Final Report: Developmental Effects of Dietary Soy Phytoestrogens

EPA Grant Number: R825721
Title: Developmental Effects of Dietary Soy Phytoestrogens
Investigators: Hughes, Claude L. , Davis, Vickie , Tyrey, Lee
Institution: Duke University Medical Center
EPA Project Officer: Carleton, James N
Project Period: November 1, 1996 through October 30, 1999
Project Amount: $574,747
RFA: Endocrine Disruptors (1996) RFA Text |  Recipients Lists
Research Category: Endocrine Disruptors , Health , Safer Chemicals


This project was directed towards determining whether intake of dietary soy phytoestrogens by rodents from mid-gestation throughout lactation to weaning will affect the progeny's sexually dimorphic development of various hormone-dependent target tissues.

Summary/Accomplishments (Outputs/Outcomes):

There is widespread concern that developmental exposure to endocrine active chemicals (EACs) may adversely affect reproductive, neurobehavioral, and immunological development or predispose those exposed towards various cancers later in life. If these concerns about developmental exposures to EACs are provable, then there must be evidence of effects in animal models of EACs at properly scaled levels of exposure during sensitive timeframes, and demonstration that exposure of the human fetus to comparable quantities of these same EACs occurs during biologically comparable sensitive timeframes. Our recent studies have continued to show effects that can be produced by EACs in rodent models and quantitation of in utero exposure of the human fetus to these types of chemicals.

First Study. This study focused on the effects of diethylstilbestrol (DES) at marginally effective doses and genistein (GEN) at levels comparable to the upper range of potential dietary intake on some markers of sexual development in Long-Evans hooded rats. Dams, sires, and pups were maintained on a purified casein-based diet throughout the study. Dams were randomly assigned to each of the five treatment (daily doses) groups: Control (Corn oil); Low DES (0.5 g/kg BW); High DES (5.0 g/kg BW); GEN (15 mg/kg BW ); and GEN + DES (GEN at 15 mg/kg BW and DES at 0.5 g/kg BW). Mating generated 43 control pups, 47 pups exposed to low DES, 29 pups exposed to high DES, 43 pups exposed to GEN, and 43 pups exposed to the combination of GEN + DES. Dams were gavaged from Gestation Day (GD) 14 through weaning on Post-Natal Day (PND) 21 with either corn oil alone (unexposed controls) or the estrogen agent in corn oil, with the exception that DES was not administered GD 20 through birth. On PND 1, litter size, sex, birth anogenital distance (bAGD), and birth body weight (bBW) were recorded, and individual pups were tattooed for identification. At weaning (PND 21), weaning anogenital distance (wAGD) and weaning body weight (wBW) were documented. The age of puberty was determined by inspecting the offspring for the day of vaginal opening or preputial separation. Upon vaginal opening, vaginal cytology was assessed by daily saline lavage to monitor initial cyclicity. The reproductive organs of male pups age 50-70 days were removed and weighed. At birth, no treatments affected the duration of gestation, litter size, or birth anogenital distance/birth body weights ((bAGD/bBW) of pups of either sex. The ratios of bAGD/cube root of bBW were calculated and with the data expressed in this fashion, no effects of treatments were found. At weaning, the ratio of weaning AGD to weaning body weight (wAGD/wBW) differed significantly between the control group and each of the estrogenic treatments in both sexes, with larger wAGD/wBW values associated with each of the estrogenic treatments. Expression of the data in the form of the ratio of wAGD/cube root of wBW did not change the finding of effects in any of the treatment groups in either sex, except for the one group exposed to the high dose of DES, in which case the values for both sexes no longer differed from respective controls. Observations at puberty showed that males exposed to High DES and GEN alone exhibited earlier onset of puberty (about 3 days earlier than the controls). The onset of puberty was unaffected by estrogen treatment in all groups of females except low DES, which showed an earlier onset of puberty by approximately 1 ? days. Observations in the post-pubertal interval showed that the ratios of male reproductive organ weights/body weight (testes, epididymis, seminal vesicle, and prostate) were unaffected by estrogen treatment in all groups except high DES, which increased testicle weight and decreased epididymis, seminal vesicle, and prostate weights. Length of the initial vaginal cycles during the first 3 weeks after vaginal opening was 4 days for all groups except for the females exposed to high DES. This group did show cycles, but there were more cycles with additional days of cornifed smears yielding a mean cycle length of 6 days.

This experiment showed that the reference estrogen DES at low doses and the common dietary phytoestrogen GEN at levels comparable to the upper range of human exposure affect some, but not all, of the markers of sexual development that we assessed in Long-Evans hooded rats.

