, Jeune, Villemur, France: Pates aux Oeufs Macaroni 13 9 4 76 2 69 2 364 1910 Pates aux Oeufs Nouillettes . 14 4 5 75 .7 68 9 365 1913 Petites Pates au Gluten 18 6 1 70 4 61 2 365 1910 Vermicelle au Gluten . 18 4 4 72 4 65 8 367 1913 Jireh Diabetic Food Co., Xcw York: Macaroni 16 9 9 71 4 58 8 361 1913 Eugene Locb, New York: Home-made Noodles 41 8 5 5 41 7 36 7 384 Said to be identical with Waukesha Hepco Dodgers. f>28 FOODS AND THEIR COMPOSITION TAHLK 224. (Continued). Date z C. t of analysis. Manufacturer and Brand. c | Z ; J || S - - 5. 02 O MACAKONI, NOODLES, ETC. Cont'd. 1906 Pure Gluten Food Co., New York: Gum Gluten Macaroni 41 4 1 46 3 46.2' 360 1911 Gum Gluten Noodles . 36 6 2 4 51 4 42 . 374 PEANUT BUTTKK. 1899 Atlantic Peanut Refinery, Phila.. 28 7 46 4 16 o 6 . 2 598 1913 .1. W. Beardsley Sons, New York: Acme Red Brand 28 2 48 3 15 2 4.0 608 1913 Beech-Nut Packing Co., Canajoharie, N. Y 29 4 46 6 16 4.5 613 1913 A. C. Blcnner & Co., New Haven (distributed by) 29 7 47 9 14 3 4.6 607 1913 D. W. Brooke, Newark, N. J. 29 5 48 5 14 9 4.3 614 1913 Dillon & Douglass, New Haven (dis- tributed by) : Perfection 29 1 42 8 20 1 4.8 582 1913 H. J. Heinz Co., Pittsburgh, Pa. 28 9 47 3 15 2 4.0 592 1913 The Kellogg Food Co., Battle Creek, Mich 28 1 49 7 14 7 3 . 4 619 1913 Francis H. Leggett & Co., New York: Premier 29 7 43 7 18 8 6 . 5 587 1 9 1 3 MacLarcn Imperial Cheese Co., Detroit ^lich * Flagle 32 1 44 9 16 o 1 .3 597 1913 Nut Products Co., New Haven: Penolia 27 0. 51 3 13 o 3.9 625 1899 Peanolia Food Co., New Haven: Peanolia 29 9 46 7 13 3 5 . 6 593 1913 S. S. Pierce Co., Boston: Acharis Brand 28 7 48 3 14 G 5.1 608 ALMOND PASTE. 1902-3 Chapman, Chicago . . 1902-3 Henry Heide, New York . 1902-3 Spencer, New York . . 13.1 12.7 13.5 1913 NTTS. California Paper Shell Almonds (ed- ible portion) (Lawrence Co., Boston) Jireh Diabetic Food Co., New York: Diatetic Pine Nuts (Pifinolias) . The KolloKK Food Co., Battle Creek, Mich.: Pine Nuts ..... MALTKD NTTS. 1901 The Kellojm Food Co., Battle Creek, Midi.: Malted Nuts . . . . 23. 1913 Nashville Sanitarium Food Co., Nashville, Tenn.: Malted Nut Food 24. 1 Determined by the diastase method, without previou> washing with water, and calculated as starch, COMPOSITION OF SO-CALLED DIABETIC FOODS TABLE 224. (Continued). 520 Date of analysis. Manufacturer and Brand. J Fat, per cent. Carbollydrate, per rent. x C OTHER NUT PREPARATIONS. The Kellogg Food Co., Battle Creek, Mich.: 1913 Nut Bromose (Meltose and Nuts) 17.1 26.8 39.4 3.2 467 1906 Nut Butter (Sanitas) . . . 28.8 50.5 13.9 9. 1 1 625 1906 Nut Meal (Sanitas) . . . 29.0 51.7 12.1 8.9' 630 1906 Nuttolene (Sanitas) 12.7 21.8 6.3 272 1906 Protose (Sanitas) .... 22.6 9.2 3.6 188 1913 Nashville Sanitarium Food Co., Nashville, Tenn.: Nut Butter ...... 28.0 52 . 6 13.0 3.8 637 1913 Nutcysa 12.9 21.0 6.3 trace 266 1913 Nutfoda 20.8 8.0 6.8 trace 182 CHOCOLATE. 1913 Brusson Jeune, Villemur, France: Chocolat with Added Gluten a la Vanille 15.9 49.7 26.4 9.2 617 1913 Fromm & Co., Dresden: Conglutin- Diabetiker-Schokolade .... 17.6 39.1 32.7 4.3 553 1914 Karl Goldscheider, Carlsbad: Feinste Dessert-Schokolade f. Diabetiker; "9.98 per cent. carbohydrates" .... 11.4 57.6 25 . 4 5.0 665 1914 Feinste Mocca-Schokolade f. Diabetiker; " 10.26 per cent. carbohydrates .... 10.2 60.2 23.5 4.1 677 1914 Feinste Nuss-Schokolade f. Diabetiker; " 11.32 per cent. carbohydrates" .... 14.6 54.4 23 . 3 6.9 641 1914 Feinste Orange-Schokolade f. Diabetiker; "9.98 per cent. carbohydrates" .... 11.4 57.6 24.9 5.0 664 1910 Groetzsch, Frankfurt: Esschokolade (Orange) 10.8 60.7 17.2 12.0 658 1910 Kochscholkolade .... 25.3 25.1 26.1 15.9 432 1903 Plasmon Co., London: Plasmon Chocolate . .... 20.2 25.1 48.0 trace 499 1903 Rademann's Nahrmittelfabrik, Frankfurt: Diabetiker Chokolade 17.5 57.6 16.9 3.8 656 COCOA. 1913 Charrasse Gluto-Cacao .... 21.5 232 FOODS AND THEIR COMPOSITION bran the salt and agar-agar solution (hot). Mold into two cakes. Place in pun on oiled paper, and let stand half an hour; then, when firm and cool, hake in moderate oven thirty to forty minutes. The bran muffins naturally will be far more palatable if butter and eggs are added. This may be done providing the patient allows for this in the diet. If the patient is not upon a measured diet, then considerable latitude can be employed in making the bran cakes. BRAX CAKES FOR DIABETICS. Carbo- Protoin, Fat, hydrate, Food. Amount. grains. gnuns. grains. C'alorios. Bran ... 2 cups Molted butter . 30 grains . . 25 . . 225 Eggs (whole) 2 . . . 12 12 . . 150 Egg-white (1) . 25 grams 3 . . . . 12 Halt .... 1 teaspoonful Water. 15 37 393 Tie bran in cheese-cloth and wash thoroughly by fastening onto the water tap, until the' water comes away clear. The bran should be frequently kneaded so that all parts come in contact with the water. Wring dry. Mix bran, well beaten whole eggs, butter and salt. Beat the egg white very stiff and fold in at the last. Shape with knife and tablespoon into three do/en small cakes. If desired one-half a gram of cinnamon or other flavoring may be added. Each cake contains: protein, O.o gram; fat, 1 gram; calories, 11. Cracked Cocoa. Cracked cocoa (cocoa nibs) makes a most, useful drink for diabetic patients. This is not generally appreciated by the profession. The sample of cracked cocoa (cocoa nibs) used has been purchased of the- S. S. Pierce Co., Boston. It was analyzed by Professor Street, with the following result: Moisture 2. S3 Protein 14.09 Fat 51.12 Fiber 4.32 Ash 3.88 Starch ... 7.48 Reducing sugar, as dextrose, direct none Reducing sugar, as dextrose, after inversion . . . 0.94 The cocoa is prepared for the table by adding a cupful of the cracked cocoa to a quart of water and letting it. simmer on the back of the stove all day, adding water from time to time. Professor Street was good enough to analyze the infus'on, and wrote me: ''The cocoa prepared according to directions contained 0.0o2 per cent, of reducing sugar as dextrose direct and O.foJS per cent, of total reducing sugars." DIETETIC SUGGESTIONS, RECIPES AND MENUS 533 LEMOX JELLY (DIABETIC). Carho- Protcin, Fat, hydrate, Food. Amount. grams. grams. grams. Calories. Lemon juice . . 