Abstract
The influence of serum lipids and proteins on partitioning of individual polychlorinated biphenyl (PCB) congener peaks between human serum and adipose tissue lipid was assessed using regression analysis. Subjects were 55 repair workers who were either currently or previously exposed, and 56 comparison workers without occupational exposure to PCBs. Seven congeners (2,4,5,4'-tetra CB, 2,3,5,2',3',6'-hexa CB, 2,3,4,2',3',6'-hexa CB, 2,3,4,2',4',5'-hexa CB, 3,4,5,3',4'-penta CB, 2,3,4,5,2',3',4'-hepta CB, and 2,3,4,5,2',3',4',5'-octa CB) which had been quantified in both serum and adipose tissue in at least one-third of the total study population were selected for evaluation. Initially the crude correlations between the serum PCB and certain candidate variables were assessed; more than one congener was correlated with serum cholesterol, albumin, triglycerides, low-density lipoproteins, very low-density lipoproteins, age, body fat content, and average servings of fish per day. Stepwise regression of log serum congener concentration on log adipose congener concentration and these variables was performed. Only cholesterol, albumin, and average servings of fish per day were significant for at least one congener peak. Congeners behaved in two groups, depending on their order of chromatographic elution. For 2,4,5,4'-tetra CB, 2,3,5,2',3',6'-hexa CB, and 2,3,4,2',3',6'-hexa CB, log serum concentration was not significantly correlated with log adipose congener concentration, and there was no consistent pattern for significant candidate variables. For congeners 2,3,4,2',4',5'-hexa CB, 3,4,5,3',4'-penta CB, 2,3,4,5,2',3',4'-hepta CB, and 2,3,4,5,2',3',4',5'-octa CB, log serum congener concentration was consistently significantly correlated with log adipose congener concentration and serum cholesterol positively, and with serum albumin negatively. For these four congeners the explanations of variation (R2) in serum congener concentration using adipose congener concentration alone were 3, 23, 20, and 30%, respectively, and after adding cholesterol and albumin were 13, 56, 43, and 46%. Thus we conclude that serum cholesterol and albumin can influence the distribution or partition of PCBs between serum and adipose tissue.
Collapse