| 5 | WATCHING BRIEF ON ENDOCRINE DISRUPTORS |
The Commission directed the Great Lakes Science Advisory Board, in its December 1997 memorandum concerning Priorities for 1997-1999, to keep a watching brief on the issue of endocrine disruptors. A watching brief is intended to update the Commission on the most recent developments occurring with respect to an issue relevant to Great Lakes research or policy. Governments have reviewed the evidence concerning endocrine disruptors (United States Environmental Protection Agency 1997; Health Canada 1998) and described their research programs (Reiter et al. 1998). In 1996 the U.S. Congress amended the Food Quality Protection Act and the Safe Drinking Water Act to require the U.S. EPA to establish a screening program to determine which pesticides and other substances might have effects on human endocrine systems. The U.S. EPA set up the Endocrine Disruptors Screening and Testing Advisory Committee (EDSTAC) that met over the following two years and reported in September 1998. U.S. EPA published its proposed endocrine disruptors and screening program, largely based on the EDSTAC report in December 1998 in the Federal Register. In Europe, the European Commission set up a Working Group on Endocrine Disrupters under the Scientific Committee on Toxicity, Ecotoxicity and the Environment (1999) that concluded that existing guidelines for toxicology testing were not able to detect all endocrine disruptors and recommended enhanced or new testing guidelines. Through the Environment Directorate of the Organization for Economic Cooperation and Development (1997), an appraisal has been undertaken of test methods for detecting chemicals that disrupt sex-hormones and a survey conducted of the regulatory activities in member countries. Similarly, the Japan Chemical Industry Association (1997) undertook a review, on a global basis, of the scientific and regulatory strategies for chemicals that mimic hormones and put forward proposals for research, including international cooperation.
The SAB has reported extensively on this issue over the past 10 years in its reports to the Commission. During the decade, the various investigations about the effects of persistent toxic substances in the Great Lakes on human health have been continually reviewed, integrated and published in the scientific literature and disseminated in more accessible documents for the public (Johnston et al. 1998). There is, however, a convergence of the various pieces of scientific evidence, and this convergence supports a conclusion of the widespread occurrence of chemicals that interfere with the endocrine system and of the associated effects. These pieces of scientific evidence include the striking consistency between the epidemiological evidence and the experimental studies undertaken to investigate toxicity and mechanisms of action. In addition, exposure of human and wildlife populations to chemicals that interfere with the endocrine system has lead to reproductive and developmental effects. The issue of chemicals in the Great Lakes frequently relates to various aspects of endocrine disruption and can be addressed as three central questions.
What Constitutes an Adverse Effect?
In the early 1980s, a cohort of infants was established in western Michigan to investigate the effects of maternal consumption of Lake Michigan fish prior to and during pregnancy (Jacobson et al. 1984). Subsequent testing showed deficits in memory (Jacobson et al. 1985) and cognition (Jacobson and Jacobson 1993, 1996). Ten years later, a second cohort was established at Oswego, in upper New York state, to replicate the original study with infants whose mothers had eaten Lake Ontario fish (Lonky et al. 1996). These findings are consistent with studies on other developmental neurotoxicants with similar or other modes of action. For example, it is well established that children who are exposed to lead, either from gasoline or from paint, suffer a loss of I.Q. points, and there is a well-established dose-response relationship between the amount of lead in the blood and these deficits in the development of intelligence (Agency for Toxic Substances and Disease Registry 1988; Centers for Disease Control 1991). A child who is moderately exposed to lead will not show signs of disease that could be diagnosed, but may suffer an irreversible loss of development of normative functioning, such as is measured by I.Q. (Needleman and Gatsonis 1990).
These several studies show that infants and children who are more highly exposed to pollutants, particularly from maternal consumption of fish prior to and during pregnancy, have suffered loss of cognitive development. More recent evidence indicates that there are other normative functions that are affected by exposures to pollutants and that would not be diagnosed using the disease model of medicine. These normative functions include memory, emotion, response to frustration and decision making. There is a need for a methodology for assessing the effects of endocrine disruptors on community health, as recommended in the section on The Definition of The Threat to Human Health.