Second Study. The second study was performed to test our hypothesis that man-made chemicals and phytoestrogens can be quantified in human amniotic fluid during the second trimester. Two groups of amniotic fluid samples obtained at routine amniocentesis were assayed by gas chromatographic/mass spectrometric (GC/MS) analysis. The first group of amniotic fluid samples (n=53) from women (n=51) with a mean (SEM) age of 36.5 ? 0.5 years and between 15 and 23 weeks of gestation were analyzed for common PCB congeners, the DDT metabolites p,p'-DDE, and o,p'-DDE as well as the pesticides hexachlorobenzene (HCB) and the three isomers of hexachlorocyclohexane ( , , and -HCH). The limit of quantitation (LOQ) for PCBs was 0.01 ng/mL and for the other organochlorines contaminants it was 0.1 ng/mL. The contaminants -HCH with a mean (? SD) concentration of 0.15 ? 0.06 (ng/mL) and p,p'-DDE with a mean (? SD) concentration of 0.21 0.18 ng/mL were detected in the amniotic fluid. PCB-specific congeners were detected with a much lower frequency, and levels were in the range of the LOQ. Overall, one in three amniotic fluid samples tested positive for at least one environmental contaminant. The second group of amniotic fluid samples (n=62) from women (n=56) with a mean (? SD) age of 36.7 ? 5.2 years and between 15 and 23 weeks of gestation were analyzed for the common phytoestrogens daidzein, genistein, formononetin, biochanin-A, and coumestrol. The LOQ for each of these phytoestrogens was 0.50 ng/mL. The mean (? SD) concentration of daidzein and genistein in amniotic fluid were 1.14 ? 1.04 and 1.37 ? 1.00 ng/mL with maximum levels of 5.52 and 4.86 ng/mL, respectively. Overall, 26 (46.4 percent) and 34 (60.7 percent) of the women had quantifiable levels of daidzein and genistein, respectively in their amniotic fluid. There were no differences in the ethnic background, age, or gestational age of women who had detectable levels of phytoestrogens compared to those who did not.

In Los Angeles, approximately one in three fetuses are exposed to xenobiotic endocrine active environmental chemicals, and more than one in two fetuses are exposed to phytoestrogens in utero. Our approach of using amniotic fluid to quantify the presence of hormonally active exogenous factors provides a viable and proximate (to time of reproductive tract differentiation) estimate of fetal exposure to hormonally active compounds. Although the consequences of these exposures for the human fetus remain unknown at this time, our demonstrations in (a) effects of these EACs in an animal model at levels of exposure that mimic those found in humans in North America during sensitive timeframes; and (b) exposures of the human fetus to pertinent quantities of these same EACs do occur during biologically comparable sensitive timeframes, sustain concerns about potential adverse health effects of developmental exposures to EACs for the human fetus/neonate.

Third Study. The major emphasis of our last phase of this work has focused on developmental effects of dietary soy phytoestrogens in target tissues in mice. We have initiated developmental trials in mice as planned and have gathered the initial sets of observations. Analyses of tissues obtained from these experiments are ongoing.

In this study, we addressed the issue of whether in utero or lactational exposure to these phytoestrogens can modify reproductive development by exposing mice during fetal and/or neonatal development. Dams were treated with diethylstilbestrol (DES, 0.03 g/kg/day), daidzein (40 mg/kg/day), or genistein (4 and 40 mg/kg/day) during pregnancy (GD 14-18) and/or lactation (birth-weaning). All treatments were given by gavage to the dams to prevent direct exposure to the pups. The dams and the offspring were fed a semipurified diet to prevent unintentional exposure to isoflavones from the mouse chow. Endpoints measured at birth, weaning, and/or sexual maturity (2 months of age) included ano-genital distance, body weight, uterine weight, testes weight, onset of puberty, and estrus cycling. The potent estrogen, DES, was found to alter many reproductive outcomes in both male and female mice, as expected. For some of these endpoints, daidzein and genistein, mimicked DES; however, unique or opposite responses were also evident for each of the phytoestrogens, including the low dose of genistein. Changes were evident for both male and female pups with in utero, lactational, and both exposures to all four treatments compared to the appropriate control group. The window of exposure also was important for type of response elicited by the treatments; that is, in utero and lactational exposure often had different effects. Altered outcomes during sexual maturity, such as estrogen cycling and uterine or testicular weight, indicated that treatment effects continued to be evident after the end of the exposure window (birth or weaning). These data indicate that in utero and lactational exposure to isoflavones can influence reproductive development of mice. Therefore, exposure of human fetus and/or nursing baby to these isoflavones has the potential to influence immediate and future reproductive development.


The following conclusions were reached:

(1) In addition to endogenous steroids, the human fetus is exposed to mixtures of naturally occurring and xenobiotic EACs.

(2) Exposures to mixtures of EACs occur at times that are known to be critical developmental intervals for the human fetus.

(3) Developmental exposure of rodents to some EACs at levels of exposure that mimic those found in humans in North America produces target tissue effects that indicate changes in cell function.

Journal Articles:

No journal articles submitted with this report: View all 2 publications for this project

Supplemental Keywords:

endocrine disruption, rodent, reproduction, development, female, animal, animal feed, birth weight and drug effects, carcinogen, DES, pharmacology, dose-response, drug effects on growth and development, estrogens, non-steroidal, health effects, isoflavones, toxicity, rats, Long-Evans, sex differentiation, sex maturation., RFA, Health, Scientific Discipline, Environmental Chemistry, Health Risk Assessment, Endocrine Disruptors - Environmental Exposure & Risk, endocrine disruptors, Susceptibility/Sensitive Population/Genetic Susceptibility, Children's Health, genetic susceptability, Biology, Endocrine Disruptors - Human Health, adverse outcomes, central nervous system, puberty, sensitive populations, adolescence, childhood development, epidemiology, health risks, females, exposure studies, exposure, gender, brain chemistry, children, animal models, developmental processes, phytochemicals, growth and development, reproductive processes, environmentally caused disease, biological effects, dietary soy phytoestrogens, female, dietary exposure, pregnancy, rodent, developmental toxicants, endocrine disruption, environmental hazard exposures

Relevant Websites: Exit EPA icon Exit EPA icon

Progress and Final Reports:

Original Abstract
  • 1997
  • 1998 Progress Report