key: cord-032979-jle66lmn authors: ROMMEREIM, D. N.; ROMMEREIM, R. L.; SIKOV, M. R.; BUSCHBOM, R. L.; ANDERSON, L. E. title: Reproduction, Growth, and Development of Rats during Chronic Exposure to Multiple Field Strengths of 60-Hz Electric Fields date: 1990-04-17 journal: Fundam Appl Toxicol DOI: 10.1093/toxsci/14.3.608 sha: doc_id: 32979 cord_uid: jle66lmn Reproduction, Growth, and Development of Rats during Chronic Exposure to Multiple Field Strengths of 60-Hz Electric Fields. ROMMEREIM, D. N., ROMMEREIM, R. L., SIKOV, M. R., BUSCHBOM, R. L., AND ANDERSON, L. E. (1990). Fundam. Appl. Toxicol. 14, 608–621. A study with multiple exposure groups and large group sizes was performed to establish whether exposure to 60-Hz electric fields would result in reproductive and developmental toxicity. A response model was developed from previous results and tested in groups of rats exposed to electric fields at various field strengths. Female rats were mated, and sperm-positive animals randomly distributed among four groups: sham-exposed or exposed to 10, 65, or 130 kV/m, 60-Hz vertical electric fields. Animals were exposed for 19 hr/day throughout the experiment. During gestation, exposure to the higher field strengths resulted in slightly depressed weight gains of dams. Offspring were born in the field and remained with their dams through the suckling period. Numbers of pups per litter and pup mortality did not differ among the exposure groups. Dams exposed at 65 kV/m lost slightly more weight through the lactation period than the control group. Male pups exposed to higher field strengths gained slightly less weight from 4 to 21 days of age than did sham-exposed animals. At weaning, two F(1) females per litter (randomly selected) continued on the same exposure regimen were mated at 11 weeks of age to unexposed males, and euthanized at 20 days of gestation. Uterine contents were evaluated, and all live fetuses were weighed and examined for external, visceral, and skeletal malformations. Fertility and gestational weight gain of F(1) females were not affected by exposure, nor was prenatal viability or fetal body weight. No significant increase in the incidence of litters with malformations was observed. Although no developmental toxicity was detected, exposures produced physical changes in the dams, evidenced as a rust-colored deposit on the muzzle and ears (chromodac-ryorrhea) that increased in incidence and severity at 65 and 130 kV/m. Incidence of chromodac-ryorrhea was not significantly different between sham-exposed rats and those exposed at 10 kV/m. BUSCHBOM, R. L., AND ANDERSON, L. E. (1990) . Fundam. and Appl. Toxicol. 14, 608-621. A study with multiple exposure groups and large group sizes was performed to establish whether exposure to 60-Hz electric fields would result in reproductive and developmental toxicity. A response model was developed from previous results and tested in groups of rats exposed to electric fields at various field strengths. Female rats were mated, and sperm-positive animals randomly distributed among four groups: sham-exposed or exposed to 10, 65, or 130 kV/m, 60-Hz vertical electric fields. Animals were exposed for 19 hr/day throughout the experiment. During gestation, exposure to the higher field strengths resulted in slightly depressed weight gains of dams. Offspring were born in the field and remained with their dams through the suckling period. Numbers of pups per litter and pup mortality did not differ among the exposure groups. Dams exposed at 65 kV/m lost slightly more weight through the lactation period than the control group. Male pups exposed to higher field strengths gained slightly less weight from 4 to 21 days of age than did sham-exposed animals. At weaning, two F| females per litter (randomly selected) continued on the same exposure regimen were mated at 11 weeks of age to unexposed males, and euthanized at 20 days of gestation. Uterine contents were evaluated, and all live fetuses were weighed and examined for external, visceral, and skeletal malformations. Fertility and gestational weight gain of F| females were not affected by exposure, nor was prenatal viability or fetal body weight. No significant increase in the incidence of litters with malformations was observed. Although no developmental toxicity was detected, exposures produced physical changes in the dams, evidenced as a rust-colored deposit on the muzzle and ears (chromodacryorrhea) that increased in incidence and severity at 65 and 130 kV/m. Incidence of chromodacryorrhea was not significantly different between sham-exposed rats and those exposed at 10 kV/m. O 1990 Society of Toxicology. Effects on survival, growth, and development used to detect effects of electric field exposure of laboratory animals (mice and rats) exposed on