A comparison of periodontal disease among rural Amish and non-Amish adults


J Clin Periodontol 1994; 21: 386-390
Printed in Denmark . All rights reserved

Copyright © Murtksgaard 1994

Clinical pencdontBioiy
ISSN 0303-6979

A comparison of periodontal
disease among rural Amish and
non-Amish adults

Robert A. Bagramian\
Mahassen M. Farghaly\
Dennis Lopatin\
MaryFran Sowers^ Salam A. Syed^
and Julie L. Pomerville^
^School of Dentistry, University of Michigan;
^School of Public Health, University of
Michigan, Ann Arbor, Michigan 48109-1078

Bagramian RA, Farghaly MM, Lopatin D, Sowers MF, Syed SA and Pomerville
JL: A comparison of perio dont cd disease among rural Amish and non-Amish adults.
J Clin Periodontol 1994; 21: 386-390. © Munksgaard, 1994.

Abstract. Periodontal disease can be more efficiently studied within a homogene-
ous population where genetic influences and lifestyles are similar enough to negate
their effect on the disease process. This study focuses on an Amish population
in southern Michigan who isolate themselves from outside influences and their
non-Amish neighbors. A total of 425 Amish and 290 non-Amish were contacted
resulting in 330 Amish and 215 non-Amish who were examined in their homes
giving a participation rate of 76.2%. Ages ranged from 18 to 82 years. Prevalence
of periodontal disease tended to be higher among males and increased with age.
There were slightly more Amish females (52%) than non-Amish (49%). Means
of periodontal conditions for Amish were 1.35 mm for attachment loss, 2.59
mm for pocket depth, 0.24 for calculus, 0.77 for plaque and 0.74 for gingivitis.
For non-Amish, the means were 1.03 mm for attachment loss, 2.38 mm for
pocket depth, 0.40 for calculus, 0.95 for plaque and 0.87 for gingivitis. It is of
interest that the Amish do not practice routine oral hygiene. Only 36.8% of
Amish reported brushing at least 1 x a day compared to 84.6% of non-Amish.
Similarly, only 8.2% of Amish reported flossing at least 1 x a week compared to
40.8% of non-Amish.

Key words: Amish; periodontai disease;
prevalence; epidemiology; attachment ioss.

Accepted for pubiication 21 Juiy 1993

The Amish in this study isolate them-
selves from the outside world by living
without electricity, telephones, or auto-
mobiles and marrying within their
society. As reported previously (Bagra-
mian et al. 1992), the Amish maintain a
secluded lifestyle as a result of religious
values established in the 16th century in
Switzerland. The extent of isolation and
self-sufficiency of the Amish com-
munity is demonstrated by the existence,
of Amish schools, which children attend
to 8th grade. Only a few of the Amish,
8%, in this study completed high school
in public schools. Households of 10 or
more comprising extended family mem-
bers are common (McKusick 1980, Hos-
tetler 1980). This secluded population
provides a unique opportunity to inves-
tigate the impact of a variety of factors
on prevalence of periodontal disease.

Low levels of periodontal disease
found among the Amish in pilot studies

(Bagramian et al. 1988) formed the basis
for the present intensive study. Our pre-
vious paper assessed dental examin-
ations conducted on an Amish farming
community during the 2-year period
1989-90. The purpose of this paper is
to report periodontal disease data col-
lected for Amish (Â  = 330) and non-
Amish controls (Â  = 215) over the entire
3-year period, 1989-91, Assessments of
behavior, attitudes toward oral hygiene,
and knowledge of periodontal disease
are examined along with various meas-
ures of periodontal disease.

iVIethods

Amish farms are interspersed with non-
Amish farms in the area targeted for this
study. A map of the Amish community
identified Amish and non-Amish farms.
Letters were mailed to all area residents
informing them of our study and sol-

iciting participation. The 2 populations
were approached in as similar a fashion
as possible. For Amish respondents,
several steps were involved in gaining
entry into individuals' homes to conduct
the interview and exam. The 1st entry
point was the religious leaders called
Bishops. Each Bishop has responsibility
for apprixmately 30 Amish families and
has ultimate authority over access to
Amish. After receiving a Bishop's ap-
proval, the team was able to contact
individual families, where the 2nd entry
point was the head of the household.
Once the household head agreed, the
investigators were free to solicit partici-
pation of individual family members.