30 c.c. . . . . 3 12 Water .... 50 c.c. Gelatin . . . 4 grams 4 . . . . 16 Saccharin (to sweeten) Cream .... 30 c.c. 1 12 1 116 5 12 4 144 Soften gelatin in a part of the cold water. Heat the remaining \vater and lemon juice and pour over the gelatin. Stir until dis- solved. Add saccharin, strain into cups. Serve with cream. BAVARIAN CREAM (DIABETIC). Carbo- Protein, Fat, hydrate, Food. Amount. grams. grams. grams. Calories. Cream, 40 per cent. 90 c.c. 3 36 3 348 Water .... 10 c.c. Egg (i) ... 50 grams 6 6 . . 78 Gelatin ... 2 grams 2 . . . . 8 Saccharin (to sweeten) Flavoring (to taste) 11 42 3 434 Soften the gelatin in cold water, then add to the cream, which has been heated. Stir until dissolved, pour on the beaten egg, cook like soft custard, turn into mold and chill. ICE CREAM (DIABETIC). Carbo- Protein, Fat, hydrate, Food. Amount. grams. grams. grams. Calories. Cream, 40 per cent. 90 c.c. 3 36 3 348 Water .... 10 c.c. Egg (1) ... 50 grams 6 6 . . 78 Saccharin (to sweeten) Flavoring (to taste) 9 42 3 426 Make a soft custard of the egg, 50 c.c. of the cream, and the water. Whip the remaining 40 c.c. of cream and fold into custard. The saccharin may be added to the egg. The flavoring should be added last. Agar-agar Jelly. One-quarter of an ounce sufficient to make one quart of jelly. Agar-agar may also be added to broths. Miss E. Grace McCullough, Dietitian at the Peter Bent Brigham Hospital, has given me several practical suggestions about the preparation of hospital diabetic diets. Many of these have been incorporated in what follows. Thrice-cooked Vegetables. The vegetables are cleaned, cut up fine, soaked in cold water and then strained. The vegetables are 534 FOODS AND THEIR COMPOSITION then tied up loosely in a large square of double cheese-cloth large enough so that the corners of the cloth, after it has been tied up \vith a string, make conveniently long ends, and also large enough to allow the vegetables to swell without sticking together. They are then transferred to fresh cold water, placed on the fire, and brought to the boiling-point, at which temperature they are main- tained for from three to five minutes. This water is then poured oil' and replaced by fresh, and the vegetables again boiled a similar length of time. Three changes of water are usually sufficient to remove the carbohydrate, as has been proven by Professor Wardall's preliminary experiments. The pots for the vegetables should be of sufficient si/e to hold a large quantity of water, and in a hospital, vegetables enough for the daily supply of six patients. Vegetables thus cooked will keep in cold storage two or more days, and the reheating of the same in a steamer is a simple affair. If the vegetables are cooked with the cover left off the pot they will be lighter in color and the flavor not so strong. Miss McCullough has adopted several expedients by which variety in the 5 per cent, vegetables is obtained, and thus the monotony of the diet avoided. She suggests that the large outer stalk slightly green covering of cauliflower be carefully cleaned, cut into half-inch pieces and boiled until tender, and frequently this is transferred from four waters. Similarly, the green outside leaves and any small pieces of lettuce may be shredded and served like spinach. Chard in season can be purchased by the bushel, cut, and then chopped up. Rhubarb retains its acid flavor and has proven so acceptable an addition to the diet that in future it should be canned by the cold-water method for subsequent use. The flat, large, celery stalks with any or all the leaves, whether yellow or green, chopped fine, serve excellently well. White, green, and red cabbage is cut fine and served as cold slaw. Diabetic patients should be urged whenever possible to have a garden and to raise suitable vegetables for themselves for the ensuing winter. One of my patients does this and thus provides himself with the best of celery, cabbage, lettuce, etc. This patient eats a slice of cabbage, cut as one buys cheese in a grocery store, for breakfast each morning, and by this means keeps the bowels perfectly regular. Canned vegetables which have been of the most service at the Peter Bent Hrigham Hospital are of four varieties: soup asparagus, broad, flat, cut string beans, the tender, green, stringless bean, and the white wax beans. The pods are separated from the beans, the latter being used for the benefit of other patients. Soup aspara- gus proved to be excellent for hospital use. It is a by-product of the factory and consists of the broken-off tips and the shorter SEVEN MENUS FOR A SEVERE DIABETIC 535 thin stalks which are unfit for the standard size. The pieces are about one inch long and are all edible. Squab. A squab when carefully boned yields 50 grams of meat. This is broiled in an oiled paper case to prevent evaporation, and when served with the -escaped juices proves a favorite dish for patients. It contains about 12 grams protein and 5 grams fat. Boiled Dinner. Corned beef, with cabbage and one other vege- table, served together as a boiled dinner, is most acceptable to male patients. A portion containing 50 to 75 grams of meat and 100 grams of each vegetable makes an excellent meal. Corned-beef hash made of meat and vegetables in the same proportion could also be served for variety. The proper seasoning of the food is a great help to the diabetic patient. So many articles are excluded from the diet that the great variety which is possible in the preparation of the food by the help of seasoning is overlooked. Horseradish, to be sure, contains 10 per cent, of carbohydrate, but it would take at least two tea- spoonfuls' to contain a gram, and probably far more. Sour pickles are allowable, and other pickles made from the group of 5 per cent, vegetables, provided one is assured that they have been prepared without sweetening. Mint, capers, curry, tarragon vinegar, onion bay leaf, and cloves may all be used as seasoning, and tomato and onion stewed, to which bay leaf and cloves may be added and then thickened with Irish moss, serves as a sauce. SEVEN MENUS FOR A SEVERE DIABETIC. For the menus and the recipes which make them possible I am greatly indebted to Miss Alice Dike, Instructor in Household Economics at Simmons College, and to Case Xo. 7(55. The directions given were as follows: Carbohydrate, Protein, Fat, Daily dietetic prescription. 