Exposure to perfluoroalkyl substances during pregnancy and child behaviour at 5 to 9years of age

Abstract

We examined associations between prenatal exposure to perfluorohexane sulfonic acid (PFHxS), perfluoroheptanoic acid (PFHpA), perfluorononanoic acid (PFNA), and perfluorodecanic acid (PFDA) - and child behaviour (SDQ-total) and hyperactivity (sub-scale) at 5-9years of age in birth cohorts from Greenland and Ukraine. Pregnancy serum samples (N=1023) were analysed for perfluoroalkyl substances (PFASs) and categorised into tertiles and also used as continuous exposure variables. Problem behaviour and hyperactivity were assessed, using the Strength and Difficulties Questionnaire (SDQ) and categorised as normal/borderline and abnormal. Associations were analysed using multiple logistic and linear regression. High compared to low prenatal PFHxS exposure was associated with 1.16 (95% confidence interval (CI): 0.08; 2.25) point higher SDQ-total (more problem behaviour) in Greenland and 0.80 (CI: 0.06; 1.54) point higher SDQ-total in the combined analyses, whereas no association was present in Ukraine alone. One natural log-unit increase in prenatal PFNA exposure was associated with 0.90 (CI: 0.10; 1.71) points higher SDQ-total in Greenland and 0.72 (CI: 0.13; 1.31) points higher in the combined analysis and no association in Ukraine. Prenatal PFAS exposure was unrelated to problem behaviour (abnormal SDQ-total). In the combined analysis, odds ratio (OR) (CI) for hyperactivity was 1.8 (1.0; 3.2) for one natural log-unit increase in prenatal PFNA and 1.7 (1.0; 3.1) for one natural log-unit increase in prenatal PFDA exposure. Findings are compatible with weak effects on child behaviour of prenatal exposure to some PFASs although spurious results are not entirely unlikely. The associations were strongest in Greenland.

Keywords: Behaviour; Child; Child development; Cohort study; Fluorocarbons; Prenatal exposure, delayed effects.