Since Amish participants were re-
cruited door-to-door, non-Amish were
also recruited door-to-door. Every non-
Amish house was approached to ensure
the sample was as similar to the Amish
as possible. Unhke the Amish, these in-



Periodontal disease and Amish 387

dividuals could be accessed directly. Sol-
icitations were made during the evening
as well as the daytime to avoid biasing
the sample.

A total of 425 adult Amish men and
women over 18 years of age were con-
tacted and of these, 330 were examined,
giving a 77.6% response rate. Of the 290
non-Amish contacted, 215 responded, a
response rate of 74.1%. All subjects
were reimbursed for participation and
no dental treatment was provided. Con-
ditions similar to the field were used
to calibrate 3 examiners to a reference
periodontist, and x-statistics between
each of the examiners were calculated
for level of agreement on attachment
loss and pocket depth. For attachment
loss the xs ranged from 0.64 to 0.70,
and for pocket depth, they ranged from
0.90 to 0.93, well within the limits in
other studies and national surveys
(Brown et al. 1989).

Electricity was not available in
homes, requiring examinations to be
performed with a battery operated head
light and portable dental chair. Meas-
urements were taken from 6 standard
sites of all existing teeth except 3rd mo-
lars: distal-buccal, mid-buccal, mesio-
buccal, disto-lingual, mid-lingual and
mesio-lingual. The following indices
were included: Silness & Loe (1964)
plaque index, Loe & Silness (1963) gin-
gival index, Ramijord (1976) calculus
index, and modified Ramfjord (1976)
PDI for pocket depth and attachment
loss measurements. A Hu-Friedy no. 11
periodontal probe was employed to
measure loss of periodontal attachment
and pocket depth to the nearest milli-
meter.

A 1-h, comprehensive questionnaire
was administered by a trained inter-
viewer. Items regarding demographic
characteristics, former health problems,
current medications, allergies, tobacco
use, frequency of physician visits, and
family oral health history provided an
overall picture of the respondents' gen-
eral health. Oral health attitudes and
practices were probed through a series
of questions on the respondent's oral
health history, recent dental care, daily
oral hygiene habits and satisfaction with
dental services. The 23-page instrument
was pre-tested on adult patients at the
University of Michigan School of Den-
tistry over a 3-week period. Internal
validity was checked through repetitive
questions with altered wording or a re-
verse response and modified.

A knowledge score was computed

based on 8 questions that evaluated the
respondent's ability to define peri-
odontal disease. An index of oral hy-
giene was also created from 4 questions
dealing with brushing, flossing, rinsing,
and periodontal disease.

Severity of periodontal disease was
examined to identify differences for
Amish compared to controls by ident-
ifying the proportion of people with ^ 1
site at various disease levels (Kingman
et al. 1988) and by assessing whole-
mouth means. Disease was delineated
as having ^ 1 tooth with 4 mm of
attachment loss or 5 mm of pocket
depth or a mean greater than 1.5 mm
attachment loss or 3.0 mm pocket
depth.

Logistic regression was applied to de-
termine the relationship between ex-
planatory variables: age, gender, Amish
or non-Amish, oral health care; knowl-
edge of periodontal disease, whether or
not the person had mean plaque or
mean gingivitis scores > 1 or mean cal-
culus scores > 0 ; and presence or ab-
sence of disease. Each explanatory vari-
able was entered into the model sequen-
tially to determine the extent of any
additional impact on the model ability
to predict the probability of disease. In-
teractions between explanatory vari-
ables were also entered into the model.
Variables that did not show a significant
dependent relationship with indepen-
dent variables, as evidenced by the p-
values for the Wald / - statistic, were
dropped from the model. This iterative
process resulted in identification of ex-
planatory variables that significantly re-
late to periodontal disease. Odds ratios
were calculated.

