uch of the current policy debate about endocrine disruptors has centered on the role of the precautionary principle because there is so much uncertainty as to the effects of anthropogenic substances on the hormonal systems of humans. In fact, the precautionary principle forms a component of the "description" of an endocrine disrupter developed by the Endocrine Disruption Screening and Testing Committee. What does that mean, however?

The Principle has been defined as "when an activity raises threats of harm to human health or the environment, precautionary measures should be taken even if some cause and effect relationships are not fully established scientifically". The principle includes taking action in the face of uncertainty; shifting burdens of proof to those who create risks; analysis of alternatives to potentially harmful activities; and participatory decision-making methods.

In essence, the precautionary principle switches the questions asked in environmental and public health policy. Under current regulatory regimes we ask "How safe is safe;" "What level of risk is acceptable;" and "How much contamination can a human (usually a healthy adult male) or ecosystem assimilate without showing any obvious adverse effects?" The precautionary principle invites a new set of questions which reflects the need to avoid harm before it is done: "How much contamination can be avoided while still maintaining necessary values?;" "What are the alternatives to this activity that achieve a desired goal (a service, product, etc.)?;" and "Do we need this activity in the first place?" The questions asked about a problem determine the solutions one might seek.

While on its face, precautionary principle may seem simple, in practice the principle has a rich, nuanced complexity. Some critics feel that implementing the principle will create added costs to society and that that the principle leads to decisions not based on "sound science." The complexity of the principle gets to the heart of traditional environmental science and decision-making and requires an expanded repertoire of scientific and policy tools.

Endocrine disruption provides an excellent case for implementing the precautionary principle for several reasons:

• While scientific understanding of how synthetic chemicals might affect the endocrine system is growing our knowledge is still very limited. There is great uncertainty about which, how, and to what extent synthetic chemicals disrupt hormone systems. This uncertainty is not simply reducible by measuring a set of generally-agreed upon biological parameters. Rather, it stems from limited knowledge about the ways that altered hormone function may be expressed at the cellular, organismic, population levels. Scientific knowledge is especially limited in understanding both exposure and the variability of responses among populations and individuals. This extremely limited scientific knowledge can be described as a situation of ignorance (we do not know what we do not know).

• While the limited evidence on the effects of endocrine disrupters may provide assurances of safety to some, this lack of evidence by no means indicates that these substances pose no risk to humans. It simply means that science has yet to fully study and understand the range of potential effects. Recent studies demonstrate that for the vast majority of high volume industrial chemicals we have minimal or no screening level toxicological information. Yet we continue to allow these chemicals in commerce and humans continue to be exposed to them with little knowledge of what we know or do not know or even what questions to ask.

• The effects of synthetic chemicals on the hormone system appear to occur at very low doses and may be cumulative or even synergistic. These effects are often time-dependent rather than dose dependent, complicating the traditional "dose makes the poison" axiom of toxicology and risk assessment. This means that low-level exposures (possibly even single exposures) during critical periods of development may lead to long lasting effects in an organism. Recent evidence indicates that the effects of endocrine disrupters may not follow simple linear dose response curves further complicating our understanding of implications of exposure to these chemicals to humans.

• Synthetic chemicals that may disrupt the endocrine system are widely dispersed in the environment. Thus, humans are exposed to at least some suspected endocrine disrupters, possibly at significant levels. The geographic distribution of potential effects spans the globe.

• The effects of synthetic chemicals on the endocrine system have the potential to be serious, irreversible, multi-generational and difficult to follow. Certain impacts of synthetic chemicals on the endocrine system may be subtle on an individual basis (for example, a reduction in sperm count) but have large implications for a population. Effects of an early exposure, for example in utero, may not be observable for many years after exposure and may be virtually impossible to trace back to the earlier exposure due to the multiplicity of confounding chemicals and stressors.

So how can the precautionary principle be included in the "description" of an endocrine disrupter and how can the principle and precautionary action be applied to potentially endocrine disrupting substances? In describing an endocrine disrupter using the precautionary principle, one must consider not only whether there is the possibility of an association (note here that we look for association and not causal relationships) between exposure and effects but also the magnitude of potential harm and opportunities for prevention. In understanding whether there is a possible association between exposure and adverse effects we must rely on multiple sources of scientific evidence and disciplines (in-vitro screens, in-vivo laboratory evidence, observational studies, structural analogs, professional judgement, etc.) to establish whether the weight of the evidence points to the possibility of effects. We must also examine subtle pieces of qualitative and quantitative evidence that provide clues as to the possibility of an association. If the magnitude of potential harm is great or there are available opportunities for prevention, it may be possible to partially or entirely bypass the costly and often contentious determination of causality that is central to current decision-making.

Early evidence taken from a variety of scientific disciplines and uncertainty should provide the rationale to act under the precautionary principle, rather than to simply study the problem further. The principle places the burden on manufacturers to demonstrate that the substances they produce will not cause unnecessary harm. For new substances this means pre-market testing. If a chemical passes tiered screening tests and regulatory review it would be allowed on the market. For novel chemicals (with unknown effects) or ones in classes with suspected health effects, the manufacturer would be responsible for securing a performance bond to cover the costs of the worst case effects should damage occur. The manufacturer would also be responsible for continuous monitoring, notification, and action should evidence point towards potential endocrine disruption at a later point in time. If the chemical were a potential endocrine disrupter, the manufacturer would be responsible for developing and analyzing alternatives to that substance.

For existing substances, evidence of potential endocrine disrupting effects would lead to the application of various regulatory tools including: toxics use reduction planning, chemical use restriction or phase-outs, information disclosure, or financial mechanisms (such as taxes or liability) The manufacturer would have the burden of demonstrating that such measures are not necessary. Measures would be more or less strict depending on the extent of exposure, types and magnitude of potential effects, economic impacts of regulatory measures, and availability of safer alternatives.

Because of the sheer number of substance that might be endocrine disrupters and the resources and time needed to study each one, a precautionary approach would encourage policy makers to make decisions based on types and classes of chemicals rather than each individual chemical. For example, the Swedish government issued a policy of phasing out persistent and bioaccumulative substances in products by the year 2007 because of their inherent characteristics. They have established a long time frame to plan for the potential social and economic impacts of such a transition. Goal setting, as is done in Sweden, is one way of implementing the precautionary principle that encourages innovation and planning rather than simply analysis and prediction of future impacts.

Precaution would also encourage manufacturers and decision-makers to examine the "service" provided by synthetic substances that are potential endocrine disrupters. For example pesticides provide insect control and nonylphenol provides lubrication. Are there other ways that these "services" could be provided with safer materials? Indeed, under a precautionary approach, potentially safer alternatives require as much scrutiny as the possible endocrine disrupters as well as continuous monitoring and action should negative effects occur at a later time.

The precautionary principle should be a guide both to say yes and no to activities. It has the potential to stimulate innovation, while addressing the limitations in science to provide definitive answers regarding causality. In the face of enormous uncertainty regarding endocrine disrupting substances, multi-disciplinary scientific approaches and innovative public policy, along with democratic values have to be unified in the decision-making process. Given the potential for endocrine disruption to cause effects at critical junctures in development (we would all agree on the need to protect the fetus and developing child) and to impact future generations, the precautionary principle dictates that we institute a priority commitment to taking caution in the face of uncertainty.