1 gnis. K'ns. gms. Five per cent, vegetables, 300 grains 10 5 Eggs, 2 12 12 Bacon, 30 grams 5 15 Butter, 30 grams 25 Cream, GO grams. 40 per cent. ... 2 2 24 Meat, 120 grams 32 20 Lister roll (2) 12 12 12 68 10S The calories furnished amount to about 1200 a maintenance diet for a patient weighing 40 kilograms and a sufficient diet for a patient of 50 kilograms when in bed. 1 These represented the dietetic orders for one week, and from the foods mentioned in the list the menus which follow were prepared. 530 FOODS AND THEIR COMPOSITION Soft-boiled egg, 1. Fried bacon, 30 grains. Lister roll and butter, S grains. Coffee and eream, 30 grains. FIRST DAY. Breakfast. Lunch. Roast beef, GO grams; grated horseradish. String beans, 75 grains, and butter, 7 grains. Lettuce and cucumber salad, 50 grains. Rhubarb jelly and meringue (rhubarb, 25 grams, and 2 white of egg). Dinner. Chicken, GO grains. Cauliflower, 75 grams, and butter, 7 grains. Celery and olives, 75 grains. Lister roll and butter, S grams. Coffee, Spanish cream (egg 1 and cream 30 grams). SECOND DAY. Breakfast. Shirred egg, 1. Fried bacon, 20 grains. Lister roll and butter, S grams. Coffee and cream, 30 grams. Lunch. Roiled haddock, GO grams. Cucumber sauce, 25 grains. Rutter, G grams. Spinach, 75 grams, and butter, 8 grams, and \ egg. Lettuce, 30 grams. Coffee jelly whip. Dinner. Lamb chops, GO grams; tomato sauce, 45 grams. Asparagus, 75 grams and butter, S grams. Dandelion greens, 50 grains, and bacon, 10 grams. Lister cream puff and custard. THIRD DAY. Breakfast. Egg, 1; scrambled with tomato, 50 grams. Bacon, 20 grams. Lister roll and butter, S grams. Coffee and cream, 30 grams. Lunch. Vegetable hash (corned beef, 10 grams; cabbage 1 , NO grams; onion, 10 grams; beet, 10 grams; bacon, 10 grams). Lett uce, 30 grams. Lister roll and butter, S grams. Tea. Dinner. Steak, SO grams, and butter, 7 grams. Rroiled pepper, 25 grams. Cauliflower, 75 grams, and butter, 7 grams. Wine jelly and egg and cream sauce (egg, 1, and cream, 30 grams). SEVEN MENUS FOR A SEVERE DIABETIC 537 FOURTH DAY. Breakfast. Liver, 40 grams, and bacon, 15 grams. Lister roll and butter, 10 grams. Coffee and cream, 15 grams. Lunch. Ham omelet (egg, 1, and meat, 20 grams). Bacon, 15 grains. Salad, 150 grains (celery, cabbage, lettuce). Lister roll, butter, 10 grams. Cracked cocoa and cream, 15 grams. Dinner. Roast lamb, 60 grams, and mint sauce. Sliced tomatoes, 75 grams. String beans, 75 grams, and butter, 10 grams. Vanilla ice-cream (egg, 1, and cream, 30 grams). FIFTH DAY. Breakfast. Scrambled egg, 1, and dried beef, 20 grams. Lister roll and butter, 6 grams. Coffee and cream, 20 grams. Lunch. Spinach soup (spinach, 25 grams; cream, 15 grams; yolk 1 egg, stock). Bacon, 30 grams; fried with eggplant, 125 grains. Coffee. Dinner. Steak, 100 grams, and water cress, 25 grams; "Maitred'Hotel" butter, 10 grams. Vegetable marrow, 125 grams, and butter, 8 grams. Lister roll and butter, G grams. Cracked cocoa whip (white 1 egg and cream, 25 grams). SIXTH DAY. Breakfast. Fried fish cakes and butter, G grams (fish, 40 grams; egg, 1; cream, 15 grams). Sliced cucumbers on lettuce, 75 grains. Coffee and cream, 15 grams. Lunch. Fried egg, 1, and bacon, 30 grams. Lister roll and butter, 10 grams. Cold slaw, 75 grams. Tea. Dinner. Broiled swordfish, 80 grams (drawn butter sauce, 7 grams, and parsley). Brussels sprouts, 100 grams, and butter, 7 grams. Tomato jelly salad, 50 grams. Lister roll and whipped cream, 30 grams (flavored with coffee). SEVENTH DAY. Fasting. ,")3S FOODS AM) THEIR COMPOSITION RECIPES USED IN PREPARING THE PRECEDING MENUS. Grated Horseradish Sauce. 1', teaspoonfuls grated horseradish. i teaspoonful vinegar. i teaspoonful salt. Cayenne. 2 teaspoon! uls cream or water. Mix first four ingredients and add cream beaten stiff. Cucumber Sauce. (Irate 25 grams cucumber and season with salt, pepper, and vinegar. Tomato Sauce. Stew 45 grams tomato, season with salt, pepper, clove, and bay leaf. Irish or sea moss may be used for thickening. Parsley Sauce. 7 grams butter. 1 teaspoonful chopped parsley. Salt and pepper. Add parsley lo melted butter just before serving. Mint Sauce. J cup finely chopped mint leaves. \ cup vinegar. 1 grain saccharin. Add saccharin to vinegar and dissolve, pour over mint and let stand thirty minutes on back of range. Let cool before serving. Maitre d' Hotel Butter. 10 grams butter. Salt and pepper. 1 teaspoonful chopped parsley. i teaspoonful lemon juice. Put butter in bowl and with wooden spoon work until creamy. Add seasoning and lemon juicy slowly. Coffee Spanish Cream. 1 scant teaspoonful gelatin soaked in 1 tablespoon ful cold water and dissolved in 5 tablespoont'uls hot coffee. Add )-!() grams- cream and pour on slightly beaten yolk of egg. Cook liko soft custard and pour while hot on stiffly beaten white of egg. Saccharin. Rhubarb Jelly with Meringue. 1 teaspoonful gelatin soaked in 1 tablespoont'ul cold water and dissolved in sauce made by cooking rhubarb in enough water to make 7 tablespoonfuls. Serve garnished with beaten white of egg flavored with vanilla. Saccharin. Coffee Jelly Whip. Make the same as plain coffee jelly, but just before it hardens beat in an egg beaten until fluffy. Saccharin. RECIPES USED IN PREPARING PRECEDING MENUS 530 Lister Cream Puff. Lister biscuit with soft custard poured over it. The soft custard is made as follows: 30 grams cream. \ egg. 2 tublespoonfuls water. Saccharin and flavoring as desired. Wine Jelly with Custard Sauce. 1 scant teaspoonful gelatin soaked in 1 teaspoonful cold water and dissolved in 4 tablespoonfuls boiling water and flavored with 3 tablespoonfuls wine and saccharin. Serve with sauce used above for Lister cream puff. Cracked Cocoa Whip. 1 scant teaspoonful gelatin, soaked in 1 tablespoonful water, dissolved in 5 tablespoonfuls strong hot cocoa. When cooled to the consistency of thick cream, pour slowly on the beaten white of an egg, beating all the time. Mold and chill. Spinach Soup. 25 grams spinach. 