UCSF develops game-changing method to translate research on toxic chemicals and prevent disease


By David Tuller, Dr.PH.

Everyone wants to know more than we currently do about the long-term effects of everyday exposures to toxic chemicals. Even obstetricians, who could be expected to have a handle on the science, report not knowing how to advise their pregnant patients, according to a recent survey led by colleagues at the University of California at San Francisco’s (UCSF) Program on Reproductive Health and the Environment.

Why has it been so difficult to draw authoritative conclusions about the impacts of toxic exposures? For one thing, we lack data. In the case of pharmaceuticals, drug companies are required to prove safety and efficacy with clinical trials in which some participants are given the medication and others are not. Such experiments are impossible with chemicals. The kinds of data that the Food and Drug Administration uses to make decisions about pharmaceuticals is simply not available—and not obtainable—for chemicals.

That means relying on other sources of information. Laboratory and animal experiments can demonstrate a chemical’s negative impacts on living organisms, but interpreting what those findings mean for human beings is difficult. Human observational studies provide solid evidence of associations between chemical exposures and specific health outcomes but drawing conclusions about causes is more challenging. That’s because of the possibility that some other, unmeasured factor—another environmental exposure, a genetic predisposition, or something else—was responsible for the outcomes observed.

In the field of medicine, the push for “evidence-based” decision-making—clinical care based on solid reviews of the available data rather than on doctors’ hunches and instincts—has transformed practice in recent decades. Without the same kind of evidence base as medicine, environmental health has lagged behind in applying findings from the science to public policy. The field lacks a rigorous method for assessing the quality and strength of the data streams from both laboratory experiments and human observational studies—and emerging with conclusions that are solidly grounded in the scientific literature.

To address that methodological gap, UCSF’s Program on Reproductive Health and the Environment has been leading an initiative to develop such a method, with participation from other academic centers, government agencies, and non-governmental organizations. In four papers just published by Environmental Health Perspectives, the researchers describe the method, called the Navigation Guide, and road-test it with a widely used chemical that virtually everyone in the U.S. is routinely exposed to—perfluorooctanoic acid, or PFOA.

The goal of the Navigation Guide is to help reduce the time lapse between scientific discoveries and the development of appropriate health policies. Speeding the process of translating findings into action could help to significantly reduce disease prevalence and health care expenditures. The U.S. health care system, for example, was estimated to have spent more than $75 billion on childhood illnesses associated with environmental exposures; changed policies leading to lower exposure levels could both improve health outcomes and trim costs.

“Due to deficiencies in the current regulatory structure for manufactured chemicals, a failure or delay in acting on the science means that exposure to toxic chemicals persists while evidence of harm mounts,” reports one of the papers. “To the extent that science informs public policy to prevent harm, a robust method to synthesize what is known about the environmental drivers of health in a transparent and systematic manner is a necessary foundational step to making the science actionable.”

In the series of EHP papers, the researchers tackled the question of whether PFOA exposure affects fetal growth in humans. The chemical has been used widely in industrial and consumer products for more than 50 years, and it persists in the environment. Virtually all Americans are exposed to it, and numerous studies have examined its relationship to fetal growth.

The Navigation Guide methodology involves four steps:

  1. To specify a study question relevant to decision-makers about the health risks posed by a chemical or class of chemicals
  2. To conduct a systematic search for published and unpublished evidence regarding the study question
  3. To rate the quality and strength of the evidence of toxicity, including integrating separate reviews of the human and non-human evidence
  4. To grade the overall strength of the recommendations for proposed public health policies by integrating the evidence of toxicity with information about other factors, such as exposure levels, the availability of less toxic alternatives, and public values and preferences.

In the PFOA case study, the team was able to synthesize the human and non-human findings and issue a clear, “strength-of-evidence” conclusion: developmental exposure to PFOA, wrote the researchers, “adversely affects human health based on sufficient evidence of decreased fetal growth in both human and non-human mammalian species.” The chemical, they determined, is “known to be toxic.”

So what are the health policy implications of that direct, evidence-based statement? That question—the heart of Step 4 of the Navigation Guide—has not yet been fully explored. Creating policy from science is always a messy business, given that people approach the problem from a vast array of political, social and economic perspectives. But no effective action is possible without a neutral, authoritative statement of what the science tells us. And now, at least in the case of PFOA’s effects on fetal growth, we know.

David-TullerDavid Tuller, Dr.PH. is academic coordinator of UC Berkeley’s joint masters program in public health and journalism. He was a reporter and editor for 10 years at the San Francisco Chronicle, served as health editor at Salon.com and frequently writes about health for The New York Times.


Leave a Comment

Give us a sign that you're human * Time limit is exhausted. Please reload the CAPTCHA.