Results

Amish respondents were younger (81%
under 45) than non-Amish (54% were
under 45) and less educated; 8% had a
12th grade education compared to 50%)
of non-Amish. There were slightly more

female Amish (52%) than males (48%).
Gender for non-Amish was almost
evenly spHt with 49%o female and 51%
male.

A larger proportion of non-Amish
had a full dentition at all age levels ex-
cept for ages 25-34 years (Tables 1, 2).
While 64%o of Amish age 18-24 years
had a full dentition, 78%o of non-Amish
at this age had all of their teeth. Simi-
larly, only 10%o of Amish over 55 years
had a full dentition while 19%) of non-
Amish of this age had all of their teeth.
The mean number of teeth present,
however, differed little except for the
age group 45-54 years, where the mean
was 19.2 for Amish and 24.2 for non-
Amish.

Mean attachment loss for Amish was
1.35 mm while mean pocket depth was
2.59 mm. For non-Amish, the means
were slightly lower: 1.03 mm for attach-
ment loss and 2.38 for pocket depth.
Mean calculus scores were low with an
overall mean of 0.24 for Amish and 0.40
for non-Amish. Plaque and gingivitis
means showed somewhat higher levels
but were still relatively low: 0.77 and
0.74 for Amish and 0.95 and 0.87, re-
spectively, for non-Amish.

Among Amish, females had lower
attachment loss levels than males at
every age except the youngest. Gingi-
vitis and plaque showed similar age- and
gender-related increases, while the
change in pocket depth levels was not
as dramatic. For Amish, attachment
loss rose from 0.96 mm at 18-24 years
to 2.18 mm at age 54 years and above
(Table 3). For non-Amish, attachment
loss levels increased consistently with
age from 0.68 mm at 18-24 years to
1.69 at age 54 and above (Table 4).

The distributions for oral hygiene
practices and knowledge of periodontal
disease showed significant differences
between Amish and non-Amish respon-
dents. Given their secluded life style, the
Amish performed as expected, scoring
lower than non-Amish on the identifi-

Tabie 1, % distribution of numbers of teeth and mean number of teeth by age among Amish

Age group
(years)

18-24 25-34
No. Teeth

35-44 45-54 55 +
iN=40)

5-10
11-19
20-27
28 +

mean no. of teeth

0.0
0.9

35.1
64.0

27.3

0.0
0.0

48.8
51.2

26.6

2.9
11.6
65.2
20.3

23.6

14.3
21.4
60.7
3.6

19.2

10.0
17.5
62.5
10.0

20.8



388 Bagramian et al.

Table 2. % distribution of numbers of teeth and mean number of teeth by age among
non-Amish

No. teeth

5-10

2 0 - 2 7 o : ^ , v . : : : • • : ; • . ,

28 +

Mean no teeth

18-24
(7V=23)

0.0
0.0

21.7
78.3

27.6

25-34

0.0
2,1

53,2
44,7

26.7

35-44

2.1
4 3

59,6
34,0

25,2

45-54
(TV=40)

2.5
7.5

70.0
20.0

24.2

55 +
(/V=58)

12,1
15,5
53.4
19.0

21.6

Table 3. Mean periodontal disease measures by age among Amish

Index

plaque
calculus
gingivitis
pocket depth
attachment loss

18-24
( y v = i i i )

0.72
0.17
0.66
2.54
0.96

25-34
(yV=82)

0,72
0.19
0,74
2,64
1,28

Age group
(years)

35-44
(A'=69)

0,74
0,22
0,77
2,68
1,48

45-54
(A^=28)

0,97
0.41
0.85
2,59
1,78

55 +
(7V=4O)

1,00
0,44
0.86
2,54
2,18

Total
(A^=545)

0,77
0.24
0.74
2,59
1,35

Table 4. Mean periodontal disease measures by age among non-Amish

Index

plaque
calculus
gingivitis
pocket depth
attachment loss

18-24
(7V=23)

0.65
0.11
0.65
2.42
0.68

25-34
(yV=47)

0,88
0,35
0,88
2,48
0,78

Age group
(years)