reproduction and development in a larger to electric fields during gestation or neonatal species (Sikov el al., 1987) . Swine were life have previously been reported (ICnicker-chronically exposed to a 30 kV/m, 60-Hz bocker«a/., 1967; Marino el al., 1967 Marino el al., , 1980 ; electric field over two generations. Maternal Sikov el al., 1984) . However, the magnitudes weight gain during pregnancy, number of of observed changes were small, and effects embryonic implants, and number of offhave been inconsistent across studies. spring per litter were not affected. Survival More recently, a multigeneration screening rates of the offspring and their growth curves study utilizing Hanford Miniature swine was were also indistinguishable between groups. However, comparison of offspring from dams exposed or sham exposed for 18 months showed mean body weights of male fetuses and mean body weights of female piglets to be less in the exposed group. In addition, an increase in frequency of malformations was observed in fetuses of exposed females. A clear association of the changes with exposure, per se, could not be demonstrated because of a number of inconsistencies in response patterns and the possibility of interacting factors . Subsequently, rats were used to investigate the possible effects of electric field exposure on development and reproduction. Two, essentially identical, multiple-generation rat experiments were conducted . Copulatory behavior, gestational weight gains, pup weights at birth, and their subsequent growth curves were not affected by exposure. In one of the replicate experiments, there was a statistically significant decrease in fertility and a significant increase in the fraction of litters with malformed fetuses among exposed animals. However, in the second replicate, no significant developmental effects were detected in fetuses of dams exposed throughout gestation. A clear association of reproductive or developmental changes with exposure could not be demonstrated. Effects detected in the first experiment were not seen in the second and were attributed to either normal biological variation or a response threshold near 65 kV/m for induction of malformations. Accordingly, in the study reported here we tested for a treatment-related response, utilizing multiple exposure groups, including 65 -cV/m (effective field strength) as used in our arevious experiments; 130 kV/m which was :wice as strong as that used in the previous ,vork and had been shown to be tolerated by •ats (Rommereim et al, 1989) ; the lowest evel of exposure (10 kV/m) was chosen to deineate the response curve at lower field strengths or possibly to determine a no-obierved-effect level; and the fourth group was sham-exposed as a control. Additionally, the chance of obtaining equivocal results due to biological variation was minimized by using large group sizes with the initial group populations of 68 female rats per treatment group. To obtain a statistical power of 0.8 or greater in comparisons of malformation incidence between groups, the study was designed to obtain a minimum of 80 litters for teratological evaluation per exposure group. Exposure system. Rats were exposed to 60-Hz electric fields on systems consisting of five parallel-plate electrodes. The metal plates were rectangular and electrically insulated from one another. When energized, they provided uniform (±5%), vertical, 60-Hz electric fields at the strengths utilized in this experiment. Four such exposure systems (racks), electrically isolated from each other, were located within one room. Four tiers of rat cages were provided on each system; each tier contained eight polycarbonate modules, each with six individual rat cages (12.5 cm wide, 25.0 cm long, 10 cm high). During parturition and litter-rearing, rats were housed in modules divided into three compartments, so that the cages were then twice as large as the standard cages. Floors of the cages were made of wire mesh and were an integral part of the lower electrode so that rats were in electrical contact with the reference ground. A small amount of Antron III (a conductive carpet material from E. I. Dupont Co.) was placed in the cages during the period in which litters were delivered and reared. The exposure system design minimizes sources of phenomena associated with electric fields of intensities used in this experiment (Dietrich and Patterson, 1988) . Current conduction, transient charge, and transient energy through the water system to a rat while drinking were tested, and values were below known levels of perception. Electrode voltages and spacing were such that flashover from rats within cages did not occur. Corona, as detected by radio noise, was not present. Harmonic distortions of input voltages were less than 0.5%. Cage floor vibration