Persistent organic pollutants and maternal glycemic outcomes in a diverse pregnancy cohort of overweight women

Abstract

Background: Animal and human studies suggest certain persistent organic pollutants (POPs) may impact glucose metabolism; however, few epidemiologic studies have examined environmental determinants of glycemic outcomes during pregnancy. Our objective is to evaluate associations between exposures to individual and mixture of POPs and measures of prenatal fasting glucose, insulin, and insulin resistance during pregnancy in overweight women.

Methods: A cohort of overweight and obese pregnant women (N = 95) was recruited from California. Blood samples were collected during late first or second trimester (median = 16 weeks' gestation; range = 10-24 weeks). Exposures included serum concentrations of polybrominated diphenyl ethers (PBDEs) and hydroxylated metabolites (OH-PBDEs), polychlorinated biphenyls (PCBs), and poly- and perfluoroalkyl substances (PFASs). Outcomes included serum concentrations of fasting plasma glucose, fasting plasma insulin, and calculated homeostatic model assessment of insulin resistance (HOMA-IR). Generalized linear models were used to evaluate cross-sectional associations between individual and aggregate POPs and mean percent difference in fasting glucose, fasting insulin, and HOMA-IR. Bayesian kernel machine regression (BKMR) was used to assess the relative importance of each exposure to the association with our outcomes, using conditional and group posterior inclusion probabilities (PIPs).

Results: Study participants were racially/ethnically diverse and nearly half were below the federal poverty level. Across PBDEs and OH-PBDEs, the direction of associations with fasting glucose, fasting insulin and HOMA-IR were varied. A doubling of PCB-138, PCB-153, PCB-180, and ∑PCBs concentrations was associated with a 2.10% mmol/L (95%CI: 0.49%, 3.74%), 2.10% mmol/L (95%CI: -0.14%, 4.39%), 2.10% mmol/L (95%CI: 0.12%, 4.12%), and 2.81% mmol/L (95%CI: 0.38%, 5.31%) increase in fasting glucose, respectively. Exposure to individual PCBs was positively associated with both fasting insulin and HOMA-IR. All PFAS were inversely associated with fasting glucose, fasting insulin, and HOMA-IR. In BKMR models of fasting glucose, all four chemical classes were important contributors to the overall mixture, with PFASs identified as the most important contributor.

Discussion: Prenatal PCB exposure was positively associated while certain PBDE and PFAS analytes were inversely associated with fasting glucose concentrations in overweight women. Further examination of the relationship between POPs exposure and glycemic functioning in a larger study population of women during pregnancy is warranted.