Endocrine Disruption:
Patterns in science; goals in policy
J.P. Myers, Ph.D.
Annual Meeting
American Association for the Advancement of Science
Philadelphia, PA
16 February 1998

Send us your comments!

The evolution of endocrine disruption as a scientific issue manifests six broad trends:

  1. The potential ranges of systems affected has broadened–from just those effects mediated by interactions with the estrogen receptor to include other hormone systems, sometimes receptor mediated, sometimes not, including testosterone, thyroid, retinoic acid, progesterone, etc., and indeed now to include chemical message systems mediating interactions among symbiotic organisms, including, potentially, communication between the nitrogen fixing bacteria Rhizobium and its host leguminaceous plants.
  2. The time frame for concern has been extended, from an initial focus on direct impact on the exposed adult, to in utero effects that cause life-long damage, including some patterns which may not appear until beyond puberty. This has special significance for how medical researchers and epidemiologists must approach their sciences, because it forces them to ask how in utero hormonal experience may affect health in adulthood. And this not only complicates the science, it also builds unavoidable time lags into the research.
  3. The geography of contamination has become global. The whole world is downwind or downstream, especially because of global atmospheric transport. Work with the Inuit in northern Canada, in the food chains of arctic Russia, sampling in seemingly pristine high elevation tropical habitats all show that there is no place that has escaped some level of contamination. Thus while we can be pleased that regulatory steps to eliminate exposure in the US to PCBs, DDT and other endocrine disrupters have have brought significant benefits to the US public, much remains to be done even with these classic contaminants.
  4. The level of contamination of concern has gone lower and lower as a result of careful laboratory work with developing embryos and fetuses. In what may be the most extreme example, work by Fred vom Saal at the University of Missouri suggests that bisphenol A adversely affects prostate development at levels many thousand times below current government mandated safety levels. Whether this safety factor is required needs replication; but this work clearly raises the bar significantly.
  5. The nature of the chemicals involved has broadened dramatically. For example, Perez et al. have an article in press in the journal of the National Institute of Environmental Health Sciences (Environmental Health Perspectives, March) showing for the first time that bisphenol-F is estrogenic. This is of special interest because bisphenol-F is one name for bakelite, a commercial product patented in 1909 which quickly moved into commerce as one of the first of the miracle plastics. The fact that bakelite is estrogenic pushes back the date of widespread human exposure to xenobiotic estrogens by as much as three decades, almost to the beginning of the 20th century.
  6. As health scientists and epidemiologists have looked for adverse health impacts in people, they have found patterns consistent with the hypothesis that fetal development has been misdirected by endocrine disrupters. This summer’s publication by the Centers for Disease Control of a doubling of penis deformities in US boys since 1970 is just one example. This result, it should be noted, was not discovered by accident and then associated post-hoc with endocrine disruptors. The CDC scientists chose to look for changes in the rate of a specific penile birth defect, hypospadias, because (1) it was known that endocrine disruptors produce hypospadias in laboratory rats if the exposure takes place at a key moment in fetal development and (2) the deformity is not one in which trends might be caused or obscured by changes in diagnostic procedures. Shanna Swan’s definitive study of global sperm count, published in the November 1998 issue of Environmental Health Perspectives, decline is another.

Is there scientific certainty about cause and effect in these cases? No. But Occam’s Razor positions endocrine disruption as one of the prime explanatory hypotheses: We know that in laboratory animals these same patterns can be reproduced by fetal exposure. We know that the developmental control of sperm count and genital formation in lab animals is highly similar to humans. We know that some percentage of the human population is exposed to the relevant range of contamination. It is not unreasonable to posit endocrine disruption as a possible cause. If fact, what is unreasonable is to dismiss it.

In response to these trends and the underlying science, Congress and the Administration have responded with several policy initiatives which I would regard as constructive first steps but incomplete.

Significant increases in the amount of money available for research, particularly at EPA and NIH. This is crucial and welcome. This work should help resolve which of the currently identified risks represent substantial threats to human and ecosystem health, and which can be rejected.

The administration, in part because of concern about this issue, has moved to reorient federal contamination standards to protect the most vulnerable in the population–the developing fetus and young children. This step would have been justified even without concern for endocrine disruption.