velocity was less than 0.5 mm/sec measured at the voltages used in this study. Rats were exposed to one of four field strengths (0, 10, 65, 130 kV/ m) for 19 hr daily, 7 days per week. Animals and their respective exposure regimens were rotated weekly from one system to another to limit the potential effect of rack location within the room on rat growth and development. Personnel performing animal care, manipula- tions, or evaluations were not aware of an individual rat's exposure status. Animals. The experiment was conducted using Charles River (Sprague-Dawley-derived) rats obtained from the Raleigh, North Carolina, facility. Approximately 400 female and 100 male rats were received at 7 weeks of age and were group-housed in standard wirebottom cages for 4 weeks of quarantine and acclimation. Midway through the acclimation period, five females and five males were randomly selected for evaluation of health status. Gross necropsy and examination of histologic sections from major organs did not disclose any unusual lesions. Cultures of nasopharynx, lung, and cecum for bacterial pathogens were also negative. Serum was tested and found not to contain antibodies to Sendai virus, H-l virus, rat coronavirus (RCV/SDA), or Mycoplasma pulmonis. All female rats were identified by ear tattoo and weighed. Animal rooms were illuminated on a 14-hr light, 10-hr dark cycle (lights on from 1000 hr to 2400 hr). Cage board beneath the cages was changed daily. Room temperature and relative humidity were recorded continuously: the ranges were 69-75"F and 30-60% RH, respectively. Food and water were provided ad libitum to all animals at all times. For breeding, male rats were individually housed in standard hanging wire cages, and two females were placed with each male overnight. Following a breeding session, females were vaginally lavaged with physiological saline, and smears were prepared and examined for presence of sperm. Those animals found to be positive were weighed and randomly assigned (blocking on weight) into four groups, and exposure was begun on the day of mating. Experimental protocol. As indicated, the four treatment groups, each on an individual system, were shamexposed or exposed to 60-Hz electric fields of 10, 65, or 130 kV/m. The experimental design of the study is outlined in Fig. 1 . In each group, 68 females were mated and started on exposure over a period of 10 consecutive days. These females were designated as "F o pregnancy." They were allowed to complete gestation, deliver, and rear their offspring (the F| generation) through weaning. Rats that had copulated but had not delivered by 2 days after the expected delivery date (i.e., 24 days after coitus) were killed, and their absence of pregnancy and associated ovarian status were evaluated. The F, pups were sexed, counted, and weighed as a litter (by sex) at 1 and 4 days of age. Litters were randomly reduced to a maximum of eight offspring (six females and two males, when possible) at 4 days of age and maintained with their dams until weaning at 21 days of age. Dams and offspring were weighed individually at 7, 14, and 21 days after birth. At weaning, two F, female offspring were randomly selected from each litter for continued exposure. These F, females were identified by ear tattoo, individually caged, and weighed at 4, 7, and 11 weeks of age. (The F o dams, F, males, and other excess F, female offspring did not continue in this experiment.) Ten F o dams, selected at random, were sampled to determine health status in the same manner as for the original health screening; no evidence of health problems was detected. For the breeding of the F, females, a new group of males was obtained from Charles River Laboratories. They underwent quarantine and health screening as described for the initial shipment of animals. When the F, females were 3 months of age, two females from different treatment groups were placed with one male for the last 2 hr of the daily dark period. During the dark period, movement of animals to breeding cages by technicians was facilitated with low-intensity red-light illumination. After the 2-hr period, females were examined for the presence of sperm. All females were then returned to their respective exposure systems, and exposures continued until the day of termination. The day on which sperm was detected was designated 0 dg. Mating procedures were repeated for up to 8 consec-utive days for females in which sperm was not detected. Females that were determined to have mated were not returned to males. After the breeding phase of the experiment was completed, females in which sperm was not detected were euthanized (13 days after the last day of possible mating), and their pregnancy status was determined (Salewski, 1964) . Body weight was recorded at 0, 7, 