Monograph on the Potential Human Reproductive and ... - OEHHA

Strengths/Weaknesses: Strengths of this study were
the bisphenol A determinations that were made and the
anchoring of animal exposure levels to human exposures.
The design appears sound with a good range of
endpoints measured. Small numbers of animals were
sacrificed at several time points and cellular analyses
were performed; these numbers were too small for a
definitive cancer evaluation and were, in fact, too small
for definitive conclusions to be reached for most of the
adult reproductive endpoints. Statistics are not described
in enough detail to determine how data from multiple
sampling points were evaluated. This experiment represents
a thorough screening study.
Utility (Adequacy) for CERHR Evaluation Process:
This study is inadequate based on insufficient sample
size (3–5/group).
Takagi et al. (2004), supported by the Japanese
Ministry of Health, Labor, and Welfare, examined the
effect of perinatal bisphenol A exposure on the reproductive
and endocrine systems of rats. Nonylphenol was
also examined but will not be discussed. Sprague–
Dawley rat dams were fed a soy-free diet (Oriental Yeast
Co.) prepared according to the formula for NIH-07. At
weaning, the offspring were fed CRF-1 diet (Oriental
Yeast Co.), which contains soybean and alfalfa-derived
proteins. Rats were housed in polycarbonate cages
containing wood chip bedding. Dams were randomly
assigned to groups, and 5–6 dams/group were fed diets
containing bisphenol A (96.5% purity) at 0, 60, 600, or
3000 ppm from GD 15 (GD 0 5 day of vaginal plug) to
PND10 (PND 1 5 day of birth). The study authors
estimated bisphenol A intake at B5, 49, and 232 mg/kg
bw/day during the gestation period and B9, 80, and
384 mg/kg bw/day during the lactation period. Dose
levels were based on results of preliminary studies, and
selected with a goal of achieving weak to moderate
toxicity in dams at the highest dose. In a separate study,
rats were fed diets containing ethinyl estradiol at 0 or
0.5 ppm from GD 15 to PND 10. On PND 2, offspring
were counted, sexed, and weighed and anogenital
distance was measured. Litters were culled to 6 pups
on PND 10, and pups were weaned on PND 21. Five
pups/sex/group (1/sex/litter) were selected for necropsy
on PND 21 and brain, adrenals, testis, ovary,
and uterus were weighed. Eight offspring/sex/group (at
least 1/sex/litter) were selected for evaluation in adulthood,
and these rats were observed for age and body
weight at puberty. Estrous cyclicity was observed from 8–
11 weeks of age. Offspring were killed at 11 weeks of age,
on the day of diestrus for cycling female rats. Brain,
pituitary, thyroid, adrenal mammary gland, epididymis,
prostate, seminal vesicles, ovary, uterus, and vagina were
weighed and examined histologically. The testis was
fixed in Bouin solution, and other organs were fixed in
10% neutral buffered formalin. The volume of the
sexually dimorphic nucleus of the preoptic area (SDN­
POA) was measured. It appears that endpoints were
assessed in 8 adult rats/sex/group, with the exception of
histopathological evaluations, which were conducted in 5
rats/sex/group. The litter was considered the experimental
unit in statistical analyses of data from PND 21
offspring, and the individual animal was considered the
statistical unit for data obtained from adult offspring.
Homogenous numerical data were analyzed by ANOVA
and Dunnett test, and heterogeneous numerical data
Birth Defects Research (Part B) 83:157–395, 2008
BISPHENOL A
255
were analyzed by Kruskall–Wallis H-test and Dunnetttype
rank sum test. Data for histopathological lesions and
vaginal cyclicity were analyzed by Fisher exact probability
test or Mann–Whitney U-test.
Maternal body weight gain was significantly decreased
the high-dose bisphenol A group during gestation,
but there were no effects on body weight gain
during lactation or food intake. In offspring evaluated on
PND 2, there were significant decreases in body weight
in low- and high-dose males [13 and 22%] and in highdose
females [20%], but there were no effects on number
of live offspring or anogenital distance. Body weight gain
was lower in high-dose males [21%] and females [29%]
from PND 2–10. Increased relative brain weight as a
result of growth retardation was reported in high-dose
offspring evaluated on PND 21. [Data were not shown
by study authors.] Exposure to bisphenol A did not
affect onset of vaginal opening, preputial separation, or
estrous cyclicity. Body weight of males was significantly
lower [by 9.3%] at adult necropsy. Weights and histopathology
of brain, pituitary, thyroid, adrenal mammary
gland, epididymis, prostate, seminal vesicles, ovary
uterus, and vagina in adulthood were unaffected in rats
from the bisphenol A group. [Organ weight data were
not shown by study authors.] Bisphenol A did not affect
SDN-POA volume. Effects observed in offspring from
the ethinyl estradiol study included reduced numbers of
live offspring, increased male:female ratio, decreased
body weight and body weight gain, accelerated vaginal
opening, delayed preputial separation, increased estrous
cycle irregularities, and histopathological alterations in
pituitary, ovary, uterus, vagina, and mammary gland.
The study authors concluded that bisphenol A did not
affect endocrine or reproductive system development of
rats at doses that induced maternal toxicity.
Strengths/Weaknesses: Strengths include the range of
doses and endpoints measured and the use of the ethinyl
estradiol comparator group. The study used small
sample sizes of dams (n 5 5–6/group) and inadequate
statistical procedures to control for litter effects.
Utility (Adequacy) for CERHR Evaluation Process:
This study is considered inadequate for the evaluative
process, based on sample size and statistical procedures.
Akingbemi et al. (2004), supported by NIEHS, USEPA,
NICHHD, and NIH, conducted a series of studies in
Long–Evans rats to determine the effects of postweaning
and perinatal exposure to bisphenol A on testicular
steroidogenesis. In vitro studies were also conducted and
are described in Section 4 because cells used in the
studies were obtained from adult animals. Rats were fed
Purina chow, which contains soybean meal, and given
drinking water in polycarbonate bottles. Pregnant and
nursing dams were housed in polycarbonate cages lined
with wood bedding, but no information was provided on
caging used at the other life stages. To reduce leaching of
bisphenol A, the cages were washed, rinsed, and dried at
least twice/week and were discarded once they began
getting cloudy; water bottles were cleaned daily. Corn oil
vehicle was used for bisphenol A and was administered
to control animals. Rats were stratified according to body
weight and randomly assigned to treatment groups. RIA
methods were used to measure steroid hormone concentrations
in serum or testicular fluid. RT/PCR methods
were used to examine changes in mRNA expression.
Statistical analyses included ANOVA with multiple