Fish ingestion and congener specific polychlorinated biphenyl and p, p'-dichlorodiphenyldichloroethylene serum concentrations in a great lakes cohort of pregnant African American women

Abstract

A cohort of low income, city dwelling, pregnant African American Women (delivered from 1994–1999) was assembled to identify factors related to organochlorine exposure through consumption of Great Lakes resources. The cohort is known as the Great Lakes Cohort of Pregnant African American Women (GLCPAAW). Pregnant women from metropolitan Chicago, IL area clinics were administered a questionnaire on diet, demographics, and health history. Weight, height, and serum lipids were measured at delivery along with serum organochlorines such as PCBs and DDE. Congener specific concentrations of PCBs and p, p'-DDE found in the maternal serum are reported. Dominant PCB congeners found in the serum of the pregnant women at delivery included PCB 101, 118, 138, 153, and 180. The high prevalence and magnitude of PCB 101 (greater than the limit of detection in > 80% of the women in the cohort) are unique characteristics of this cohort. Great Lakes fish has been identified as a source of exposure to organochlorines in several studies. Spearman correlations and robust regression models were utilized to identify the impact of Great Lakes fish ingestion on cohort serum organochlorine concentrations. Several potential confounders of the relationship between serum organochlorines and Great Lakes fish consumption were identified. Covariates related to organochlorines in correlations as well as regression models included age, body surface area, fish ingestion, lipids, parity, race and smoking. Lower chlorinated PCB congeners do not follow the same trends as the higher chlorinated congeners and DDE. The higher chlorinated PCB congeners (PCB 138, 153, and 180) and DDE were correlated with age while the lower chlorinated congeners were not. PCB 153 and 180 regression models included age as a significant covariate. None of the higher chlorinated congeners correlated to race, while both lower chlorinated congeners were correlated to race. Race was also significant in both lower chlorinated congeners' regression models. PCB 101, a lower chlorinated congener seldom found in human serum, is readily found in the cohort. Airborne PCB exposure as well as diminished metabolism of PCB 101 in African Americans may explain the increased presence of PCB 101 and it's correlation with race. High end sport fish consumers (≥ 1meal per week) carried elevated levels of DDE and higher chlorinated PCB congeners (138, 153, and 180) compared to non-sport fish eaters. Unexpectedly, DDE was correlated more consistently with fish ingestion and age (a marker of bioaccumulation) in comparison to PCBs. Small correlations were found between serum PCBs and fish ingestion (Spearman correlation = 0.19 for total PCBs and fish meals per year). Additionally, Serum PCBs in low end Great Lakes sport fish consumers were not higher than non-consumers. These findings suggest the women of the cohort are being exposed to PCBs through other routes in addition to Great lakes sport fish. One major route of exposure may be Chicago air. The observed trends amongst individual PCB congeners has important ramifications because lower chlorinated congeners or their metabolites may be mediators of toxicity. Organochlorine exposure through Great Lakes fish ingestion was clearly identified in high end fish consumers while associations with race, metabolism, and possible airborne exposures pose new questions.