There is no doubt, based on the Great Lakes monitoring data, that the levels of persistent toxic substances have declined in the past 30 years (Pekarik and Weseloh 1998). However, there has been an increasing knowledge of the dangers posed by exposures to these substances. There have been questions posed as to whether the risks from persistent toxic substances in the Great Lakes are higher than elsewhere. There is now a much better understanding of the subtle effects on the endocrine system, neurological development and functioning (Colborn et al. 1998), and the development of the immune system (Voccia et al. 1999) from low level exposures to a range of chemicals. The replication of the findings from the Michigan cohort in the studies of the Oswego cohort underscores the robust nature of this association. The findings of other behavioural sequelae in the more highly exposed infants from the Oswego cohort reemphasize the significance of these exposures for those involved in human health protection. In addition, there are indications that these biologically significant effects could occur in human populations at levels of exposure close to background (Johnson and De Rosa 1999).
There is considerable evidence for the effects of critical pollutants on human reproduction at existing concentrations in the Great Lakes. In one study being undertaken in New York state, there is a significant reduction in the timing of the menstrual cycle in women who consumed more than one meal per month of contaminated fish from Lake Ontario (Mendola et al. 1997). In another study (Courval et al. 1999), which is being undertaken in Michigan, about 15 percent of 625 married couples, who were trying to conceive for at least 12 months, were unsuccessful. It seems that the strongest association is found between the increase in the rate of conception failure in these couples and the fish consumption by the men. Though both men and women consume Great Lakes fish for most of their reproductive years, men tend to consume much more fish than women.
There are several substances in the Great Lakes that interfere with the development of the immune system in exposed organisms, including humans (Weisglas-Kuperus et al. 1995), resulting in greater susceptibility to bacterial and viral infections and to cancer, as well as inducing abnormal immune responses, such as hypersensitive reactions, including allergies and autoimmune diseases. While most research on immunocompetence has been undertaken on gulls, terns, seals, polar bear, beluga whales and osprey, the focus is now on populations that are heavily dependent on wild fish and game for cultural and economic uses and for sport.
There is evidence of auditory impairment in Inuit children in the Arctic exposed to contaminants from consumption of wild foods (Julien et al. 1987). Similarly, from laboratory experimentation with rats exposed to high doses of Arochlor 1254, there is evidence of auditory loss (Goldey et al. 1995). Several species of organisms, such as whales, use low frequency sound for communication and echo-location. From laboratory experiments with rats, researchers have established that the mechanism of action of this auditory loss is through the PCBs causing a decrease in the circulating levels of thyroxine (hypothyroxinemia). The same effect can be induced with propylthiouracil that interferes with thyroxine production, and can be reversed by feeding thyroid hormone. During the EDSTAC meetings, there were questions about the specific mechanism, within the thyroid economy, that lead to this hearing loss. There are twelve known ways that chemicals can interfere with thyroid production, transport, transformation, metabolism and excretion. This instance poses the question of the level of specificity required for decision making to protect the public good.
This particular instance demonstrates the serious difficulties for regulatory officials in deciding when there is enough information to prohibit the manufacture, distribution, use, sale and disposal of a chemical or when to undertake preventive and remedial action. For example, during the EDSTAC meetings, there was extensive discussion of the definition of endocrine disruptors and particularly of what constituted an adverse effect. The evidence from the Jacobsons' research (Jacobson and Jacobson 1996) of a loss of 6.2 I.Q. points indicates that the result of the exposure of the infants was a deficit in cognitive function. However, it has been argued that the loss of 6.2 I.Q. points is within two standard deviations of the norm and therefore the deficit cannot be described as an adverse health effect. The question posed by this conclusion is whether this definition of an adverse effect is adequate to protect the public good from the effects of endocrine disruptors (Colborn and Clements 1992; Colborn et al. 1998, 1999; Johnson et al. 1998; De Rosa et al. 1999).
