Abstract
BACKGROUND: Polychlorinated biphenyls (PCBs), a diverse class of chemicals, are hypothesized mammary carcinogens. We examined plasma levels of 17 PCBs as individual congeners and as a mixture in association with breast cancer using a novel approach based on quantile g-computation. METHODS: This study included 845 White and 562 Black women who participated in the population-based, case-control Carolina Breast Cancer Study Phase I. Cases (n = 748) were women with a first diagnosis of histologically confirmed, invasive breast cancer residing in 24 counties in central and eastern North Carolina; controls (n = 659) were women without breast cancer from the same counties. PCBs were measured in plasma samples obtained during the study interview. We estimated associations [covariate-adjusted odds ratios (ORs) and 95% confidence intervals (CIs)] between individual PCB congeners and breast cancer using multivariable logistic regression. We assessed PCB mixtures using quantile g-computation and examined effect measure modification by race. RESULTS: Comparing highest and lowest tertiles of PCBs resulted in ORs of 1.3 (95% CI = 0.95, 1.8) for congener 74, 1.4 (95% CI = 1.0, 1.9) for 99, 1.3 (95% CI = 0.91, 1.8) for 194, and 1.2 (95% CI = 0.90, 1.7) for 201. Among all women, we estimated a joint effect of the PCB mixture with an OR of 1.3 (95% CI = 0.98, 1.6) per tertile change. In race-stratified analyses, associations for tertiles of PCB mixtures were stronger among Black women (OR = 1.5; 95% CI = 1.0, 2.3) than among White women (OR = 1.1; 95% CI = 0.81, 1.6). CONCLUSION: Our results are consistent with the hypothesis that exposure to PCB mixtures increase the risk of breast cancer, but studies of populations with different exposure profiles are needed.