Associations of prenatal fine particulate matter mixtures with neurodevelopmental outcomes in early childhood: component- and source-specific insights

Abstract This study investigates independent and joint effects of fine particulate matter (PM2.5) components on early childhood neurodevelopment and explores emission sources of key toxic components. We included 165 mother-infant dyads from Southern California. Annual average concentrations of 15 PM2.5 components, including carbonaceous components, secondary inorganic salts, and trace elements, were estimated for the birth year. Neurodevelopment across cognitive, language, motor, social-emotional, and adaptive behavior domains was assessed at age 2 using Bayley-III Scales. Mixture effects and key contributors were evaluated using weighted quantile sum (WQS) and Bayesian kernel machine regression (BKMR). Source inference was conducted through inter-component clustering and spatial analysis. Linear regression showed PM2.5, sulfate (SO4 2−), nitrate (NO3 −), ammonium (NH4 +), copper (Cu), nickel (Ni), lead (Pb), and vanadium (V) were inversely, while calcium (Ca) and zinc (Zn) were positively, associated with adaptive behavior scores (p < 0.05). WQS showed negative associations between the mixture and adaptive behavior (p = 0.02–0.06), with Ni, Cu, V, and SO₄²⁻ as key contributors. BKMR showed similar trends. Ni, V, and SO4 2− likely originate from heavy oil combustion, and Cu from brake wear. Findings suggest that PM2.5 components, particularly from traffic and marine fuel combustion, may adversely affect adaptive behavior in early childhood.