Safe's 1999 op-ed
in the New England Journal of Medicine accompanies an important, very
study involving over 120,000 women from 11 different states by
Hunter et al. The authors of this study report no relation
between the risk of breast cancer and contamination levels of PCBs
and DDE (a persistent breakdown product of DDT) in women's blood.
Safe concludes on the basis of this result that weakly estrogenic
chemicals do not increase the risk for breast cancer.
first glance, this study might appear to support Safe's argument,
but in reality it does not. The study is the best to date available
on association between breast cancer risk and adult exposure to
PCBs and DDE. In and of itself, that is very important, but it
does not warrant Safe's sweeping conclusion.
Safe's conclusions extrapolate the data from two contaminants
to all 'weakly estrogenic' compounds. This argument is illogical
to begin with, to use data from two compounds to conclude
no others are involved. Not only is it logically incorrect, and
a classic ploy used by industry interests defending their products,
the argument has been proven startlingly wrong by later
studies showing that another weakly estrogenic compound, dieldrin,
is associated with an increase in risk of breast cancer and higher
mortality rates of breast cancer victims. If Safe's argument
were correct, this result would have been impossible.
is also illogical because DDE and the most common forms of PCB
found in human tissue are not weakly estrogenic. DDE
is an anti-androgen, while the common, persistent forms of
PCBs are anti-estrogens.
Thus concluding that studies from just two compounds, neither
of which are weakly estrogenic, disprove the hypothesis that weak
estrogens are involved in breast cancer is logically ludicrous.
got his toxicology wrong here, which is surprising given his
long experience in toxicology. It's akin to concluding on the
basis of toxicology experiments with water that oil
isn't toxic. The hypothesis that weakly estrogenic compounds increase
risk of breast cancer rest on well-established observations that
life-time exposure to estrogen increases breast cancer risk. On
that basis, one would not expect DDE or PCBs to be involved in
breast cancer [although there might be other mechanisms through
which they could be involved.]
this study looks at exposure measured in adulthood in relation
to risk of breast cancer. It says nothing about developmental
exposure, either during fetal development or during puberty
(when human breast tissue is undergoing rapid development). While
it is possible that the origins of breast cancer lie in exposure
during adulthood, it is equally plausible that the antecedents
occur much earlier in life. To argue that a study finding no
relationship between adult exposures and adult risk rules out
developmental effects is obviously wrong.
fourth, Safe argues that the time trends of breast cancer vs.
DDT and PCBs argue against a role of these compounds in causing
breast cancer. Breast cancer has been increasing in the US just
as DDT and PCBs have been declining because of restrictions on
their use. Superficially this sounds plausible. But if the chemical's
impact is through disruption of development, then this argument
must be rejected, at least at this time. Women entering into
the prime years of risk for breast cancer today, in the year 2000,
were exposed in the womb and during childhood development to the
highest levels of DDT and PCBs ever experienced by any generation
in history. The rise in breast cancer rates, if associated with
developmental exposures to these compounds, should continue to
rise for at least two more decades, until women born to women
whose earlier experience was long after the peak periods of use.
ends this editorial with rhetorical flourish: "However, it
is incumbent on scientists, the media, legislators, and regulators
to distinguish between scientific evidence and hypothesis, and not
to allow a "paparazzi science" approach to these problems."
It would seem even more important to avoid science conflicted
by vested interests.