Prenatal exposure to per- and polyfluoroalkyl substances and child adiposity at age 5 years: a multipollutant analysis

Abstract

BACKGROUND AND AIM: Prenatal exposure to per- and polyfluoroalkyl substances (PFAS) is linked to lower birthweight and adiposity at birth. In some but not all previous studies, prenatal PFAS predict greater childhood BMI. However, few have directly measured child adiposity. We hypothesized that prenatal PFAS would be positively associated with child adiposity, evaluated using a precise method. METHODS: Pregnant women enrolled in the Healthy Start study (2009-2014). Blood samples were collected at a median of 27 weeks of gestation. Eleven PFAS were quantified in maternal serum; five were detected in 75% of participants and analyzed here. Child body composition at age 5 years was assessed using air displacement plethysmography. Adiposity was calculated as fat mass/total body mass x 100%. Three statistical approaches estimated associations between prenatal PFAS and child adiposity: single pollutant linear regression models; multipollutant linear regression model; and Bayesian Kernel Machine Regression (BKMR). Multipollutant models included five PFAS and a common covariate set. RESULTS:Among 340 participants, median child adiposity was 20.2% [interquartile range: 8.6%]. Pairwise Spearman correlations between PFAS were moderate to high (range: 0.32-0.78). In single pollutant models, each ln-unit increase in perfluorooctanoate (PFOA) was associated with 1.3% (95% CI 0.2, 2.5) greater child adiposity; no other PFAS was associated with adiposity. In the multipollutant model, PFOA was more strongly associated with adiposity than in the single pollutant model and there was an inverse association between perfluorodecanoate (PFDA) and adiposity; all variance inflation factors were 5. BKMR confirmed PFOA as the only PFAS with a posterior inclusion probability 0.3, but also showed a weak inverse association with PFDA. No pairwise interactions between exposures or strong non-linear trends were detected. CONCLUSIONS:Prenatal exposure to PFOA was positively associated with child adiposity at age 5. Multipollutant linear regression and BKMR produced similar results in this relatively simple scenario. KEYWORDS: PFAS, Obesity and metabolic disorders, Mixtures analysis, Children’s Environmental Health