15 grams cream. Yolk of 1 egg. | cup beef or chicken stock. Add stock to cooked spinach and cook five minutes. Then rub through sieve. Beat yolk of egg with cream. Add spinach and stock and return to double boiler. Cook one minute and serve at once. INEXPENSIVE MENUS. Diet for Day. Carbohydrate, Protein, Fat, pins. grns. Kins. Five per cent, vegetables, three times washed, 300 grams Eggs, 2 12 12 Bacon, 30 grams 5 15 Oleo or butter, 50 grams } Q 041 Lard or crisco, 45 grams / Meat, 120 grams 32 20 Hepco cakes, 2 . 12 12 61 100 FIRST DAY. Breakfast. Fried egg, 1, and bacon, 30 grams. Hepco cake, 1, and oleo, 15 grams. Coffee. Dinner. Boiled dinner. Corned beef, 80 grams. Cabbage, 150 grams. Oleo, 10 grams. Pickle. Hepco cake, 1, and oleo, 15 grams. Tea and coffee. 540 FOODS AND THEIR COMPOSITION Supper. Vegetable and compel beef hash with fried egg: Corned hoof, 40 grains. Cabbage, lf>0 grains. Oleo, 10 grains. Tea or coffee. SKCOXD DAY. Breakfast. Egg, 1; scrambled with tomato, 50 grams. Bacon, 1") grains. Ilepco cake, 1, and oleo, 15 grams. Tea or coffee.. Dinner. Hamburg steak, SO grams. Onions (30 grams > fried in 10 grams oleo, GO grams. Green.s, 90 grains, with egg, 1, and oleo, 10 grams. Hepco cake, 1, and oleo, 15 grams. Tea or coffee. Supper. Meat (liver), 40 grams, with bacon, 15 grams. Cold slaw, 100 grams (cabbage, vinegar, salt, pepper). THIRD DAY. Breakfast. Boiled egg, 1. Bacon, 30 grains. Ilepco cake, 1, and oleo, 15 grams. Coffee,'. Dinner. Boiled cod, SO grants, with oleo, 10 grams, and vinegar. String beans, 150 grams, and oleo, 10 grams. Hepco cake, 1, and oleo, 15 grains. Supper. Sardines, 40 grams, with hard-boiled egg, 1. Sauerkraut, 150 grams. Tea or coffee. PICNIC LUNCHES. FIRST DAY Dinner. Lister sandwich: 1 Lister roll, chicken, (JO grains, cucumber, 75 grams. 1 lard-boiled egg. Olives. Tea or coffee. Supper. Sardines, (JO grams. Lister roll and butter. Lettuce, radish, and celery, 75 grams. Ripe tomato, 50 grams. RECIPES USED IN PREPARING PRECEDING MENUS 541 SECOND DAY. Dinner. Sliced veal loaf sandwiches (1 Lister roll). Dressed cabbage, 75 grams. Custard (-J- egg). Coffee. Supper. Salad (cold halibut, egg, \, cucumber, 75 grams). Lemon or rhubarb jelly. Brazil nuts. THIRD DAY. Dinner. Cold lamb chop. Tomato. Olives and pickles. Lister cream puff. Supper. Salad: egg. Lister sandwich: Lister roll, cold bacon, lettuce. Coffee Bavarian cream. FOURTH DAY, Dinner. Egg baked in tomato with cheese on top. Ham sandwich: 1 Lister roll. Swiss chard. Coffee jelly. Supper. Sandwich: cold roast beef, 1 Lister roll, lettuce and horseradish. Rhubarb sauce. 542 FOODS AX1) THEIR COMPOSITION G. TABLES OF EQUIVALENTS. TABI.K 225. COMPARATIVE SCALES OF KILOCKAMS AND POUNDS, CENTIMETERS AND INCHES. (I am indebted to Mr. Edward Clark, of (he Massachusetts Institute of Technology, for the preparation of these scales.) Kilograms to pounds. OntimrtiTs t, > frr-t and inc Kg. Lb. Kg. Lb. Kg. Lb. Cm. ^_ hi. Ft. 20 -41 rM ~I -110 125- 200- ou . -43 -270 10 42 -105 120- 190- -75 - 41 115- 6 is -40 45~ -100 - 250 180 - : 70 -39 no- O" -210 : 17 its 170- - i 37 105- - 230 -65 30 90 i 16 35 W ~ 100- 220 1GO " - - -34 So 210 GO 5 95- 150 15 -33 ^_ 32 90- -200 - 11 -31 35- 140- 55 -190 -30 - 75 35 - -29 - 130 13 - _ ISO - 1 SO -50 : 28 iO _ -27 170 4 120 - 12 30 75 -26 - (15 ~ 100 - -45 - -25 70- - . - 110 _ 11 24 - (JO 150 - -23 (W - - 110 -40 1(1 -22 '~~>~ - 55 10(1 - _ -21 150- - - 130 _ - 3 ~ 20 - 50 DO * _ -35 _ - 55 ~ - _ 120 - -19 _ - - 32 18 - 15 8 20 50~ -110 80 ~ Kg. Lb. Kg. LD. Kg. Lb. Cm. In. Ft. EQUIVALENTS USEFUL IN DIABETIC WORK 543 EQUIVALENTS USEFUL IN DIABETIC WORK. 1 TABLE 226. 1 gram O 2 (T. = C. Tension, 760 mm. mercury) = 0.7 liters. 1 gram CO 2 (T. = C. Tension, 760 mm. mercury) = 0.5091 liters. 1 liter O 2 (T. =0 C. Tension, 760 mm. mercury) = 1.4285 gm. 1 liter CO 2 (T. =0 C. Tension, 760 mm. mercury) = 1.9642 gm. According to Zuntz 2 the quantity of oxygen which has been used and the quantity of carbon dioxide which has been produced in the combustion of protein represented by the excretion of 1 gram of urinary nitrogen is, at C. and a pressure of 700 mm. Hg, as follows: 1 gram urinary nitrogen equivalent to 8.44 grains O 2 , or 5.91 liters. 1 gram urinary nitrogen equivalent to 9.33 grams CO 2 , or 4.75 liters. 1.0 ounce 28.4 grams 1.0 pound 454.0 " 2.2 pounds 1.0 kilogram 1 fluidounce 29.6 cubic centimeters 1 quart 946.0 1000 grams 1.0 kilogram 1000 cubic centimeters 1.0 liter 1.0 meter 39. 37 inches 1000.0 meters 3281. 00 feet 1609.3 meters 5280.00 feet or 1 mile. 1 horizontal kilogram-meter requires expenditure of 0.0005 calorie. 426.5 vertical kilogram-meters represent the heat equivalent of 1 calorie. 'See also Tables 39, 40, 41, 42, 103, 105, 109, 110, 111, 112, 120, 124. Zuntz u. Schumburg: Physiologic des Marsches, Berlin, Hirschwald, 1901, p. 361 CASE INDEX. 1 Case I . Reason for not giving drugs 399 4. Severe acidosis . 158, 164 1500 c.c. alkaline solution temporarily brought out of coma . 397 7. Trauma 62 8. Tendency of glycosuria to increase 83 Volume of urine little index to severity 170 Negative carbohydrate balance due to stored carbohydrate 27.5 Prolonged use of high fat diet 278 Intermittent claudication preceded hemiplegia .... 418 " 10. Disadvantage of treatment with drugs; description of case . 400 11. Angina pectoris; sudden death 418 17. Apparent tolerance for milk 377 18. Gall-stones 295 29. Operation for appendicitis 342 " 30. Importance of examining mixed twenty-four-hour specimen of urine 169 38. Specific gravity not always high in presence of sugar . . 173 58. Gangrene of long duration after amputation 423 70. Blood-pressure fell with diet ; complication of gangrene . . 417 " 90. Success of treatment in diabetic of advanced years . . . 468 " 102. Successful pregnancies 449 " 105. Survived mumps 326, 403 " 106. Pregnancy with severe diabetes; healthy child .... 452 " 127. Temporary disappearance of sugar with removal of fibroid 296, 447 " 129. Temporary precedes permanent diabetes 286 " 135. Loss of weight on withdrawal of sodium chloride . . . 