Endocrine Disruptors

Generations at Risk (Greater Boston PSR) -- How Environmental Toxins May Affect Reproductive Health in Massachussetts

EPA's Endocrine Disruptor Screening and Testing Advisory Committee Home Page (EDSTAC)

Links to Other Sites

Endocrine Disruptors:
The State of the Science

Ted Schettler MD, MPH
Greater Boston Physicians for Social Responsibility

What is an endocrine disruptor?

The endocrine system utilizes circulating hormones to help integrate the functions of individual organs and the nervous and immune systems. Complex interactions among these integrating systems are responsible for control, regulation, and maintenance of homeostasis, reproduction, development, and behavior.

Hormones exert their effects through binding to a specific receptor. In turn, the activated receptor initiates a cascade of biochemical events. In many cases, the hormone-receptor complex interacts directly with DNA, triggering production of gene products. An endocrine disruptor is an agent that produces reversible or irreversible biological effects in individuals or populations by interfering with hormone function.

These agents may act though any of a number of mechanisms. They may interfere with the synthesis, storage, release, secretion, transport, elimination, binding, or action of endogenous hormones. They may temporarily or permanently alter feedback loops involving the brain, pituitary, gonads, thyroid gland, or other organs. Their action is not limited to receptor binding.

Analyses which examine only receptor binding ignore non-receptor mediated events such as effects on transport proteins or hormone metabolism. Limited analyses may also fail to demonstrate that weak hormone agonists may, in the whole animal or at certain doses, actually behave as hormone antagonists by blocking a receptor from occupancy by a more potent compound.

Timing, duration, and amount of exposure. Organization vs. activation

Timing, duration, and amount of exposure are each important determinants of the outcome. There are windows of vulnerability during fetal development in which small exposures to endocrine disruptors may have profound effects not observed in adults.

Studies of the intrauterine position of mice during fetal development show that slight fluctuations of steroid hormone levels influence genital morphology, timing of puberty, sexual attractiveness, sexual behavior, aggressiveness, and activity level of offspring.

Also in mice, a single small maternal dioxin exposure on day 15 of pregnancy causes delayed testicular descent, impaired sperm production, abnormal hormone levels, and altered behavior in male offspring.

And in humans, DES daughters who developed clear-cell vaginal and cervical cancer were exposed in utero during the first 16 weeks of pregnancy.

These observations highlight the differences between organizational effects in fetal life and activational effects in adult life. Although the adult may be able to compensate for small exposures through established feedback mechanisms, the developing fetus does not have the same capacity - and in some instances, in utero exposures that would have no effect in adults may permanently alter certain characteristics -- including reproductive, neurological, and immune system parameters.

Dose-response curves

The dose-response curves for endocrine disruptors sometimes vary from those to which we are accustomed.

Male mice exposed in utero to DES, DDT, or methoxychlor, an organochlorine pesticide, show aggressive behavioral changes. They mark their territorial boundaries with urine more vigorously than control animals. This effect is reduced rather than increased at the larger doses of DES.

This unusual dose-response is also seen in rats exposed to PCBs in utero. Very small doses have no effect on litter size or physical development but cause impaired learning and hyperactivity in offspring. Higher exposure levels have no impact on learning and activity, but litter size is decreased and neuromuscular maturation delayed.

Adverse effects

The question of what represents an "adverse effect" is a matter of considerable debate and disagreement. Just what is an adverse effect? How certain are we that we will know adverse effects when we see them? What do we look for? In individuals? In populations?

The answers depend on whom you ask? For some, there is little disagreement. We would probably all agree that young adult cancers or learning disabilities resulting from in utero exposures are adverse. But there is no general agreement about other observations.

For example, male or juvenile fish exposed to alkylphenol polyethoxylates, estrogenic substances commonly found in sewage effluent, produce two proteins normally produced only by female fish - a yolk protein, vitellogenin, and an eggshell protein. We do not yet know if these chemicals have impacts on the reproductive success of these fish. Does that mean that these observations are unimportant, or should they be interpreted as a signal that may represent important but incompletely understood effects?

Historical perspective

Historically, when considering agents which are hormonally active, we have sometimes failed to listen to answers to questions we have asked - or failed to ask the right follow-up questions. For example, from 1933 to 1938, a series of journal articles from laboratories in Great Britain reported estrogenic properties of a series of chemicals, some of which, like polyaromatic hydrocarbons, were also carcinogenic. The investigators were studying the estrogenic activity of compounds which lacked the basic ring structure of estradiol at the same time that they were refining their understanding of the basic estrogen molecule.

