Prenatal exposure to heavy metals poses risks to fetal brain development, yet the joint effects of these metals remain unclear, with inconsistent findings across statistical models. This study investigates the joint effect of prenatal exposure to cadmium (Cd), nickel (Ni), mercury (Hg), and lead (Pb) on infant neurodevelopment using various statistical approaches. The study included 400 mother-infant pairs. Heavy metal levels were measured in maternal urine samples at the 12th week of gestation, and infant neurodevelopment at 40 days was evaluated by the Bayley Scales of Infant and Toddler Development. Generalized Additive Models (GAM), Multivariable Linear Regression (MLR) with restricted cubic spline (RCS), Bayesian Kernel Machine Regression (BKMR), and Weighted Quantile Sum (WQS) regression were applied to explore the associations between heavy metal exposure and neurodevelopmental outcomes. GAM revealed a significant linear relationship for Cd with cognitive scale (p = 0.045) and expressive language (p = 0.043). MLR confirmed that Cd was negatively associated with both cognitive scale (β = -1.47, p = 0.044) and expressive language (β = -0.32, p = 0.019) and RCS presented a non-linear association between Pb and language scale (p = 0.001). BKMR suggested a negative but non-significant association with most outcomes. WQS indicated a significant adverse effect of metal mixture on expressive language (β = -0.26, 95% CI = -0.44, -0.07), identifying Cd and Ni as the primary contributors. Prenatal exposure to heavy metals have detrimental effects on infant neurodevelopment, especially on language development.
Keywords: Bayesian kernel machine regression; Heavy metal; Metal mixture; Multiple pollutants; Neurodevelopment; Weighted quantile sum.
Copyright © 2025. Published by Elsevier Ltd.