A main issue of the single pollutant approach in epidemiologic research is that it is prone to confounding. For example, the health effects of PCBs are subject to confounding by methylmercury if participants were co-exposed to both toxicants from fish consumption. This example also suggests that beneficial nutrients such as omega-3 fatty acids may confound the toxic effects by PCBs and methylmercury [2], [3]. Therefore, a positive association in a single pollutant approach may be observed if the single pollutant is a proxy for other co-pollutants or a mixture of pollutants. Alternatively, if individual pollutants have relatively small effects but multiple pollutants as a whole influence the disease risk, the single-pollutant approach may not capture the true effects [4].