127 155. Mother of Case Xo. 203, in whom diabetes was apparently cured 52 " 173. Operation at seventy-three years, later cerebral hemorrhage with unusual recovery 445 " 181. Intermittent fasting; my first case so treated 314 Description of case, with dietary and urinary charts . . . 493 " 190. "Is tendency of untreated glycosuria to increase?" ... 84 " 194. Gain in tolerance when sugar-free 85 1 In the Case Index are not included references to those cases mentioned in Tables 30, 36, 49, 81, 82, 85, 90, 91, 100, 101, 108, 139, 140, 147, 148, 149, 174, 175, 180, 181, 183, 185, 195, 205, and 206, 35 ( 545 ) 540 CAKE INDEX Case 2015. Favorable influence of heredity 52 220. Unset of coma after Castro-intestinal attack 1C>4 " 234. Died with cardiac incompctcncy . 419 " 23"). Loss of energy in excretion of sugar and acid bodies . . . 147 Temporary precedes pennanent diabetes 280 24t>. Decrease in acidosis after oatmeal 374 " 252. Coma upon restriction of diet . 1(14 " 289. Decrease in acidosis after oatmeal . . 37/5 2 l .)f>. Large amount of urine and sugar ... .... 1/57 Exceptionally long duration 4(>3 " 304. Coma following active exercise 21)1 307. Pregnancy with severe diabetes and albumin and casts; patient survived, child lost 452 309. Call-stones 295 Pregnancy case with small amount sugar 448 310. Typical onset of coma 1C>4 318. Pregnancy case; small amount of sugar: favorable outcome 44!) 332. KlTect of levulose and oatmeal on respiratory quotient . . 143 " 340. Volume of urine little index to severity . . ... 170 343. Loss of both legs from gangrene 423 344. Loss of energy in excretion of sugar and acid bodies . . lf)8, 253 Excretion of ammonia 189 Minus carbohydrate balance . . . . 27t) Carbohydrate balance 1 and acidosis in relation to oatmeal days .... 282, 374 Disappearance of acidosis with development of tuberculosis 2!)7 Diabetes and tuberculosis 40!) " 347. Renal involvement 170 Left hospital in good condition; coma shortly after . . 304 Arteriosclerosis 41!) " 348. Operation for removal of prostate 70, 442, 447 Constipation relieved by sawinii wood 434 "Is tendency of untreated glycosuriu to increase?' 1 . . 84 Improvement with diet 2!) KiTect of development of nephritis 53 Cardiac incompetcncy . . .... 41!) 355. Angina pectoris 311 Cangrene 423 35t>. "Is tendency of untreated glycosuna to increase?" . . 84 371. Decrease in acidosis after oatmeal . .... . 374 43C>. Niece of patient with marked heredity 47 Description of case; notable increase in tolerance for carbo- hydrate 3 Estimated more comfortable upon reduced diet .... 254 " Hoi. Large volume of urine; description of case 171 " 1157. Sugar appeared with responsibility 290 " 11SS. Indirect harm from trauma 03 " 1191). Large volume of urine when sugar-free 170,283 Marked acidosis, but disappeared without alkalis . . 217, 390 Metabolism of protein while fasting 208 Intermittent fasting 314 Description of case treated with levulose . . . 3 1C), 379, 393 " 1199. Chart of child two years old 403 " 1209. Chart of child who conquered tendency to eat candy . . 402 " 1213. Description of case treated with levulose; girl of seventeen years 320, 384 " 1220. Not severe case, but continual excretion of a little sugar . 298 " 122S. Inadequate measure of fat tolerance 321 " 1233. Description of case treated with levulose 320, 385 Condition of kidneys as shown by phenolphthalein excretion 423 " 1237. Illustration of ambulatory treatment 300 Easily became sugar-free with protein at normal level . . 311 " 1239. Drainage of abscess of lung 320, 342 Changes in pulmonary epithelium in relation to alveolar air 239 " 1204. Tonsillitis 403 " 12i')7. Treatmentwithlevulo.se . .... 314 " 1287. Feels better if keeps under limit of carbohydrate tolerance . 325 INDEX. ABBREVIATIONS, 485 Abscess of lung, operation, recovery, 443 Acetone and diacetic acid, quantitative test for, 191, 196 Huppert-Mess- inger, 191 in blood, 214 Wishart's test for, 214 qualitative test for, 185 quantitative test for, 194 Acid bodies, elimination of source of supply, 392 mode of elimination of, 156 percentage relation to one another, 155 source of, 155 substances favoring combus- tion of, 392 intoxication, treatment of, 387 alkalis in, dangers of, 394 Acidosis, culmination of, in coma, 164 dependence upon fat in diet, 281 in diabetes, similarity to that in normal individuals, 156 during fasting of a fat woman, 152 effect of oatmeal cure on, 374 induced in normal individual, influence of, 111 in mild, severe and extreme cases of diabetes, 157 nature of, 148 qualitative tests for, 185 quantitative tests for, 186 relation to coma, 148 safeguards of body against, 161 treatment of, results of, 234 upon a fat-protein diet of a normal individual, 153 variation of, with age, 163 with condition of kidneys, 163 with onset, 163 Agar-agar, 533 Akoll, 525 Albumin, tests for, 200 Alcohol in coma, 393 in diabetes, 282 Alcohol, effect of, on body processes, 223 respiratory quotient for, 227 Alimentary glycosuria, 17 Alkalinity of blood, titratable, test for, 222 Alkalis, dangers of, 394 Allen's hypothesis of diabetes mellitus, 76" paradoxical law, 18 Almond paste, 528 Alveolar air, carbon dioxide tension, 231 Fridericia meth- od, 233 25 mm. Hg or less, 234 Marriott's method, 237 Ambulatory treatment, 500 Ammonia, method for determination of, 188 Anesthesia, 441 in diabetes, 70, 350, 431, 438 without subsequent acidosis, 432 Appendicitis, 446 death caused by, 352 Arteriosclerosis, diabetes mellitus and, 62, 413 Aspirin, 401 Assimilability of dextrose, 257 of galactose, 258 of inulin, 259 of lactose, 258 of levulose, 258 of maltose, 258 of saccharose, 258 of starch, 254 B BACOX, 248, 249 loss of weight during cooking of, 24S Bananas, 510 carbohydrate in. 266 in treatment of diabetes, 377 Bang's method of estimation of blood sugar, 204 Barley, 510 (551) INDEX Beef, 508 Benedict's qualitative test, 175 quantitative test, 180 Blood, acetone in, 214 Wishart test for, 214 alkalinity of, tit rat able test for,222 chlorides in, test for, 212 examination of, 202 hydrogen-ion concentration in, 221 lipoids in