15, and 20 days of gestation (dg). Animals were killed at 20 dg for teratological evaluations. Teratologic evaluation: maternal and fetal examination. Females that mated were killed on 20 dg with carbon dioxide. Their body cavities were opened, and the uterus, ovaries, and abdominal and thoracic viscera were examined visually. Corpora lutea were counted, and maternal liver and uterus were weighed. Uteri were externally examined for hemorrhage, removed from the peritoneal cavity, and incised longitudinally to expose their contents. All live and dead fetuses and resorption sites were recorded. Apparently nongravid uteri were dissected and placed in a 10% ammonium sulfide solution to detect possible implantation sites (Salewski, 1964) . Live fetuses were individually weighed as were their placentas. Each fetus was examined for external malformations. Half the live fetuses in each litter were then decapitated, and their heads were fixed in Bouin's solution for examination of craniofacial structures by a modification of the sectioning methods described by Wilson (1965) . All fetuses in each litter were examined for thoracic and abdominal visceral abnormalities by a modification of the methods described by Staples (1974) and sexes were determined by internal examination. The fetuses were eviscerated and fixed in ethanol, and skeletons were stained with alizarin red S (Staples and Schnell, 1964) for examination. Statistical methods. Sample size determination for this study was based on an assumed angular dose-response model. The model was 2 arcsin YP, = 0.451 + 0.00296F, where P] is the proportion of litters with one or more malformed fetuses at field strength F, kV/m. The intercept and slope estimates were calculated from results of previous studies where P, was approximately 0.05 at 0 kV/m and 0.10 at 65 kV/m. If this model were valid, then for this study the predicted proportion of litters with one or more malformed fetuses would be approximately 0.05 at 0 kV/m, 0.057 at 10 kV/m, 0.10 at 65 kV/m, and 0.168 at 130kV/m. Data from continuous-type variables were evaluated by analysis of variance (ANOVA), utilizing the general linear model (GLM) procedure from the Statistical Analysis System (SAS, 1985) . Repeated measures, such as maternal weight measurements, were analyzed by repeatedmeasures ANOVA. Fetal body weights, with the litter as the experimental unit, were analyzed by nested ANOVA, which takes into account effects of treatment, litter, and sex. Data calculated as a response proportion, e.g., proportion of implants resorbed, were analyzed by ANOVA after arcsin transformation of the proportion (Freeman and Tukey, 1950) . When significant treatment effects were detected by ANOVA, intergroup differences were delineated utilizing Tukey's multiple comparison test (Tukey, 1953; Kramer, 1956) . Orthogonal contrasts were used to test for a trend in the response with increased exposure intensity. Results that differed at the p < 0.05 level were considered to be statistically significant. Rates of rat breeding per group were analyzed by an actuarial life-table method (Culter and Ederer, 1958) . The response criterion was the day an animal mated. Response curves were tested by a generalized Wilcoxon test (Breslow, 1970) to determine differences among groups. Data from binary-response-type variables (e.g., number of pregnant females, number of malformed fetuses, number of litters with malformed fetuses) were evaluated by x 2 test. When the x 2 test was significant, pairwise comparisons were made using Fisher's exact test (Siegel, 1956) . The Cochran-Armitage test was conducted to test for linear trend of response with increased field strength (Cochran, 1954; Armitage, 1955) . The P4F program of BMDP Statistical Software (Dixon, 1983) was used for tests of binary response data. Of the 68 mated female (F o ) rats per group, the pregnancy rate was 88 to 96% among the four groups (Table 1) . Exposure had no detectable effect on the percentage of animals that maintained pregnancy. The initial random assignment was blocked on weight; therefore essentially the same mean initial body mass existed in the four groups (Table 1) . Gains in body mass throughout gestation appeared to decrease with increasing field strength for the two higher-level exposed groups of F o females. Although approaching significance by the trend test (p = 0.065), this decrease was not statistically significant, as the overall test for equality of means indicated no significant differences among groups (p = 0.081). Parturition did not occur in four pregnant dams, three of which (each from a different exposure group) had one pup per litter and a fourth died, with a pup lodged within the birth canal. Mean litter size and pup mortality incidences in the period between 1 and 4 days of