Among the compounds tested, which showed estrogenic activity in animal tests, was one we now call bisphenol-A.Today it is a major component of polycarbonate plastics, epoxy resins, and flame retardants. More than a billion pounds of bisphenol-A are produced annually in the US, Europe, Japan, Taiwan, and Korea. Epoxy resins containing bisphenol-A are used to coat the inside of food cans and as dental sealants. Laboratory tests show that bisphenol A leaches out of the epoxy coating and sealants into food and saliva. How large a concern is this? We don't yet know. But since 1938 we have known that bisphenol-A is estrogenic.

DDT was recognized to have estrogenic effects in chickens in 1950. Though it is now banned from use domestically, one ton of DDT per day passed through US ports in 1996. Large quantities are produced and used elsewhere in the world.

The diethylstilbestrol (DES) story is, of course, another tragic example of failure to understand the consequences of human exposure to a hormonally-active substance. DES was given to millions of pregnant women for over twenty years before its adverse effects in DES daughters and sons were recognized and its failure to do what it was intended to do was acknowledged. In the early 1950's, soon after human use of DES was initiated, investigators performed a double-blind, placebo-controlled study on the therapeutic value of DES during pregnancy. The study clearly indicated that DES did not prevent spontaneous abortions. In fact, DES use was associated with increases in abortions, neonatal deaths, and premature births - and we now know of the wide range of other effects which became apparent after birth. If the unusual malignancies had not occurred and finally been recognized in young women, how long would it have been before other adverse effects were recognized and attributed to DES exposure?

A diverse and ubiquitous group of compounds

The list of known endocrine disruptors varies in length, depending on individual interpretations of the evidence. Approximately 40 pesticides are on the list, including herbicides, insecticides, and fungicides, many of which are in widespread commercial use. They are joined by a series of industrial chemicals, some of which, like alkylphenols and phthalates, are produced in enormous quantities for multiple uses and found throughout the world's ecosystems.

Dioxins and furans are produced as unwanted byproducts of industrial activity such as medical and municipal waste incineration, pulp and paper bleaching, fuel combustion, and PVC manufacturing. Polychlorinated biphenyls (PCBs) now banned from production in the US, were used as insulators in electrical equipment and for a variety of other purposes, including as a coating on the inside of grain silos. These organochlorines bioconcentrate high in th food chain, persist for decades in the environment, and are transported globally. Their half-lives in humans are measured in years. In some areas of the country, houses are built on landfill contaminated with PCBs. General Electric waste sites along the Hudson River in NYperiodically spew plumes of PCBs into the river, renewing sediment contamination. When water levels fall, PCBs evaporate from exposed mud into the atmosphere and may then be transported globally. Dioxins, PCBs, and other persistent organochlorine compounds are found throughout the world . For some, human and wildlife tissue levels are at or above those known to be associated with adverse health effects. Significant exposure to these substances will continue indefinitely without definitive coordinated global action.


The answer to "what to look for in studies of endocrine disruptors" depends on what we are able to see and what we care about. There is always the temptation to look for what is familiar, perhaps developing new techniques of analysis. But this keeps us from looking elsewhere - at effects which are less well understood but which may be extraordinarily important.

Subtle effects on neurological development and behavior are quite difficult to study. Adverse effects which are delayed, sometimes by decades in humans, are difficult to link with early exposures. And some, such as the effects of exposures on skeletal development and maintenance, have been studied very little.

Sources of information include laboratory, wildlife, and human epidemiological studies. Population-wide trends in certain diseases and biological measurements add important data.

We know alot about a few substances like dioxin, PCBs, and DDT - and much less about many others. But these few compounds have been instructive.

Acting through the Ah receptor, in laboratory animals, in utero exposure to some dioxins and PCBs reduce fertility, testis weight, and alter hormone levels, interfering with the functioning of feedback loops which normally correct these alterations. Similar hormone level changes are seen in occupationally-exposed humans.

Some dioxins and PCBs alter thyroid hormone function, acting through several mechanisms, including competition for binding to the thyroid hormone receptor or competition for binding to transport proteins, particularly transthyretin. This may well explain neurological abnormalities seen in the offspring of animals exposed during pregnancy.

Thyroid hormone is essential for normal brain development. It increases the rate of neuronal proliferation, stimulates differentiation at appropriate times, influences neuronal migration, and stimulates the formation of axons and dendrites. Only thyroxine (T4), bound to transthyretin, is capable of entering the fetal brain. Some PCBs lower T4 levels and displace T4 from transthyretin. Longitudinal studies of humans over time show that even transient hypothyroidism at birth is associated with IQ deficits later in life.