diabetes, 100 in health, 90 influence of carbohydrate bal- ance on, 104 fasting on, 103 in normals and diabetics com- pared, 100 test for, 200 non-protein nitrogen in, 208 plasma, carbon dioxide in, test for, 214 Van Slykes', 217 solids of, total, 223 sugar, as clue to character of disease, 91 during treatment, 93 effect of dextrose on, 257 estimation of Bang's method, 204 Lewis-Benedict method, 203 influence of carbohydrate bal- ance on, 91 duration of diabetes on, 90 excretion of sugar in pre- ceding twenty-four hours on, 91 in severe diabetes, 92 threshold, 91 Van Slyke-Fitz method, 194 Body protein, storage of carbohydrate as, 139 Boiled dinner, 535 |3-oxybutyric acid in blood, 190 quantitative tests for, 190 Black's, 190 Shaffer's, 190 diacetic and acetone in urine and blood, 194 Bran biscuits for constipation, 531 Bread, 510 bran, 505, 531 carbohydrate in, 207 Breads, gluten, 500 hard, 523 light, 500 soft, 522 substitutes for, 505 Breakfast foods, 527 Bright's disease, diabetes mellitUS and, 419 Broths, 271, 284 Broths, analyses of, 272 Buckwheat flour, 510 Butter, 509 Buttermilk, 510 CALORIC needs of children, 243 in diabetes, 251 of diabetic, 321 in health, 240 values which every doctor should know, 249 Calories, lost, 322 Cancer, deaths from, in 1910, 342 Caramel, 401 Carbohydrate, assimilability of, 254 balance, 141, 275 in bananas, 200 in bread, 207 content of foods, 511 in diabetic diet, estimation of, 259 effect of exercise on, 144 in prevention of coma, 393 in fruit, 204 in grapefruit, 200 increased assimilability of, in absence of fat, 273 of protein, 273 influence of rate of administration of, 138 in lemons, 200 in milk, 207 in nuts, 204 in oatmeal, 207 in oranges, 204 possible storehouses for, 139 in potatoes, 203 in relation to weight, 141 respiratory quotient for, 220 following ingest ion of, 142 in ripe olives, 2(50 storage of, as blood sugar, 130 body protein and, 139 as glycogen, 135 tolerance for, determination of, 317 improvement in, 391 utilization of, 140 in vegetables, 259 Carbohydrates, gain in tolerance for, 84 influence of, on weight, 125 storage of, in diabetes, 134 Carbon dioxide in blood plasma, test for, 214 tension of alveolar air, 231 Carbuncles in diabetes, 429 multiple, death caused by, 353 Cardiorenal complications, death caused by, 302 Casoid, 523 Casts, 200 "showers" of, 201 INDEX 553 Cataracts, 438 Cereal, coffee infusion, 511 Cesarean section. 453, 457 Chart, dietary and urinary, 481 history, diabetic, 480 Cheese, 509 Children, diabetes in, 52, 458 fat requirement of, 461 healthy, nitrogen and salts in urine of, 460 protein requirement of, 460 Chlorides in blood, test for, 212 in urine, test for, 201 Chocolate, 511 and cocoa, 529 Cholesterol, 96, 99 and blood sugar compared, 107 esters, 97 Clams, 509 Classification for prognosis, 299 for treatment, 297 Cocoa, 511 cracked, 532 Coma, alcohol in, 393 carbohydrate in prevention of, 393 deaths from, in 1916, 342 with, causes of, 69 ether, 70 impaired kidneys, 70 infections, 71 influence of fat-pro- tein diet in, 72 mental excitement, 72 rapid loss of body fluid, 73 without, causes of, 65 cancer, 67 cardiorenal, 67 inanition, 68 infections, 68 pernicious anemia, 67 suicide, 67 tuberculosis, 67 vascular, 67 liquids in, importance of, 397 loss of weight during, 129 prior to, 129 relation of, to acidosis, 148 treatment of, results of, 388 of threatening, 387 Conjugal diabetes, 49 Conservation of energy in a diabetic, 398 Constipation, bran biscuits for, 531 in diabetes, 433, 475 exercises for, 434 Cornmeal, 510 Crabs, 509 Crackers, 510 Cream, 510 Crockery, diabetic, 251 DANGER zone, diabetic, 303 Death, average age at, in United States, 25 with coma, causes of, 69 without coma, causes of, 65 Deaths, causes of, in 191(5, 338 Dextrose, antidiuretic, in health, 19 assimilability of, 257 diuretic, in diabetes, 19 effect of, on blood sugar, 257 nitrogen ratio, 130, 274 in excess of 3.65 to 1, 133 Diacetic acid, qualitative test, for, 185 Diabetes mellitus, alcohol in, 282 Allen's hypothesis of, 76 apparent increase of, due to more accurate statistics, 25 appendicitis and, 446 arteriosclerosis and, 62, 413 associated with organic dis- eases, 53 better understood, 31 blood lipoids in, 100 sugar in, 92 blood-pressure and, 413 Blight's disease and, 419 caloric needs in, 251 carbohydrates in, storage of, 136 carbuncles in, 429 cataracts in, 438 in children, 458 complications of, treatment of, improved, 31 conjugal, 49 constipation in, 433, 475 curability of, 51 danger zone of, 303 death with coma in, causes of, 69 without coma, causes of, 65 diagnosis of, early, 29, 287 diarrhea in, 435 diet in, carbohydrate in, 254 estimation of, 259 fat in, 276 protein, 267 dietary excesses as precursors of, 59 duration fifteen years or more, 468 of life in, 32 et aeq. statistics of, Boston, 33, 34 Joslin's, 35 Massachusetts General Hos- pital, 32 Naunvn, 35 554 INDEX Diabetes mellitus, duration of life in, si at ist irs of, von Xoor- den, 35 Ions, 358 of elderly people, 75 emotional elements in, 01 ether anesthesia in, 431 etiology of, 53, (K) multiple, 60 eyesight in, 437 fatal eases of 1916, 336 food values in, importance of, 250 furuneulosis in, 428 gangrene and, 423 glyeosuria most trustworthy symptom of, 148 gout and, (>4 heredity and, 43, 2S7 appearance in child be- fore development in parent, 44 favorable influence of, 47 in notable diabetic fami- lies, 46 hygiene and mental, 2(iO physical, 289 hypophysis and, 63 imperfect supervision of, re- sults of, 349 incidence of. 19 in Berlin, 22 in Boston, 22 in England, 23 Magnus-Levy's table of, 23 in Xe\v York City, 22 in Paris, 23 in Tokyo, 24 in United States, 20 in Wales, 23 increasing seventy of, cause of death, 349 infectious diseases and, 62, 288, 402 islands of Langerhans and. 76 in .Jewish race. 49 Joslin's plan of campaign against . 