age and between 4 and 21 days of age were similar in all groups. The sex ratio (mean percentage of male pups) in each litter was not different among the treatment groups. Pup mean body masses were similar among the groups for both sexes at 1 and 4 days of age (Table 2) . Growth curves were similar for all groups within sex, although male pups exposed at the two higher field strengths tended to gain slightly less weight (p < 0.05) in the period between 4 and 21 days of age than did those in the other groups. Mean weight gains of female pups (females selected to continue on experiment after 21 days of age) did not differ among the groups. Mean body weights of dams at Day 1 after parturition were similar in the four groups. Rats exposed at 65 kV/m lost significantly more weight during lactation than did those of the other groups (Table 2) . Throughout the postweaning period measures of toxicity or animal well-being were equivalent among groups for the selected female rats that continued on study. Only two rats died during this period, of causes unrelated to exposure (broken tooth, bladder stones), and mean body weights among the groups were the same. Breeding procedures began when the F, female rats reached 11 weeks of age. Approximately equal numbers of females were mated daily from each group, and curves, calculated from the cumulative numbers of rats mated daily, did not differ among groups. Over a 12day period, mating was detected to have occurred in 81 to 91% of the rats in the four groups, the pregnancy rate of which varied between 87 to 94%. These values were not significantly different among groups (Table 3) . Mean numbers of consecutive daily mating periods were approximately equal among groups. Mean body weights at 0 day of gestation were not significantly different among groups Data expressed as means ± SE except as noted. * One rat removed due to technician error. ' Mean number of daily periods paired with males to achieve copulation. d Days of gestation; data shown only for rats that were pregnant. ' Gestational weight gain minus uterine weighL / Liver weight divided by 20 dg body weight X 100. • p < 0.05 versus sham-exposed, Tukey's test. " Data expressed as means ± SE except as noted. * One dam aborted before litter evaluation. c One dam was excluded from evaluation due to dental malocclusion. J Data expressed as mean of litter means. (p = 0.075). Mean weight gain through pregnancy was not affected by exposure, and extragestational weight gain (an indicator of dam well-being through pregnancy) was also not affected by exposure. In teratological examinations, mean gravid uterine weights were not affected by treatment (p = 0.52). Absolute mean liver weight was significantly less (p = 0.05) in rats exposed at the highest field strength than in the sham-exposed group. However, ratios of mean liver weight to body weight did not differ. No significant differences were observed among treatment groups with respect to the total number of corpora lutea per dam, implantation sites per dam, or preimplantation loss as indexed by implantation sites per corpus luteum (Table 4 ). Postimplantation embryonic death was not affected by exposure, as indicated by the fact that none of the measures of implant resorption differed among groups. Mean number of live fetuses per litter was less in rats exposed to the 130 kV/m field, but it was not statistically different from the control group {p = 0.33). Mean fetal body weight and placental mass were the same among the groups. The sex ratio (number of males/litter) appeared to increase with field strength but did not differ significantly among groups (p = 0.08) and was within the range of expected variation. A total of 5076 fetuses in 375 litters were examined. In all cases, the number of litters examined per treatment group exceeded the 80 litters required for a statistical power of 0.8 or greater for comparison of malformation incidence. A wide variety of malformations were recorded, as would be consistent with the large number of litters examined (Table Data presented as No. of fetuses affected/No, of litters affected. * A single fetus may be represented more than once if multiple defects were present. * p < 0.05 versus sham-exposed, x 2 test. 5). Several malformations were recorded in litters from exposed groups which did not occur in the sham-exposed animals: anotia, acaudal or vestigial tail, edema, imperforate anus, umbilical hernia, retrognathia, cleft palate, rib agenesis, cleft sternum, fused ribs, and kinked tail. However, the incidence of each was low and did not differ significantly between sham-exposed and exposed groups. The predominant major vessel malformation observed was a retroesophageal aortic arch with abnormal origin of the right subclavian artery. Vertebral agenesis occurred