LAUSR

Background: Children are born with a burden of persistent organic pollutants (POPs) which may have endocrine disrupting properties and have been postulated to contribute to the rise in childhood obesity. The current evidence is equivocal, which may be because many studies investigate the effects at one time point during childhood. We assessed associations between prenatal exposure to POPs and growth during infancy and childhood. Methods: We used data from two Belgian cohorts with cord blood measurements of five organochlorines [(dichlorodiphenyldichloroethylene (p,p'-DDE), hexachlorobenzene (HCB), polychlorinated biphenyls (PCB-138, -150, -180)] (N = 1,418) and two perfluoroalkyl substances [perfluorooctanoic acid (PFOA) and perfluorooctanesulfonic acid (PFOS)] (N = 346). We assessed infant growth, defined as body mass index (BMI) z-score change between birth and 2 years, and childhood growth, characterized as BMI trajectory from birth to 8 years. To evaluate associations between POP exposures and infant growth, we applied a multi-pollutant approach, using penalized elastic net regression with stability selection, controlling for covariates. To evaluate associations with childhood growth, we used single-pollutant linear mixed models with random effects for child individuals, parametrized using a natural cubic spline formulation. Results: Prenatal exposures to p,p'-DDE and PCB-153 were selected in elastic net models for infant growth analysis, but the selections were unstable. No clear association between any of the exposures and longer-term childhood growth trajectories was observed. We did not find evidence of effect modification by child sex. Conclusion: Our results suggest that prenatal exposure to PCB-153 and p,p'-DDE may affect infant growth in the first two years, with little evidence of more persistent effects. Keywords: organochlorines, poly- and perfluoroalkyl substances, infant growth, childhood growth trajectory, longitudinal study