304 lime deficiency in. 2Sf) liquids in, 2S3 liver and, 63 metabolism in, 109 decrease in. 1 12 mild, management of, 324 Naunyn's conception of, 74 neurit is in, 436 nitrogen in, 209 obesity in, 53, 2SS in old age, 466 onset of, frequency of, by de- cades, 2S Diabetes mellitus, organic, 76 pancreas and, 63 pathological anatomy of, 76 physiology of, 7(5 pneumonia and, 403 pregnancy and, 289, 448 production of, 76 prognosis of, classification for, 299 prophylaxis in, 286 pruritus in, 427 pulse in, 121 renal, 63 respiratory quotient in, 117 salt in, 283 severe, management of, 1524 sex and, 41 skin in, care of, 427, 474 sodium chloride in, 283 syphilis and, 62, 292 duration of, 294 teeth in, care of, 431, 474 tendency congenital, 74 to increase of, 82 trauma and, 62 treatment of, 286 abandoned, results of, 347 ambulatory, 500 illustration of, 306 bananas in, 377 classification for, 297 complications of, im- proved, 31 dietetic. 299 general discussion of, 299 special. 364 summary of, 30") drugs in, 399 fasting in, 31 1 days in, 364 intermittent, 313 case of, 492 preparation for, 307 unsuccessful cases, 350 follo\v-up methods in. 327 glycerin in, 401 hediosit in, 402 hospital. 493 improvement in, 29 levulose in, experiments with, 37S of milk in, 299 " milk cure " in, 377 oatmeal, 367 pancreatic preparations in, 294 potatoes in, 376 rectal injection of sugar in. 385 rest in, 289 INDEX 000 Diabetes mellitus, treatment of, saccharine in, 401 surgery and, 295 surgical, 438 elements favoring success, 439 predisposing to failure, 438 results of, 445 tolerance for carbohy- drate in, 317 for fat in, 320 for protein in, 318 vegetable days in, 367 wheat in, 376 tuberculosis and, 404 untreated, deaths from, in 1916, 342 uterine fibroids and, 447 untreated, deaths from, in 1916, 342 Diabetic, caloric needs of, 321 conservation of energy in, 398 crockery, 251 foods, 503 composition of so-called, 519 menus, 530 muffins, 531 receipts, 530 Diabetics above age of seventy, course of, 467 Diarrhea, 435 Dietary and urinary chart, 481 excesses as precursors of diabetes, 59 Diet, composition of, 243 diabetic, carbohydrate in, 254 estimation of, 259 fat in, 276 protein in, 267 knowledge of more general, 30 management of patients', in a hos- pital, 498 method of reporting, 477 normal, carbohydrate in, 244 fat in, 247 protein in, 245 variations according to race, 245 Dietetic suggestions, 530 treatment, 299 special, 364 Diets, actual, to render patients sugar- free, 484 Doughnuts, 510 Drugs in treatment of diabetes, 399 Duration of life, increase in, 25 EDEMA, 124 Eggs, 509 Eggs, variations of, in weight, 249 Emotional elements, diabetes and, 61 glycosuria, 18 Energy, loss of, in urine, 147 Equivalents, table of, 542 useful in diabetic work, 543 Eskimo, 245, 279 Ether, death from, in 1916, 350 Exercise, effect of, 144 Expectation of life, 58 Eyeball (Riesman), 391 Eyesight, 437 FAST days, weekly, 323 Fasting, 311 cases unsuccessfully treated by, 350 days, 364 intermittent, 313 man, metabolism of, 149 preparation for (preparatory treat- ment), 307 variations in weight during first week of, 129 Fat, the "built in" fat of the tissues, 99 danger of, to diabetic, 281 diabetic diet, 276 in normal diet, 247 requirements of a child, 461 tolerance for, determination of, 320 value of, to diabetic, 277 Fat-protein diet, death caused by, 350 Fatal cases of 1916, 336 Fats, cheap, 280 influence of, on weight, 125 respiratory quotient for, 226 Fehling's qualitative test, 175 quantitative test, 179 Fermentation test, 178 Fibroid, uterine, temporary disappear- ance of sugar after removal of, 296 Fish, 509 protein in, 270 Fischer on sodium chloride retention, 128 Flours and meals, 510, 519 Folin's test, 176 Follow-up methods, 327 Food, American, composition of, 508 arranged according to carbohy- drate content, 260 carbohydrate content of, 511 of alcoholic beverages, 517 of animal products, 511 of breads, 513 of cereal breakfast foods, 513 of cereals, 512 of crackers, 513 550 IXDKX Food, carbohydrate content of fish and shellfish, 512 of flours, meals, etc., . r )12 of fruits and berries, 5 1. 1 ) canned, 510 of grapefruit, 510 of non-alcohol beverages, 510 of oranges, 510 of pastes, 513 of pastry, 513 of pickles and condiments, 515 of vegetables, canned, 513 dried, 513 fresh, 513 of wines, dry, 510 sweet, 517 diabetic, 503 composition of so-called, 510 variations in, 252 Fruits, carbohydrate in, 204 Furunculosis in diabetes, 428 G A LACTOSE, assimilation of, 258 Gall-stones, 00, 205 diabetes in association with, 205 Gangrene, diabetes and, 423 duration of diabetes preceding, 424 of life of patients with, 420 treatment of, results of, 427 Gelatin, 500, 530 Glucose, formation of, from protein, 208 tests for, 174 qualitative, 175 multiple, 177 quantitative, 178 tolerance for, 250 Glycerin, 401 in treatment of diabetes, 401 Glycogen, storage of carbohydrate as, 135 Glycosuria, alimentary, 17 diabetes and, distinction between, 18 most trustworthy symptom of dia- betes, 148 renal, 03 Glycuronic acid, 184 Gout, diabetes mellitus and, 04 Grapefruit, carbohydrate in, 200 H HKDTOSIT in treatment of diabetes, 402 Height and weight, mortality accord- ing to, 57 Hepcocakes, 531 Heredity, diabetes mellitus and, 43, 287 Hominy, 510 Hospital treatment, 403 Household measures, 250 Hydrogen-ion, concentration of in blood, 221 Hygiene, mental, diabetes and, 200 physical, diabetes and, 280 Hyperglycemia, 87 Hypophysis, diabetes mellitus and, 03 IMPERFECT supervision, results of, 340 Inanition, deaths from, 08, 338 in 1010, 338 Infections, deaths from, in 1010, 342 intercurreiit, death caused by, 352 Infectious diseases, diabetes mellitus and, 02, 288, 402 Insurance, 20, 27 Jnulin, assimilation of, 250 Islands of Langerhans, diabetes melli- tus and, 70 JEWISH race, diabetes in, 40 J>;slin's plan of campaign against dia- betes, 304 LABORATORY, importance of physi- cian's, 100 Lactose, assimilation of, 258 test for, 183 Lecithin, 00, 08 Lemons, carbohydrate in, 200 Levulose, assimilation of, 258 influence of, on respiratory quo- tient, 143 test for, 184 in treatment of diabetes, experi- ments with, .378 Lewis-Benedict method of estimation of blood sugar, 203 Life, average expectation of, in Ger- many, 20 insurance 1 , 20 diabetes discovered at exami- nations for, 107 policies issued in I'nited States, 30 rejections on account of gly- cosuria, 27 Lime deficiency in diabetes, 285 Lipoids in blood, test for, 200 Liquids in diabetes, 