Estimating chemical risk: Breadth (prevalence) may be just as important as depth (magnitude of effect)

Jennifer McPartland, Ph.D., is a Health Scientist.

Earlier this month Dr. David Bellinger at Boston Children’s Hospital published a very interesting paper in Environmental Health Perspectives offering a new way to consider the importance of various risk factors for child neurodevelopment—such as pre-existing medical conditions, poor nutritional status or harmful chemical exposures—at the population level.  “A Strategy for Comparing the Contributions of Environmental Chemicals and Other Risk Factors to Neurodevelopment of Children” argues that, in evaluating the contribution of a risk factor to a health outcome, it is critical to consider not only the magnitude of its effect on the health outcome, but also the prevalence of that risk factor in the population.

Dr. Bellinger argues: “Although a factor associated with a large impact would be a significant burden to a patient, it might not be a major contributor to the population if it occurs rarely.  Conversely, a factor associated with a modest but frequently occurring impact could contribute significantly to population burden.”  The former “disease-oriented” approach has generally been used to estimate the burden of harmful chemical exposures to population health, rather than the latter “population-oriented” approach.  Relying solely on the former approach, he contends, may result in an underestimation of the impact of a chemical exposure or other risk factor on public health. 

Dr. Bellinger illustrates his point by examining the impact of various risk factors on neurodevelopment in U.S. children.  He first estimates the drop in full-scale IQ (FSIQ) score (the size of the risk factor’s effect) associated with various risk factors, ranging from congenital heart disease to lead exposure.  He then multiplies the effect size of each risk factor by the prevalence at which that risk factor occurs in U.S. children aged 0 to 5 years (a total of 25.5 million children), to determine the total cumulative impact of each risk factor on FSIQ loss in these children.

The eye-opening results are shown in the table below.

Dr. Bellinger notes that the cumulative population burdens associated with chemical exposures (methylmercury, organophosphate pesticides and lead) are “surprisingly” large—due not so much to their effect sizes (degree of FSIQ loss), but rather to the prevalence at which children in the U.S. experience these exposures.

Dr. Bellinger emphasizes that the estimates of FSIQ loss associated with the various risk factors should not be over-interpreted, pointing to several limitations in his calculations.  For example, some of the effect-size estimates for the risk factors were derived from studies conducted on non-U.S. samples.  There could be regional differences that modify effect size in these populations versus U.S. populations.

Nevertheless, the take-home message stands:  Population burden needs to be considered just as much individual burden when evaluating the impact of a risk factor to a health outcome.

In EDF's chemicals policy work, we emphasize that the effects of chemical exposures on human health can vary across individuals and sub-populations, and hence, we need policies that protect the most vulnerable among us, such as children and pregnant women.  Such vulnerabilities can be considered to represent the “depth” of a chemical exposure.  This paper points out the other side of the same coin, namely that the “breadth”—or prevalence at which a chemical exposure occurs—is just as important in judging its significance to public health.


This entry was posted in Health Science, Regulation and tagged , , , , . Bookmark the permalink. Both comments and trackbacks are currently closed.