Causal inference considerations for endocrine disruptor research in children's health.

Substantial population exposure to endocrine disrupting chemicals, combined with available biomarkers and public concern, has resulted in an explosion of human health effects research. At the same time, remarkable shifts in the regulations governing the composition of some consumer products that contain endocrine disruptors (EDs) has occurred. However, important questions remain as to the weight of evidence linking EDs to human health end points. In this review, we critically examine the literature linking ED exposures to child neurodevelopment, focusing in particular on two model exposures to demonstrate issues related to bioaccumulative [e.g., polychlorinated biphenyls (PCBs)] and rapidly metabolized (e.g., phthalates) compounds, respectively. Issues of study design, confounding, and exposure measurement are considered. Given widespread exposure to these compounds, the potential public health consequences of even small effects on human health are substantial. Therefore, advancing our understanding of any impact calls for careful attention to the principles of causal inference.