in the sacral region. Malformation incidence was low overall, with the exception of fused sternebrae, which occurred in one fetus in the sham-exposed group versus eight in one litter of the 65 kV/m group and seven in one litter in the 130 kV/m group. Incidence of fused sternebrae was significantly higher (p < 0.05) on a per fetus but not on a per litter basis. Evaluation of total malformation incidence on a per fetus basis showed a significantly higher number of malformed fetuses in the " Data are presented as No. of fetuses affected/No, of litters affected. b A single fetus is represented more than once if multiple defects were present. * p < 0.05 versus sham-exposed, x 2 test. group exposed at 65 kV/m than was observed in the other groups. However, on a per litter basis the total incidence of malformations did not differ significantly among groups. The mean percentage of malformed fetuses per litter was the same among the treatment groups (p = 0.34). A few incidences of fetal variations were significantly higher for specific site/exposure group combinations, but no clear correlation of increasing incidence with increasing field strength was evident. For example, incidence of dilated ureter was significantly higher in fetuses of dams exposed to 65 kV/m electric fields on both a per fetus and a per litter basis ( Table 6 ). The associated variations of renal pelvic cavitation were also higher on a per fetus basis. Incidence of an additional ossification site at the first lumbar vertebra was significantly higher per fetus (but not per litter) in animals exposed to 65 kV/m; the incidence was not higher in rats exposed to 130 kV/m. Likewise, significantly increased incidences of reduced ossifications in skull, sternum, and vertebrae were recorded on a fetus and litter basis in rats exposed at 10 kV/m. Incidence of this condition in rats exposed at higher field strengths did not differ from that of the sham-exposed group. During routine observation of the F o dams, a rust-colored deposit (chromodacryorrhea) was observed on the muzzles and ear hair. Subsequently, the deposits were also observed on F| females during the 3 months of exposure following birth and the incidence and severity of chromodacryorrhea were recorded at termination. A statistically significant trend (p < 0.0001) toward an increased incidence of deposits was recorded with increasing field strength in both lactating and pregnant animals. Seventy-eight percent of nursing dams exposed at the highest field strength had detectable amounts of such deposits (Table 7) . Further details on these observations are reported elsewhere (Leung el al., 1989 ). Note. Number (%) scored within each category. " Amount of deposits defined by subjective scores as: 0, without deposits; 1, slight amounts, light dots; 2, moderate areas on nose or ears, approximately 1 to 3 mm in diameter, 3, moderate areas on nose and ears, each approximately 1 to 3 mm in diameter, 4, significant areas on nose and ears, approximately 3 mm in diameter or larger, with interrupted areas extending from nose to eyes; 5, large areas on entire head and ears. * Significantly different from controls (p < 0.0001). Throughout the study there were no indications of disease or confounding factors unrelated to exposure other than an occasional transient reduction in body mass of individual animals as a result of water-dispenser malfunction. Effects of this transient deficiency are judged to be minimal. In general, mean body weights were consistent among the groups, and body weight gains were comparable to those in previous experiments. Our observation of slightly reduced gestation body weight gain of F o dams exposed at 130 kV/m was unexpected. In our previous work, gestational body weight gain in females exposed to 150 kV/m did not differ from that of sham-exposed animals (Rommereim et ai, 1989) . Differences between the two experiments may be explained in that dams in this study received no acclimation to the electric field prior to breeding in contrast to a 1month exposure in the previous experiment. Reproductive outcome was largely unaffected, as indicated by the fact that the mean number of live pups per litter did not differ significantly among the groups. The reduced gestational weight gain of the dams in the high exposure group is consistent with a numerically lower, albeit nonsignificant, number of live pups per litter in the high exposure group; however, a chance association is an equally plausible explanation. Overall, no evidence of detrimental effects on survival or growth of the offspring was observed in animals exposed to electric fields. No maternal mortality was associated with electric field exposure; only two deaths occurred and those were attributable to factors other than the field. Most of the rats