et al. find that bisphenol A at affects characteristics of
the adult rat testes at levels as low as 20 µg/kg (parts per
billion), including a reduction in sperm count. The most surprising
aspect of this finding is that the effects were seen in rats exposed
to BPA after puberty. Because adult processes are thought to
be less sensitive than developmental process, this is of concern
not just because of the direct effect on the testis, but also because
it implies that more sensitive lifestages will be sensitive to even
lower levels. Impacts of in utero low-level exposure to BPA
have been noted for other endpoints (e.g., prostate
characteristics) as well as sperm count.
did they do? Sakaue et al. exposed 13-week old rats for
six days to different concentrations of BPA, delivered by gavage.
They compared the impacts of exposure to control animals, measuring
a series of parameters that reflect the condition of the testis,
including weight of testis, daily sperm production, and protein
action. The experiments were conducted in several rounds: first
at 200 mg/kg of BPA, and then because of the positive results at
this level, at a much lower ranges (down to 2 µg/kg).
also performed an analysis of protein expression in the testes of
males exposed to a single dose of BPA at 20 µg/kg. The analysis
was done using two-dimensional polyacrylamide gel electrophoreisis.
did they find? In their experiments examining sperm count, testicular
weight and testicular structure, Sakaue et al. found significant
affects of BPA exposure at levels as low as 20 µg/kg with
hints (not significant) in the data at effects even down to 0.2
µg/kg (see figure, below).
important background finding necessary to interpret their results
is that in the control animals, the testis continue to develop from
day 13 to day 18. Hence while these rats are post-pubertal, their
testes are still developing.
measurements of daily sperm production (right) revealed that
significant differences could be observed at 20 µg/kg.
In this experiment, BPA dose ranged from 2 nanograms per kg
(parts per trillion) to 2 milligrams/kg (parts per million).
A dose-response relationship is suggested even at 200 ng/kg.
differences from controls are indicated by the red dots (one
dot, p <0.05; 2 dots, p < 0.005). Two
control groups were used, one with no treatment, the other
exposed to the mixture of ethanol and corn oil that was used
to deliver the BPA, but without BPA.
analysis of details of the impacts of BPA led them to conclude that
the effect of the contaminant was to BPA "suppressed the increase
in germ cells per Sertolli cell" that normally takes place
as the testis matures from week 14 to week 18.
protein analysis using gel electrophoresis revealed changes in the
expression patterns of three proteins. "These changes were
repeatedly observed in different rats, indicating that administration
of 20 µg/kg could affect protein expression in the adult testis."
does this mean? According to the authors, "we show here
for the first time that environmental endocrine disrupters such
as BPA alter spermatogenesis in a linear manner in a dose range
which is perhaps relevant to the daily level of exposure in man."
findings add to the accumulating evidence indicating that regulatory
agencies should immediately reassess acceptable exposure levels
for people. Acceptable daily intake in the United States and
Japan is now set at 50 µg/kg, more than twice the level
observed here to cause adverse effects.
A website, 23 December 2001) that human exposure to BPA is far
lower than anything that is likely to cause adverse effects. They
continue to contest accumulating evidence that BPA at very low levels
does have biological impacts in animals, giving credence to negative
studies from their own laboratories and dismissing positive studies
from independent laboratories. This
same pattern prevailed for years in public health debates about
tobacco smoke and about lead.
claims that long experience with BPA has proven its safety.
This is simply not true. The epidemiological studies that would
test for the adverse effects observed in animals have not been done
in people. What the absence of data on this proves is ignorance,