Correlations among Human Plasma Levels of Dioxin-Like Compounds and Polychlorinated Biphenyls (PCBs) and Implications for Epidemiologic Studies

In studies of the potential health effects of background-level exposure to organochlorine compounds (e.g., polychlorinated biphenyls, polychlorinated dibenzodioxins, and polychlorinated dibenzofurans), investigators have often measured either polychlorinated biphenyls or polychlorinated dibenzodioxins/polychlorinated dibenzofuransbut not both. We measured polychlorinated biphenyls (including specific non-, mono-, and di-ortho congeners) and specific polychlorinated dibenzodioxins/dibenzofurans among 63 Canadian blood donors. Levels of these compounds were, in general, fairly correlated. For example, Pearson's correlation coefficient between log total polychlorinated biphenyl and log total polychlorinated dibenzodioxins was .52. These results suggest that in epidemiologic studies of health effects of background-level exposures to these compounds, the quantitative dose-response relation observed for a given compound (or class of compounds acting through a similar mechanism) may easily be miscalibrated or confounded.