that did not deliver at the expected time were not pregnant, and three of the F o dams that did not litter had only a single large pup in utero. The low incidence of this situation and its distribution among groups indicate that it is not attributable to electric-field exposure. The weight loss of lactating dams exposed to the 65 kV/m field was slightly greater than that in other groups. However, the weight loss was not proportional to increased field strength and is therefore not considered of consequence. In the F, generation females, measures of reproductive capability were unaffected by exposure: the percentage mating and rate at which rats mated (cumulative proportion on consecutive days) did not differ among the treatment groups. Although not statistically significant, the number of live fetuses per litter was numerically less (0.8 pup per litter) in rats exposed to 130 kV/m than in other groups (Table 4 ). This deficit is not accounted for by prenatal death; rather, it reflects fewer than normal ovulations (indicated by the number of corpora lutea per dam) and thus fewer implantation sites per dam. The electric field may have decreased the number of ovulations, but chance is an equally tenable explanation, as the values did not differ significantly. Low fetal body weights generally provide an excellent indicator of deleterious effects. Equivalence of this parameter among groups gives no indication of fetal toxicity. The lack of malformation differences between groups, with the litter as the basis for comparison (Fig. 2) , indicates that exposure was not teratogenic. It should be noted that the number of affected fetuses was significantly elevated in the 65 kV/m group and a specific malformation (fused sternebrae) occurred at a significantly higher incidence in fetuses exposed to 65 and 130 kV/m (Table 5 ). However, these findings do not provide credible evidence of electric-field exposure effects because the incidence of fused sternebrae in the 65 and 130 kV/m groups was limited to one litter in each group. Nor was this finding unique to litters exposed to electric fields, as fused sternebrae also occurred in an unexposed fetus. Historical control incidence of fused sternebrae approximates incidences recorded in the sham-exposed group (1 in 2600 fetuses, 1 in 170 litters). Examination of mating records among litters with malformations did not reveal significant relationships of sire or grandam to litters with malformations or to type of malformation. Two litters with malformations in the 65 kV/m group were sired by the same male rat; both litters had a single malformed fetus with affected ribs. This male sired two additional litters in this group that had no malformed fetuses and two litters in other exposure groups that were likewise without malformations. Litters with incidences of fused sternebrae were sired by different males. Two litters in the 65 kV/m group had the same grandam, and each had one affected fetus but with different malformations. No other litters with malformations had the same grandam. Because the F o generation was purchased from an animal supplier our genetic information is limited, but from the available data, no significant relationships are evident. Although a few types of malformations differ between groups, these findings are not considered to be treatment-related effects because of the absence of increased incidence in (1) total number of litters with one or more malformed fetuses (litters rather than fetuses are the unit of analysis (Manson et ai, 1984) ), (2) specific malformation or category of malformations when compared by litters, or (3) any response by increasing field strength. The teratologjcal assay yields several parallels with results of our previous replicate experiment. In the earlier experiments (Rommereim et ai, 1987) incidences of litters with malformed fetuses (combined data from replicate experiments) from 3-month-old F, dams were 5.0% for controls and 9.5% in litters from animals exposed at 65 kV/m. In the results reported here, malformation incidence was doubled in rats exposed at 65 kV/ m as predicted by the statistical model (Fig. 2) . However, lack of a comparably increased percentage of affected litters in animals exposed at 130 kV/m invalidates the hypothesis that increased field strength would result in increased numbers of terata. Results of this experiment clearly demonstrate that developmental changes did not increase with field strength exposures above 65 kV/m, mediating against the existence of a threshold for altered development. One could argue that there is a window of effect around the 65 kV/ m exposure level. At this time, however, a biological explanation for such an event does not exist. The only overt sign of an effect of electric fields in adult rats was