Polybrominated biphenyl (PBB) exposure causes a significant decrease in numbers and functional integrity of T- and B- lymphocytes. Studies of Michigan farmers who were accidentally exposed in the 1970s show persistence of the immune system changes.

DDT is weakly estrogenic but its major metabolite, DDE, is a much more powerful androgen antagonist, acting by attaching to the androgen receptor and blocking it from occupancy by naturally-occurring androgenic hormones.

Vinclozolin, a fungicide used on fruits, vegetables, ornamentals, and turf grass, is transformed in soil, plants, and animals into by-products which also block the androgen receptor and behave as androgen antagonists. Similarly, some synthetic pyrethroids, a widely-used class of insecticides, block the androgen receptor or displace testosterone from sex hormone binding globulin, the carrier protein for steroid sex hormones in the blood.

Laboratory tests with alkylphenol polyethoxylates, widely used in industry and found in detergents, paints, plastics, food wraps, pesticides, and other consumer products have some degree of estrogenic activity and may be responsible for feminization of fish observed in rivers receiving sewage treatment effluent.

Some phthalates, the most abundant manmade chemicals in the environment and found virtually everywhere throughout the world's ecosystems, also behave as weak estrogens. In one experiment, male rat exposed to drinking water containing small amounts of benzyl butyl phthalate throughout their gestation and the first three weeks of life had significantly reduced testis size and sperm counts.

Wildlife studies

Wildlife studies have demonstrated a relationship between exposures to endocrine disrupting substances and abnormal thyroid function, sex alteration, poor hatching success, decreased fertility and growth, and altered behavior.

For example, tributyl tin, used in anti-fouling paint on ships, causes imposex in molluscs. Imposex is a condition in which male characteristics are imposed on genetic females. This effect is reproducible in laboratories with levels of exposure which are actually less than are sometimes found in coastal waters.

Female fish downstream from pulp and paper mills have developed male sex organs and have altered behavior. Masculinized females have tried to mate with normal females or with each other. Males were hypermasculinized, showing aggressive mating behavior. The agents responsible have not been identified with certainty, but some suspect that they are pulp-mill chemicals which are converted to androgenic steroids.

In the Great Lakes region, nearly all birds and fish - for example, herring gulls and salmon - have abnormal thyroids. The glands are enlarged and otherwise abnormal on microscopic examination. Some believe that this is the result of low iodine levels in the Great Lakes region. But because of interlake differences in iodine content, others disagree, pointing to persistant organic pollutants as a more likely cause.

Gulls are relatively resistant to eggshell thinning effects of DDT but sensitive to chemicals inducing feminization like DDT and methoxychlor. In laboratory tests, gull eggs injected with DDT and methoxychlor at levels found in the environment produce feminized male embryos.

Great Lakes gulls and terns, as well as some western gulls, have, within the past several decades, shown supernormal egg clutches and female-female pairing. Gulls in these colonies also show excessive chick mortality, birth defects, and skewed sex ratios , with an excess of females. These effects correlate with levels of persistent organic pollutants like PCBs and DDT.

Great Lakes gull and tern chicks show a correlation between suppression of T-cell- mediated immunity and the level of organochlorine contamination of their eggs. The suppression is best correlated with PCB levels, but since PCBs tend to co-occur with other persistent organochlorines, like DDT, the exact contribution of each chemical to the observation is unknown. T-cell-mediated immunity is important for resistance to certain types of infections, the more general inflammatory response, and surveillance for tumor cells. The mechanism by which PCBs impair the immune response is not well understood. It may be via the Ah receptor or through interference with thyroid hormone function.

Gulls in Puget Sound show feminization and egg-shell thinning. DDT/DDE levels are relatively low in the Puget Sound region but levels of PCBs and polyaromatic hydrocarbons are relatively high, and birds from Puget Sound have high levels of these compounds in their tissues.

Seals from the Wadden Sea in the Netherlands have diminished reproductive success, and the fish that constitute their diets are heavily contaminated with PCBs and DDE. Studies under controlled conditions show that seals from elsewhere, fed contaminated fish from this area, also experience impaired reproduction.

Male alligators from Lake Apopka in Florida, exposed to the insecticide, dicofol, contaminated with DDT, as well as other organochlorine pollutants, show permanently altered hormone levels, diminshed reproductive success, small penises, and feminization of other reproductive structures.

A common theme that runs through many of the wildlife studies is a lack of accurate exposure assessment and exposure to co-occurring mixtures of contaminants. Many of these are ecological or cross-sectional studies in which the adverse effect and mixture of contaminants are noted concurrently. To the extent that observations in the field can be combined with laboratory studies, a better understanding of precise cause-and-effect and dose-response relationships will emerge.

