ENVIRONMENTAL ESTROGENS AND REPRODUCTIVE HEALTH - A DISCUSSION OF THEHUMAN AND ENVIRONMENTAL DATA

Estrogenic activity of certain xenobiotics is an established mechanism of toxicity that can impair reproductive function in adults of either sex, lead to irreversible abnormalities when administered during deve lopment, or cause cancer. The concern has been raised that exposure to ambient levels of estrogenic xenobiotics may be having widespread adv erse effects on reproductive health of humans and wildlife. The purpos e of this review is to evaluate (a) the nature of the evidence support ing this concern, and (b) the adequacy of toxicity screening to detect , and risk assessment procedures to establish safe levels for, agents acting by this mechanism. Observations such as adverse developmental e ffects after maternal exposure to therapeutic Levels of the potent est rogen diethylstilbestrol or male fertility problems after exposure to high levels of the weak estrogen chlordecone clearly demonstrate that estrogenicity is active as a toxic mechanism in humans. High level exp osures to estrogenic compounds have also been shown to affect specific wildlife populations. However, there is little direct evidence to ind icate that exposures to ambient levels of estrogenic xenobiotics are a ffecting reproductive health. Reports of historical trends showing dec reasing reproductive capacity (e.g., decreased sperm production over t he last 50 years) are either inconsistent with other data or have sign ificant methodologic inadequacies that hinder interpretation. More rel iable historical trend data show an increase in breast cancer rate, bu t the most comprehensive epidemiology study to date failed to show an association between exposure to persistent, estrogenic organochlorine compounds and breast cancer. Clearly, more work needs to be done to ch aracterize historical trends in humans and background incidence of abn ormalities in wildlife populations, and to test hypotheses about ambie nt exposure to environmental contaminants and toxic effects, before co nclusions can be reached about the extent or possible causes of advers e effects. It is unlikely that current lab animal testing protocols ar e failing to detect agents with estrogenic activity, as a wide array o f estrogen-responsive endpoints are measured in standard testing batte ries. Routine testing for aquatic and wildlife toxicity is more limite d in this respect, and work should be done to assess the validity of a pplying mammalian toxicology data for submammalian hazard identificati on. Current risk assessment methods appear to be valid for estrogenic agents, although the database for evaluating this is limited. In concl usion, estrogenicity is an important mechanism of reproductive and dev elopmental toxicity; however, there is little evidence at this point t hat low level exposures constitute a human or ecologic health risk. Gi ven the potential consequences of an undetected risk, more research is needed to investigate associations between exposures and effects, bot h in people and animals, and a number of research questions are identi fied herein. The lack of evidence demonstrating widespread xenobiotic- induced estrogenic risk suggests that far-reaching policy decisions ca n await these research findings. (C) 1997 Elsevier Science Inc.