283 INDEX 557 Liquids, importance of, 397 Lister's cakes, 531 Liver, 508 diabetes mcllitus and, 63 Lobster, 509 Lunches, picnic, 540 Lung, abscess of, operation, recovery, 443 M MACARONI, noodles, etc., 510, 527 Magnus-Levy's table of incidence of diabetes mellitus, 23 Maltose, 184 assimilation of, 258 Meat, protein in, 270 Menus, diabetic, 530 inexpensive, 539 seven, for severe diabetics, with recipes for same, 535-538 Metabolism in diabetes mellitus, 109 decrease in, 112 of fasting man, 149 respiratorv, total, 227 Milk, 267, 509, 510 carbohydrate in, 267 "Milk cure'' in treatment of diabetes, 377 respiratory quotient for, 185, 227 substitutes for, 507 in treatment, 290 Miscellaneous products, 530 Muffins, diabetic, 531 Mussels, 509 Mutton, 508 N NEURITIS, 436 Nitrogen, determination of, by titra- tion, 211 in diabetes, 209 excretion of, 130 non-protein, in blood, 208 tests for, 198 Non-carbohydrate diets, adaptation of body to, 154 relation of, to acidosis, 154 Nurses, directions for, 476 Nuts, 510, 511, 528 carbohydrate in, 264 malted, 528 O OATMEAL, 510 carbohydrate in, 267 "Oatmeal (lay," nutritive value of, 373 influence of, on respiratory quo- tient, 144, 145 Oatmeal, plaoo of, in present-flay diabetic therapy, 375 treatment, 367 weight of, uncooked and cooked, 375 Obesity, diabetes mellitus and, 53, 288 prognosis in presence of, 343 Old age, diabetes in, 466 Olives, ripe, carbohydrate in, 266 Onset of diabetes, frequency of, by decades, 28 Opium in treatment of diabetes, 399, 400, 436 Oranges, carbohydrate in, 264 Oysters, 509 PANCREAS, diabetes mellitus and, 63 Pancreatic preparations in treatment of diabetes, 294 Paradoxical law, Allen's, 18 Peanut butter, 528 Pentose, test for, 184 Peter's test, 180 Physiological glycosuria, 17 Pickles, 510 Pie, 510 Pneumonia, diabetes mellitus and, 403 Polariscope test, 179 Pork, 508 Potatoes, carbohydrate in, 263 influence of, on respiratory quo- tient, 143, 144 in treatment of diabetes, 376 Poultry, 508 Pregnancy, conclusions, 454 diabetes and, 289, 448 mild cases, 448 severe cases, 449 increase in tolerance for carbohy- drate during, 455 Prognosis, classification for, 299 Prostate, removal of, 447 Protein in diabetic diet, 267 in fish, 270 formation of, from glucose, 268 in meat, 270 need of diabetic for, 267 in normal diet, 245 preparations, 522 requirements of a child, 460 respiratory quotient for, 226 tolerance for, determination of, 318 Pruritus in diabetes, 427 Pulse in diabetes, 121 compared with normal, 122 R RECIPES, diabetic, 530 Rectal injection of sugar, 385 558 INDEX Renal glyposiiria, (53 Respiration apparatus, clinical, 221 examination of, 223 examination of, technic of, 223 Respiratory metabolism, total, 227 quotient, 225 in diabetes, 1 17 during prolonged fast, 117 following ingcstion of carbo- hydrate, 142 for alcohol, 227 for carbohydrate, 220 for fats, 22(5 for milk, 227 for protein, 22(5 in health, 11") influence of levulose on, 1 13 of oatmeal on. 1 14, 11.") of potato on, 143, 144 non-protein, 229 theoretical, 11!) quotients, theoretical, calculated from diet, 230 Rice, ;")!() Rye flour, 5 10 S SACCHARINE in treatment of diabetes, 401 Saccharose, assimilation of, 258 Salt in diabetes. 2S3 Salt-free diet, 12S Sausage, 508 Scallops, 50!) Septicemia, death caused by, 353 Sex, diabetes mellitus and, 41 Shellfish, 509 Skin in diabetes, care of, 427, 474 Sodium bicarbonate;, avoid, 3SS dangers of, 394 influence of, on weight, 127 mode of, administrations of, 305 chloride in diabetes, 283 influence of, on weight, 12(i relent ion of, Fischer on, 128 Soups, 51 1 Specific gravity, 1 73 fixat ion of, 1 7 1 Squab. 535 Statistics, diabetic, book constructed to facilitate, 501 Sugar, assimilability of, 254 in relation to sugar in urine, 25(5 consumption of, in United States, 59 reappearance of, 323 rectal injection of, 385 substances in urine which give rise to confusion, 1S5 Sugar, tost for use by patients, 475 Surgery, diabetes and, 438 Surgical failure, elements predisposing to, 438 success, elements favoring, 43!) Symptoms, 391 Syphilis, death caused by, 3(54 diabetes mellitus and, 62, 092 TKKTH in diabetes, care of, 431, 474 extraction of, with subsequent death, 431 Trauma, diabetes mellitus and, (52 Traumatic glycosuria, 18 Treatment, abandoned, results of, 347 ambulatory, 500 illustration of, 30(5 bananas in, 377 classification for, 297 of complicat ions, 402 dietetic, general discussion of, 299 special, 3(54 summary of, 305 drugs in, 399 fasting days in, 3(54 variations in weight during first week of, 129 of gangrene, results of, 427 glycerin in, 401 hediosit in, 402 hospital, 493 levulose in, experiments with, 378 milk in, 299 '' milk cure" in, 377 oatmeal, 3(17 opium in, 395, 100, 43(5 potatoes in, 37(1 rest in, 289 saccharine in, 101 summary of, in three successive groups of cases, 328 body weight, loss of, 329 protein, loss of, 335 calories, total, 335 vegetable days in, 3(17 wheat in, 37(5 Tuberculosis, angina, pecforis and, 107 deaths from, in 191(5, 342 decrease in severity of diabetes in presence of, 40! I diabetes mellitus and 404 U I UINAUV examinations, cause's leading to, in diabetes, 1(5(5 inexpensive, 1(55 routine, 29, 1(55, 1(57. 1(58 INDEX 559 Urine, chlorides in, test for, 201 collection of, by patient, 475 examination of, to be made on each birthday, 287 mixed twenty-four-hour quantity of, 169 reaction of, 187 specific gravity of, 173 volume of, 169 Uterine fibroids, 447 VAX SLYKE'S test for carbon dioxide in blood plasma, 217 Vegetable days, 367 Vegetables, carbohydrate in, 259 influence of cooking on, 262, 263 thrice cooked, 533 washed, 261 W WATER content of bodv, 124 Weight, 55, 57, 123 of diabetic patients, 128 influence of carbohydrates on, 125 of fats on, 125 of sodium bicarbonate on, 127 chloride on, 126 loss of, in diabetic patients, 128 during coma, 129 prior to coma, 129 variations in, during first week of fasting treatment, 129 What every diabetic should know, 470 Wheat, old in treatment of diabetes, 376 Wishart's test for acetone in blood, 214 ZWIEBACK, 510 UNIVERSITY OF CALIFORNIA LIBRARY Los Angeles This book is DUE on the last date stamped below. 315 000021667