Human epidemiologic studies

Trends in several human diseases or biologic parameters have raised concerns that exposure to endocrine disrupting chemicals may contribute to the observed changes. The increasing incidence of breast, testicular, and prostate cancer, better understanding of how certain cancers may be modulated by exposure to hormonally-active substances, and the results of animal laboratory tests combine to form a body of evidence which is provocative and can not be ignored. Evidence that sperm counts and the quality of human sperm have declined in recent decades is mounting.

Longitudinal studies of children born to mothers exposed to PCBs prior to and during pregnancy, in which their offspring demonstrate developmental delays, including reduced IQs, add further evidence.

Breast Cancer

It is likely that breast cancer develops as a result of complex interactions among hormonal, genetic, and environmental factors. The incidence of breast cancer has increased over the past several decades, and today 1 in 8 or 9 women in the United States will develop this cancer in her lifetime. There is general agreement that a woman's lifetime exposure to estrogens is an important influence on her likelihood of developing breast cancer. Early menarche and delayed menopause are associated with an increased risk.

As for the role of environmental agents, given what we know about the importance of exposures during periods of organ development, we might expect that, in this case, exposures to hormonally-active ubstances may be particularly important during fetal life, puberty, and adolescence - the time of breast development. Normal breast development depends on complex interactions among several hormones, including estrogen, progesterone, and prolactin. There are man-made synthetic chemicals to which we are all exposed which alter the levels or activity of each of these. Our understanding is incomplete.

Several epidemiological studies of women with breast cancer show a distinct significant relationship between elevated organochlorine contaminant levels and the risk of the disease. Others do not. But breast cancer is not a single disease, and in those cancers which are estrogen-receptor positive, there seems to be a better correlation.

There is a lively debate about the importance of various mechanisms by which endocrine disruptors may increase the risk of breast cancer or other hormonally-related cancers.

Altered hormone metabolism, increases or decreases in the number of hormone receptors, changes in hormone carrier protein levels, among others, are actively being studied.

Testicular Cancer

Testicular cancer is also increasing around the world, with some countries reporting a 4-5 fold increase over the past 30 years, though it still remains much less common than other malignancies. Of human studies that have examined the risk of testicular cancer associated with in utero exposure to DES, several have found an increased risk while others have not. Problems of study design include uncertainties about the timing, duration, and level of DES exposure, as well as the relative scarcity of cases, requiring large numbers of participants for statistical analysis.

Estrogen receptors are present in the male reproductive tract of mice during fetal life and diminish at birth. In utero exposure to estrogenic chemicals causes an excess of abnormally large fetal germ cells in newborn mice. Germ cells are the cells that develop into cancer in over 90%of human cases.

Epidemiological links to early hormone exposure include the following observations:

Undescended testes (or cryptorchidism) is a major risk factor for developing testicular cancer. Fetal estrogen exposure is a known cause of undescended testes (DES provides an example)

First born males have a higher incidence of testicular cancer, and first pregnancies have higher early estrogen levels than subsequent pregnancies.

Maternal overweight is a risk factor for male-offspring testicular cancer, and overweight is associated with lower levels of sex hormone binding globulin.

Most experts agree that, although the factors are complex, early malignant transformation of germ cells probably occurs in fetal life. The cells then remain dormant until activated by hormone changes in puberty.

Prostate cancer

There is general agreement that the incidence of prostate cancer is also increasing even after taking into account early diagnosis resulting from increasing use of newly-available blood tests. Approximately 40,000 men in the US die of this disease annually.

Male mice exposed to DES, estrogen, or bisphenol A in utero develop enlargement and inflammation of the prostate. Some also develop what appear to be microscopic malignancies in their genital tract.

We have known for some time that the likelihood of prostate cancer increases with age and that many elderly men who die of other causes also have incidental prostate cancer which is unrelated to their deaths. But a recent human autopsy study shows that prostate cancer may develop much earlier than previously realized. This study found that, of 152 men who died of other unrelated causes, all less than 50 years old, 34% had evidence of cancer in their prostates when examined microscopically. Of those less than 40 yrs. old, 27% had prostate cancer. These remarkably high percentages, along with the animal data, suggest that prostate cancer may be yet another malignancy whose foundation is laid early in life, perhaps even in utero, but which does not be come apparent until later when hormone changes stimulate the dormant cancer to grow.

Sperm Counts

A series of studies in Europe, the UK, and the US indicate that there has probably been a substantial decline in human sperm counts over the past 40 years. Inherent study problems such as age, period of abstinence from sexual activity, and individual and geographical variability make this challenging to investigate, but the weight of evidence indicates a significant decline.

