Polychlorinated biphenyls (PCBs) are a class of 209 chemicals with varying biological actions; the members of the class are referred to as congeners. Humans are exposed to many of these; in human tissue samples, depending on the chemical methods used, the number of congeners that can be detected ranges as high as 60 or 70. When studying health effects of PCBs, the appropriate exposure measure to use is not clear. Most studies either have used a measure of total PCBs or have examined selected individual congeners. In deciding what measure to use, it is important to realize that exposures of humans are to mixtures of relatively standard proportions rather than to individual congeners independently. This means that persons with high exposure to one PCB will tend to have high exposures to others as well. The resulting correlations among congeners impact the conclusions that can and should be drawn from studies of health effects, so it is important to know what the correlations usually are. We have examined correlations among 38 PCB congeners previously measured in milk from 497 Canadian mothers. We found that a large group of congeners were highly intercorrelated, so that separate health effects in humans could not be distinguished. We found that another group of congeners were so frequently below the minimum concentration quantifiable in the assay that few conclusions could be drawn about their human health effects. A third group of congeners were detectable in reasonable numbers of individuals and had low correlations with other congeners. This latter group has not always been included in epidemiologic studies, and should be considered for inclusion in future studies. Relationships of PCBs with organochlorine pesticides were also examined, and correlations were generally much lower.
Showing the most recent 10 out of 14 publications
