Abstract

Estimates of the half-lives of PCBs in humans derived from successive body burden measurements are reviewed and found to vary widely whether based on congener-specific or aggregate data. Variability due to differences in physiological processes among individuals and in congener properties is to be expected, but does not appear to be a complete explanation. Very short half-lives (i.e. < 1 year) are unlikely for those congeners most frequently found in human blood because the exposures required to sustain observed body burdens are too large to be easily explained. Very long half-lives ( > 10 years) may be artifacts of confounding by ongoing exposures (a common effect at low body burdens) and are also suspect. The loss of significant quantities of PCBs from the blood of occupationally exposed persons with half-lives of 2-6 years is comparatively well documented (i.e. has been observed in studies with relatively large numbers of subjects and high initial body burdens). Therefore, very long half-lives must be limited to subsets of congeners or of populations if they occur at all. The impact of the use of fixed estimates of half-lives drawn from the tails of the observed range on the evaluation of regulatory standards is shown to be substantial.