35-44
(A'=47)

1,00
0,46
0,93
2,45
0,85

45-54
(/V=40)

0,84
0.34
0,75
2,29
0,89

55 +
(yV=58)

1,12
0,47
1,04
2,43
1.69

Total
(7V=545)

0.95
0.40
0.87
2.38
1,03

Table 5. % distribution for attachment loss of ^ 4 mm by age for Amish and non-Amish

Attachment loss
Amish Non-Amish

Age
(years)

mm mm mm mm

"A, N N N

18-24
25-34
35-44
45 +

73
30
13
9

65,8
36,6
18,9
13,2

38
52
56
59

34,2
63,4
81,1
86,8

19
27
24
34

86,4
57,5
51,1
34,7

4
20
23
64

13,6
42,5
48,9
65,3

total 125 37,9 205 62,1 104 48,4 11 51,6

cation and prevention of periodontal
disease. 35% of the non-Amish sample
provided the correct response to 8
knowledge questions while only 10% of
the amish could. Over -/^ of Amish indi-
cated incorrectly that cavities were in-
dicative of periodontal disease and that
widening of spaces between teeth was
not an indicator. It is important to point
out that this lack of knowledge is evi-
dent in the reported oral hygiene prac-
tices of Amish, who show low behav-

ioral scores when asked about actual
practices.

Table 5 shows that the proportion of
persons with 4 mm or more of attach-
ment loss rose with age, as expected. A
greater proportion of Amish had 4 mm
of attachment loss at every age level.
The overall difference in the proportion
of respondents with ^ 1 tooth at the 4
mm level was 62% for Amish and 52%
for non-Amish.

Logistic regression showed a high risk

for attachment loss of 4 mm for respon-
dents who were Amish, male, older, had
calculus, and knew little about the signs
of gum disease (/?<0.05, Table 6). For
pocket depth of 5 mm, the best-fitting
model includes Amish, older, male,
having calculus, and having gingivitis as
significant risk factors. The odds ratios
presented in Table 6 show age as being
the highest risk factor of 4 mm of attach-
ment loss. For pocket depth, the greatest
risk is among those with gingivitis.

Behavior, plaque, and interactions
between the independent variables did
not reach statistical significance in the
final model for attachment loss or
pocket depth. However, when factors
related to having gingivitis, calculus, or
plaque were examined, interactions and
behavior were important. The signifi-
cant risk factors for plaque are being
Amish, older, having little knowledge
about the signs of periodontal disease,
and the interaction between Amish and
behavior (Table 7). Significant factors
related to having gingivitis are being
Amish, older, having little knowledge
about the signs of periodontal disease,
behavior, and the interactions between
Amish and knowledge of the signs of
periodontal disease. No factors were
significantly related to having calculus.

When whole mouth means were
examined, being Amish, age, and gender
(male) were significantly related to
attachment loss. These 3 variables con-
stituted the entire set of independent
variables for the final model. For mean
pocket depth, having gingivitis, being
Amish, age, and male were significant
risk factors.

Discussion

The Amish settlements in Michigan
present an exceptional opportunity to
study demographic factors and behav-
iors associated with periodontal disease.
Amish homes are interspersed among
non-Amish, thus reducing environmen-
tal influences not related to genetics or
life-style on the disease process. The
Amish and controls in this study were
predominantly farmers.

The examination of different meas-
ures of periodontal disease in this study
indicates that an assessment of severity
when measured as the presence of ^ 1
site at a specific disease level, yields
more information than an assessment of
whole-mouth means (Burt 1990). Only
Amish, gender, and age were significant-
ly related to mean levels of attachment



Periodontal disease and Amish 389

Table 6, Odds ratios for periodontal disease

Amish
gender
age (years)*

25-34
35-44
45-t-

Plaque
calculus
gingivitis
behavior
disease knowledge

Attachment

odds ratio

2.27
2 J 2

3.82
9.12

15.80
1.72
2.20
1.36
1.09
1.22

loss ̂ 4 mm

P

0.0009
0.0001

0.0001
0.0001
0.0001
0.06
0.01
0.28
0.28
0.002

Pocket d e p t h s

odds ratio

1.72
1.65

2.29
1.88
2.08
1.75
2.32
3.00
N / A
N / A

5 mm

P

0.01
0.01

0.005
0.03
0.01
0.07
0.003
0.0005
N / A
N/A

Reference group was respondents 18-24 years old.