Congress passed the Food Quality Protection Act of 1996. Again, most of this would have been justified without endocrine disruption. One of the most important aspects of that legislation is the Endocrine Disruption Screening and Testing Advisory Committee, or EDSTAC. The work that committee is undertaking is fundamentally important. It has been hard work, and we owe them a great deal for having been willing to participate, especially the academic and not-for-profit scientists whose participation is an add-on to their main employment.

I am concerned, however, that the results of EDSTAC and related efforts to apply traditional risk assessment procedures to signal disrupting chemicals will most likely fail to protect public health, for two reasons.

First, the numbers of chemicals to be assessed–both those already in widespread use as well as the many new ones that enter commerce each year–is daunting.

We have a lot of experience in a chemical-by-chemical approach to contaminant regulation, and one of the principal impacts is often to enshrine ignorance because of the laborious nature of the process… we know almost nothing about most of the chemicals in widespread use, not just with respect to endocrine disruption but also about other more classic health endpoints. For example, the Environmental Defense Fund released a report last year finding that nearly three quarters (71%) of sampled high-volume chemicals do not meet the minimum data requirements for health hazard screening set by international standards.

As EDF pointed out in their study: "Absence of data is not the same as proof of harm, of course. It is only proof of ignorance." But ignorance means that any conclusion about safety is unfounded.

Second, if there is a general message to the emergence of endocrine disruption as a policy issue, it isn’t just that we haven’t been able to catch up with all the chemicals in use and the new ones coming on the market, it is also that we are caught with a scientific and regulatory structure that for decades was ignorant of what questions to ask. Experience tells us that we do not do very well at anticipating unexpected forms of toxicity, and that commerce proceeds to ramp up production and dispersion of new compounds far more rapidly than the sciences of toxicology and epidemiology can advance. The pace at which these two sciences move forward is not commensurate with the pace of commercialization of new products. The result is that we wind up performing decades long experiments on the public and we do so even lacking adequate monitoring systems to alert us to problems.

There is an alternative approach, one that is being developed by the Chemicals Policy Committee of the Swedish government, established in 1996 in Sweden to evaluate approaches to achieving an environmentally sustainable chemical industry.

The CPC begin by observing, as I just mentioned, that we do not do very well at anticipating unexpected forms of toxicity. They go on to observe that substances that are persistent and bioaccumulative create widespread exposure, both through atmospheric and water transport as well as through trade, and that when problems are finally identified the exposures cannot easily be reduced simply by discontinuing production. They conclude that for substances that are persistent and bioaccumulative, we are virtually guaranteed to find that knowledge of adverse impacts will come to late to avoid widespread damage.

For these reasons, the Swedish Chemicals Policy Committee has proposed a series of targets for the government’s chemical policies:

Specifically with respect to endocrine disrupters, the Chemical Policy Committee has proposed that substances that meet criteria established for endocrine disrupters be phased out, without the need to achieve scientific certainty in demonstrating harm. Period. Even more dramatically, the Committee recommends that polyvinyl chloride plastic materials be phased out from commerce no later than 2007.

This approach is radically different from that being used in the United States, because it begins with the premise that precautionary measures should be taken in the face of significant potential harm even without scientific certainty. They go farther, however, than the most timid applications of precautionary measures, by acknowledging that if our century’s experience with chemical safety issues has demonstrated anything, it is that we do very poorly at anticipating deleterious effects mediated by unanticipated mechanisms of toxicity.

The SPC’s targets are directed toward industry. They are ambitious. They are intended to generate driving forces for change in the choice of chemicals and products and in requirements for products.

They have chosen a long time frame for change, out to the year 2012. They judge that ten to 15 years is a reasonable lead time for targets related to restrictions and phase-outs. Were the timetable longer, then motivation to implement new design criteria would be insufficient. Were it shorter, the targets would be viewed as unrealistic.

The targets focus on sending market signals with enough lead time to allow manufacturers to develop and implement new design criteria. The fact is that–with the significant exception of pesticides–rarely do products of the modern chemical industry require biological activity, environmental persistence or bioaccumulation as a design objective of products. By being explicit about the need to employ design criteria that explicitly eliminate these characteristics, we are more likely to encourage the development of a sustainable chemical industry which can continue delivering the benefits of modern chemical synthesis with much less risk.

Send us your comments using the form below:
What is your name? (not required)
What is your e-mail address? (not required)

Your comments:

Return to Earth & W. Alton Jones Foundation