a significant increase at 65 and 130 kV/m in the incidence and severity of a rust-colored deposit on the face and ears of mature rats exposed to 65 or 130 kV/m (Fig. 3) . These deposits occurred in the absence of any detectable effects on body weight. The material is thought to originate in the Harderian gland, located within the eye-socket, behind the eye. Phenomena of a similar nature have been documented in rats as a response to various stressors and were called chromodacryorrhea by Harkness and Ridgway (1980) . This effect appears to indicate a physiological response to exposure, perhaps to chronic irritation or stress, but the mechanism for stimulating production of the exudate by electric fields is unknown at this time. Because a significant increase in the incidence of chromodacryorrhea was not reported in earlier experiments, it is probable that exposure at 130 kV/m (twice that used in the previous work) intensified a subtle effect, prompting closer investigation and, thus, detection at 65 kV/m. Occasionally, the severity of chromodacryorrhea in individual shamexposed females was equal to that of exposed rats (Table 7) . Furthermore, not all exposed rats displayed the colored deposits, making detection of a subtle effect among groups difficult at the lower field strengths. A similar chromodacryorrhea response has been reportedly produced by various stressors (Scialli, 1988) , and stress in rodents has been shown to produce adverse effects on implantation and embryo viability, resulting in terata when sufficiently intense (Cook et ai, 1982; Rozenzweig and Blaustein, 1970; Barlow et ai, 1975) . Additionally, the incidence of chromodacryorrhea in rats exposed at 65 and 130 kV/m but not at 10 kV/m clearly indicates a response to electric field exposure; however, the fields were not stressful enough to produce developmental toxicity. The incidence of chromodacryorrhea was not significantly elevated in litters with malformations (p = 0.54). If the chromodacryorrhea produced by electric field exposure is an indicator of stress, then stress is a potential factor which may have contributed to the developmental Values are significantly different (p< 0.0001) from sham-exposed rats compared to rats exposed to 65 or 130 IcV/m. changes reported in the earlier studies. For example, when combined with other factors (disease, diet, housing, etc.), electric field induced stress may have acted synergistically to produce the developmental changes observed in the swine experiment. The effects of electric fields on developmental integrity when combined with stressors have not been specifically examined. Tests for linear trends in proportions and frequencies The relation between maternal restraint and food deprivation, plasma corticosterone, and induction of cleft palate in the offspring of mice A generalized Kruskal-Wallis test for comparing k samples subject to unequal censorship Some methods for strengthening the common x 2 tests Effects of high-frequency noise on prenatal development and maternal plasma and uterine catecholamine concentrations in the CD-I mouse Maximum utilization of the life table method in analyzing survival Design and Testing of a Small-A nimal Electric Field Exposure Facility, Final Report, EPRI Project 799-18 Electric Power Research Institute BMDP Statistical Software Univ Transformations related to the angular and the square root Exposure of mice to a strong ac field-An experimental study Extension of multiple range tests to group means with unequal numbers of replications Brown-colored deposits observed on hair of female rats chronically exposed to 60-Hz electric fields Teratology test methods for laboratory animals The effect of continuous exposure to low frequency electric fields on three generations of mice: A pilot study Power frequency electric field induced biological changes in successive generations of mice Reproduction and development in rats chronologically exposed to 60-Hz electric fields Rats successfully reproduce and rear litters during chronic exposure to 150-kV/m, 60-Hz electric fields Cleft palate in A/J mice resulting from restraint and deprivation of food and water Farbemethode zum makroskopischen nachweis von implantations-stellen am uterus der ratte SAS User's Guide: Statistics. Version 5 SAS Institute Inc Is stress a developmental toxin? Non-Parametric Statistics for Behavioral Sciences Studies on prenatal development in rats exposed to 60-Hz electric fields Developmental studies of Hanford miniature swine exposed to 60-Hz electric fields Detection of visceral alterations in mammalian fetuses Refinements in rapid clearing technic in the KOH-alizarin red S method for fetal bone The problem of multiple comparisons Methods for administering agents and detecting malformations in experimental animals