“Effect of prenatal exposure to fine airbone particles on child mental and psychomotor development: an analysis based on propensity score matching”

Abstract

Introduction: Maternal exposure to ultrafine airborne particulate matter during pregnancy has recently been identified as a potential risk factor for infant neuropsychological development. This study estimated the causal effect of prenatal exposure to high concentrations of particles less than 2.5 μg/m3 in diameter (PM2.5) on children’s psychomotor and mental scores during their second year of life, in a birth cohort from the province of Guipúzcoa (The Basque Country, Spain). Methods: We performed the analysis under a potential outcome approachon 438 mother-children pairs. We focused on the average PM2.5 level during the whole pregnancy and we considered a binary version of this exposure in respect to the median: PM2.5<17 μg/m3 (control), PM2.5≥17 μg/m3 (treatment). A logistic model on this binary exposure was specified to estimate the propensity score,including the available background variables as predictors. The average treatment effect on the exposed children (Average Treatment effect on the Treated, ATT) was estimated for both the outcomes, following a nearest-neighbor matching procedure based on the estimated propensity score. Due to the presence of missing exposures, separated analyses were performed on five complete data sets obtained applying a Multiple Imputation procedure and the five results were combined in a final ATT estimate. Results: The final ATT estimate was equal to -2.3 (90% CI: -7.1; 2.5) for mental score and -3.1 (90% CI: -7.9; 1.7) for psychomotor score, indicating that, when exposed to high level of PM2.5, children get lower test scores. The large confidence intervals partly reflect the variability due to the presence of missing data. Conclusions: Our results suggest the existence of a negative causal effect of exposure to high PM2.5 levels during pregnancy on the infant psychomotor and mental development.