Irreversible Effects at Low Doses and Some Limitations of Testing
Much of the evidence cited above is from epidemiological research undertaken decades after the substances have been released to the Great Lakes and after exposures of toxicological significance have been shown retrospectively to have occurred. In preventive rather than restoration terms, experimental toxicology is the approach used to warn of the potential dangers posed by chemicals. It is only after chemicals have been released and have caused damage that epidemiology can be used. During the EDSTAC discussions, the participants addressed recent evidence that has shown the limitations of experimental toxicology, undertaken at high doses, in protecting human health, particularly from chemicals that disrupt the endocrine system. Traditionally, when chemicals were tested for their toxicity, high doses were administered to laboratory animals and the frequency of health endpoints such as cancer, mutations, birth defects and obvious neurotoxicity were determined in the exposed animals.
Concern has shifted to endpoints that are more cryptic; measured as shifts in development of the functioning of the endocrine system resulting in altered sex steroids and changes in thyroid economy, and in the functioning of the reproductive, nervous and immune systems (Colborn et al. 1998, 1999). These changes can undermine an individual's potential, and they occur at very low doses (Welshons et al. 1999) but they neither shorten the life of the test organism nor produce the traditional endpoints of concern. Evidence now exists that a number of the contaminants found in the Great Lakes fit into the latter category. First, there appears to be no threshold at which they do not effect a change. Second, the physiological effects are manifested at very low doses that in the past were considered safe, but that are at concentrations that occur in the environment. And third, depending on time of exposure, the effects at these low doses are expressed in a non-linear, dose-response manner in that they produce an inverted U-shaped curve in which effects appear at low doses that do not appear at high doses (vom Saal et al. 1995; Welshons et al. 1999). From a mechanistic standpoint, at the high doses, the feedback to the brain shuts down the receptors so that there is no longer an effect.
EDSTAC recommended testing for estrogenic/anti-estrogenic, androgenic/anti-androgenic and thyroid activity. Recent publication of the papers from the work session on health effects of contemporary-use pesticides in the Journal of Toxicology and Industrial Health underscores the difficulty of undertaking this kind of research and testing. For example, methoxychlor, which is a substitute for DDT, causes enlargement of the prostate gland as though methoxychlor were an estrogen (Gray et al. 1999). In contrast, a series of assays for anti-androgenic activity has been performed with several new fungicides, plasticizers and some old insecticides, including methoxychlor, DDT and its metabolite DDE. In this assay, methoxychlor acted as an anti-androgen and blocked the androgen receptors in the developing male animal so that it did not develop like a male. These are the kinds of problems that have been occurring in vertebrates in the Great Lakes, and that have not yet been successfully investigated and remedied. While short-term assays for screening chemicals for estrogenic/anti-estrogenic, androgenic/anti-androgenic and thyroid activity are useful and economical, the endocrine system is much more complex than represented by these biochemical assays and there is a need to test chemicals using an embryo assay to test for a wide range of endocrine effects.
Testing of Mixtures of Substances
Recent evidence has been published showing that the mixtures of a pesticide, a herbicide and a fertilizer exhibit interactive effects on the endocrine, immune and behavioral systems (Porter et al. 1999). These effects suggested to the authors that there were deficiencies in testing requirements for pesticide registration and associated implications for human health of present trends in pesticide use. The situation is further complicated by the recent discovery that components of plastics that have been manufactured and used in increasing quantities since the Second World War, have now been shown to be endocrine disruptors. While it is clearly impossible to examine all possible mixtures of chemicals experimentally (Carpenter et al. 1998), it may be practical to test mixtures of chemicals that are likely to be associated with each other from specific activities such as crop rotation or tillage practices, or effluent discharges for their interactive threats to human health.