Tabie 7, Odds ratios for plaque and gingivitis

Amish
gender
age (years)*

25-34
35-44
45-f

behavior
knowledge score
interactions:

Amish / behavior
Amish/knowledge

Plaque

odds ratio

7.1
1.27

1.05
1.22
1.95
1.68
1.43

1.39
1.17

P

0.02
0.20

0.80
0.50
0.01
0.03
0.05

0.03
0.21

Gingivitis

odds ratio

13.46
1.51

1.25
1.12
1.32
1.63
2.05

1.32
1.42

P

0.002
0.03

0.40
0.70
0.30
0.04
0.001

0.06
0.01

Reference group was respondents 18-24 years old.

loss and pocket depth, whereas severity
measured at ^ 1 site at a specific level
showed additional factors as being sig-
nificant predictors.

Amish respondents reported lower
education levels, scored lower on
knowledge of signs of periodontal dis-
ease, and reported less oral health care
than non-Amish. The difference in dis-
ease levels, although statistically signifi-
cant, is clinically small. Overall, Amish
respondents had only 0.32 mm more
attachment loss and 0.21 mm rnore
pocket depth than non-Amish. Differ-
ences in plaque, gingivitis, and calculus
were also small.

Both measures of periodontal dis-
ease, attachment loss of 4 mm or pocket
depth of 5 mm, assessed among Amish
and non-Amish respondents, demon-
strated the expected relationships with
age and gender. The odds of having
attachment loss of 4 mm were higher
among persons with calculus and those
who knew httle about the signs of peri-
odontal disease. The highest odds ratio
for pocket depth of ^ 5 mm was for
persons with gingivitis. This may be due
to inflammation that occurs with gingi-

vitis. A pocket may be measured deeper
by the examiner if the area around the
tooth is swollen due to gingivitis. Calcu-
lus showed the 2nd highest odds ratio
for pocket depth. Calculus is formed
over a longer period of time than plaque
and gingivitis; thus, it is not surprising
that calculus was the only measure of
the 3 that was a strong predictor of
attachment loss and pocket depth.

The high odds ratio between having
gingivitis or plaque and being Amish
reflect the low behavior score among
the Amish (Table 7). It is interesting to
note that behavior was not statistically
significant for attachment loss or pocket
depth and was significant in the plaque
and gingivitis models. Levels of plaque
and gingivitis in the mouth are affected
by brushing or flossing, whereas the
connection between oral hygiene and
pocket depth or attachment loss is not
as straightforward; these results are thus
in line with current thought that con-
siders plaque and gingivitis separate
entities from the more advanced disease
of attachment loss (Williams 1990).
Oral health care and knowledge about
the signs of periodonta] disease were

significant risk factors for having
plaque along with the interaction effect
of oral health care and being Amish.
The gingivitis model was similar with
being Amish, male, high behavior score,
high knowledge score, and the interac-
tion between being Amish and knowl-
edge of periodontal disease as signifi-
cant.

The modest differetices between Am-
ish and non-Amish oral health were in-
teresting given the dramatic differences
in oral hygiene practice and knowledge
of periodontal disease. A lack of knowl-
edge may well lead to a lack of dental
care which is expected to lead to peri-
odotital disease and missing teeth. This
study of oral health indicates Amish
have slightly more disease than non-
Amish, which is in line with the reported
lack of knowledge of periodontal dis-
ease and poor oral hygiene practices
among Amish.

Acknowledgement

This work was supported by the Na-
tional Institute of Dental Research
Grant 5R01 DEO8571-01.