A recently published autopsy study of men who died of various causes unrelated to their reproductive tract compared results from 1991 with those obtained from a similar population of men in 1981, corrected for age and various other medical and social factors. In this study, the percentage of subjects with normal spermatogenesis on microscopic exam of their testes declined from 56% to 27% over the ten year period. Complete spermatogenic arrest increased from 8% to 20% over the same period.

Altered neurological development, behavior, learning

Studies in monkeys and rodents indicate that in utero exposure to dioxin, PCBs, and estrogenic chemicals can alter behavior and learning ability. Several human epidemiological studies have similar findings.

Maternal consumption of PCB-contaminated fish before and during pregnancy exposes their developing fetuses to this organochlorine during critical times of brain development.

A series of studies of ohorts of children, each consisting of between 200-700 participants who have been followed for varying lengths of time, show a distinct relationship between fetal PCB exposure and delayed motor development, abnormal reflexes, and hyperactivity.

One study of children followed for 11 years shows that those in the highest exposure group have an IQ deficit of about 6 points. Projected onto an entire population, this amounts to a large shift in the number of people at the high end and low end of a bell-shaped curve. It is a biological phenomenon with enormous social consequences.

Laboratory studies suggest that the mechanism of this adverse effect may be through interference with thyroid hormone metabolism, as discussed earlier, though this is not known with certainty.

Other health effects

There are also data linking exposure to endocrine disruptors with other medical conditions. For example, a study of primates exposed to small amounts of dietary dioxin showed dose-dependent development of endometriosis. Some laboratory animal tests show a relationship between fetal estrogenic exposures and subsequent development of auto-immunity. In general, our understanding of the relationship between fetal exposure to hormonally-active substances and effects on the brain or immune system is still very incomplete.


Laboratory, wildlife, and epidemiological studies provide a body of evidence which, though incomplete, demonstrates important health effects resulting from exposure to endocrine disrupting substances at current environmental levels.

These health effects include altered development and behavior, impaired reproductive success, thyroid dysfunction, and altered immune system function. DES caused cancer in humans following in utero exposure. Other types of cancer may be related to in utero exposures as well, but evidence is incomplete. In many cases, our understanding of mechanisms and dose-response relationships is poor and considerable research is necessary.

However, when we weigh scientific evidence, we need to explicitly recognize that the limits of science influence both the questions we ask and how well we are able to design studies to get answers. It has become almost commonplace to determine whether or not a chemical attaches to a hormone receptor, or causes a sperm count decline, or diminished reproductive success in a laboratory animal. There are reproducible and valid methods for answering those questions. However, free-ranging wildlife - or human populations and sub-populations - living in diverse environments, with complex, underlying background exposures to multiple, man-made chemicals, present study problems much more difficult to overcome.

Biological complexities include 1) interactions among the endocrine, neurological, and immune systems, 2) interactions among genetic, environmental, and social factors - interactions which we know exist and which do not lend themselves to simplistic reductionist analysis, 3) particularly susceptible species, sub-populations, and individuals, and 4) multiple, real-world background exposures overlaid with new ones .

Some endpoints may be subtle, delayed, or difficult to detect - and yet they may be the most important and the very things we care most about. Data gathering and interpretation often become more subjective and value-laden.

Our historic concentration on dioxin, PCBs, and DDT has produced a large volume of literature relating to their toxicity. Even with these chemicals, there are large gaps in our understanding of mechanisms of action. Most other compounds, some of which are in widespread commercial use, found virtually throughout the world, to which large populations of humans and wildlife are exposed, have been studied much less or not at all.

We have an understandable interest in mechanisms of action. It is part of a long-standing scientific tradition. In fact, the criteria which must be met in order to argue a cause-and-effect relationship from an epidemiological study includes identification of a plausible biological mechanism. Mechanistic understanding satisfies our intellectual curiosity, advances research, and provides arguments for philosophical and legal debates. But when we begin to use a list of criteria as the basis for making, or failing to make, public policy, we need to explicitly address embedded values and the limits of scientific inquiry.

We are engaged in a large global experiment. It involves widespread exposure of all species of plants and animals in diverse ecosystems to multiple manmade chemicals. And in flagrant disregard of the notion of informed consent which underlies medical experimentation and treatment, those exposed are frequently uninformed and have rarely given consent. Instead, the limits of science and rigorous requirements for establishing causal proof often conspire with a perverse requirement for proving harm, rather than safety, to shape public policies which fail to ensure protection of public health and the environment.

Act! | Program Update | Press Releases | Research Update | Publications | Press Releases
PSR Home
| Violence Prevention | National Security | Membership | Search | Links|

Copyrightę1997 Physicians for Social Responsibility, Inc.