Multiple pesticides in mothers' hair samples and children's measurements at birth: Results from the French national birth cohort (ELFE).

BACKGROUND: A growing body of studies now suggests that the general population is continuously and ubiquitously exposed to numerous pesticides. However, studies investigating the possible role of environmental exposure to pesticides on fetal growth have focused on a limited set of substances, despite the hundreds of modern pesticides currently available. AIM: To explore the relation between maternal hair concentrations of 64 pesticides and metabolites and their newborns' measurements at birth, with data from the ELFE French nationwide birth cohort. METHODS: We measured 64 compounds (10-100% detection) in bundles of hair 9cm long collected at birth from 311 women who gave birth in France in 2011. We assessed their associations with birth weight, length, and head circumference, adjusted for potential confounders, and used elastic net regularization to simultaneously select the strongest predictors of measurements at birth. Selected variables were multiply imputed for missing values, and unpenalized estimators were assessed by standard linear regression. RESULTS: We observed statistically significant associations between maternal hair concentrations of seven pesticides or pesticide metabolites and birth measurements (weight: fipronil sulfone; length: TCPy, bitertanol, DEP, and isoproturon; head circumference: tebuconazole and prochloraz). Analyses restricted to boys identified 12 additional compounds: 8 independently associated with birth weight (3Me4NP, DCPMU, DMST, fipronil, mecoprop, propoxur, fenhexamid, and thiabendazole), 2 with birth length (dieldrin and beta-endosulfan), and 6 with head circumference (beta-endosulfan, beta-HCH, fenuron, DCPMU, propoxur, and thiabendazole). CONCLUSION: Our results suggest that prenatal exposure to 19 pesticides or metabolites from various chemical families may influence measurements at birth. As with any exploratory research findings, results should be interpreted cautiously, until they are replicated or verified by further epidemiological or mechanistic studies.