Zusammenfassung

Ein Vergleich von Parodontalerkrankungen
zwischen Erwachsenen der landlichen Amisch-
und Nicht-Amisch Bevolkerung
Parodontalerkrankungen konnen besser in-
nerhalb einer homogenen Population unter-
sucht werden, in der genetische Einflusse und
Lebensgewohnheiten ahnlich genug sing, um
keine Auswirkung auf den Erkrankungspro-
zeB zu haben. Diese Studie konzentriert sich
auf eine Amisch-Bevolkerung im Stiden Mi-
chigans, die sich selbst auBeren Einflussen
gegeniiber isoliert und ihren Nicht-Amisch
Nachbarn. Insgesamt wurde mit 425 Amisch
und 290 Nicht-Amisch Kontakt aufgenom-
men. Das Ergebnis war, daB 330 Amisch und
215 Nicht-Amisch bei sich zuhause unter-
sucht wurden, wodurch sich eine Teilnahme-
rate von 76.2% ergab. Das Alter lag zwischen
18 und 82 Jahren. Die Pravalenz der Par-
odontalerkrankung tendierte dazu, bei Man-
nern hoher zu sein und im Alter zuzunehmen.
Es gab etwas mehr weibliche Amisch- (52%)
als Nicht-amisch (49%) -Teilnehmerinnen.
Die Mittelwerte fiir den parodontalen Zu-
stand ftir die Amisch-Bevolkerung waren
1.35 mm ftir Attachmentverlust, 2.59 mm fiir
Taschentiefe, 0.24 fur Zahnstein, 0.77 fiir Pla-
que und 0.74 fiir Gingivitis. Fiir die Nicht-
Amisch-Bevolkerung waren die Mittelwerte
1.03 mm fiir Attachmentverlust, 2.38 mm fur
Taschentiefe, 0.40 ftir Zahnstein, 0.95 fiir Pla-
que und 0.87 fur Gingivitis. Es ist von Inter-
esse, daB die Amisch keine regelmaBige
Mundhygiene durchfuhren. Nur 36,8% der



390 Bagramian et al.

Amisch berichteten, daB sie wenigstens ein-
mal pro Tag putzen, im Vergleich zu 84.6%
der Nicht-Amisch. Auch berichteten nur 8.2
der Amisch, daB sie Zahnseide wenigstens
einmal pro Woche verwenden, im Vergleich
zu 40.8% der Nicht-Amisch.

R e s u m e , - • . '•;, •' ••• • • . • ; - . " " > ' ^ .

Comparaison de la maladie parodontale chez
des adultes dans des populations rurales Amish
et non-Amish
L'etude de la maladie parodontale peut etre
plus efficace si elle est faite au sein d'une
population homogene ou les styles de vie et
les influences genetiques sont suffisamment
uniformes pour annuler Ieur effet sur le pro-
cessus pathologique. La presente etude est
centree sur une population Amish du sud du
Michigan qui s'isole des influences exterieu-
res et de leurs voisins non-Amish. On a pris
contact avec au total 425 Amish et 290 non-
Amish, ce qui a permis d'examiner 330 Amish
et 215 non-Amish a Ieur domicile, obtenant
ainsi un taux de participation de 76.2%. Leur
age allait de 18 a 82 ans. La prevalence de la
maladie parodontale tendait a etre plus elevee
chez les sujets du sexe masculin et augmentait
avec l'age. II y avait un peu plus de sujets du
sexe feminin Amish (52%) que non-Amish
(49%). Les moyennes des parametres paro-
dontaux chez les Amish etaient 1,35 mm pour
la perte d'attache. 2,59 mm pour la profon-
deur des poches, 0,24 pour le tartre, 0,77 pour
la plaque et 0.74 pour la gingivite. Chez les
non-Amish. ces moyennes etaient 1.03 mm
pour la perte d'attache, 2,38 mm pour la
profondeur des poches, 0,40 pour le tartre,
0,95 pour la plaque et 0,87 pour la gingivite,
II est interessant de noter que les Amish ne
pratiquent pas de soins reguliers d'hygiene
bucco-dentaire, Chez les Amish, 36.8% seule-
ment disaient se brosser les dents au moins
une fois par jour, contre 84.6'Mi chez les non-
Amish. De meme, 8.2'/o des Amish seulement
disaient utiliser le fil dentaire au moins une
fois par semaine. contre 40,8'M) des non-
Amish,

References

Bagramian, R. A,, Farghaly, M,, Lopatin, D,
E., Sowers, M. F., Syed, S. A. & Pomervil-
le, J, L. (1992) Periodontal disease in an
Amish population. Journal of Clinical Peri-
odontology 20, 269-272.

