Temporal trends and determinants of serum concentrations of per- and polyfluoroalkyl substances among Northern California mothers with a young child, 2009–2016

Abstract Background/Objective Human exposure to per- and polyfluoroalkyl substances (PFAS) has changed since the early 2000s, in part, because of the phase-out and replacement of some long-chain PFAS. Studies of PFAS exposure and its temporal changes have been limited to date mostly to adults and pregnant women. We examined temporal trends and determinants of PFAS serum concentrations among mothers with a young child who participated in the CHARGE ( CH ildhood A utism R isk from G enetics and E nvironment) case-control study. Methods We quantified nine PFAS in serum samples collected from 2009 to 2016 in 450 Northern California mothers when their child was 2–5 years old. Except for four compounds that were detected in less than 50% of the samples, we used multiple regression to estimate least square geometric means (LSGMs) of PFAS concentrations with adjustment for sampling year and other characteristics that may affect maternal concentrations (e.g., breastfeeding duration). We used time-related regression coefficients to calculate percent changes over the study period. Results LSGM concentrations of perfluorooctanoate (PFOA), perfluorooctane sulfonate (PFOS), and perfluorohexane sulfonate (PFHxS) decreased over the study period [percent change (95% confidence interval): 10.7% (−12.7%, −8.7%); −10.8% (−12.9%, −8.5%); −8.0% (−10.5%, −5.5%), respectively]. On the other hand, perfluorononanoate (PFNA) showed mixed time trends. Among the selected covariates, breastfeeding duration was inversely associated with the maternal serum concentrations of PFOA, PFOS, PFHxS, PFNA and perfluorodecanoate. Conclusions Our study demonstrated that body burden of some common long-chain PFAS among California mothers with a young child decreased over the study period and that breastfeeding appears to contribute to the elimination of PFAS in lactating mothers.