Bagramian, R. A., Narendran, S. & Khavari,
A. M. (1988) Oral health status, knowl-
edge, and practices in an Amish popula-
tion. Journal of Public Health Dentistry 48,
147-151,

Beck, J, D , Lainson, P A., Field, H. M. &
Hawkins, B. F. (1984) Risk factors for vari-
ous levels of periodontal disease and treat-
ment needs in Iowa. Community Dentistry
and Oral Epidemiology 12, 17-22.

Brown, J. L., Oliver, R. C. & Loe, H. (1990)
Evaluating periodontal status of US em-
ployed adults. Journal American Dental
Association 121, 226-232.

Brown, L. J., Oliver, R. C. & Loe, H. (1989)
Periodontal diseases in the U,S, in 1981:
prevalence, severity, extent, and role in
tooth mortality. Journal of Periodontology
60, 363-370,

Burt, B. A., Ismail, A. I., Morrison, E, C, &
Beltran, E, D (1990) Risk factors for tooth
loss over a 28-year period. Journal of Den-
tal Research 69, 1126-1130.

Epidemiology and Oral Disease Prevention
Program N (1988) Oral health of United
States adults: the National survey of oral
health in US employed adults and seniors,
1985-86: regional findings. NIH publi-
cation no, 87-2869,

Epidemiology and Oral Disease Prevention
Program N (1987) Oral health of United
States adults: the National survey of oral
health in US employed adults and seniors,
1985-86: national findings, NIH publi-
cation no, 87-2868.

Hostetler, J. A. (1980) Amish society, 3rd edi-
tion, Baltimore, MD: John Hopkins Uni-
versity Press.

Ismail, A. I., Morrison, E. C , Burt. B. A,,
Caffesse, R, G, & Kavanagh, M, T. (1990)

Natural history of periodontal disease in
adults: fmdings from the Tecumseh peri-
odontal disease study, 1959-1987. Journal
of Dental Research 69, 430-435.

Kingman, A., Morrison, E., Loe, H. &
Smith, J. (1988) Systematic errors in esti-
mating prevalence and severity of peri-
odontal disease. Journal of Periodontal Re-
search 59, 707-713.

Loe, H., Age, A., Hans, B. & Martyn, S.
(1978) The natural history of periodontal
disease in man. The rate of periodontal
destruction before 40 years of age. Journal
of Periodontology 49, 607-620.

Loe, H., Anerud, A., Boysen, H. & Morrison,
E. (1986) Natural history of periodontal
disease in man: rapid, moderate and no
loss of attachment in Sri Lankan laborers
14 to 46 years of age. Journal of Clinical
Periodontology 13, 431-445,

Loe, H. & Silness, J. (1963) Periodontal dis-
ease in pregnancy (I). Prevalence and se-
verity. Acta Odontologica Scandinavica 21,
533-551,

McKusick, V, A, (1980) The Amish. Endeavor
4, 52-57.

Ramijord, S. P (1976) The periodontal dis-
ease index (PDI). Journal of Peri-
odontology 2>^, 602-611,

Sillness, J. & Loe, H. (1964), Periodontal dis-
ease in pregnancy (II). Correlation be-
tween oral hygiene and periodontal con-
dition. Acta Odontologica Scandinavica 22,
121-135,

Williams, Ray C, (1990) Periodontal disease.
The New England Journal of Medicine 322,
373-381.

Address:

R. A. Bagramian
Department of Periodontics/Prevention

and Geriatrics
School of Dentistry, University of Michigan
Ann Arbor, Michigan 48109-1078
Supported by National Institue of Dental
Research. 5ROI DE08571-01
USA