Residential early-life air pollution exposure and oculomotor markers of cognitive workload in children: a prospective birth cohort study

Background Growing evidence indicates that early brain development is particularly vulnerable to environmental influences, with early-life air pollution exposure linked to attention problems, anxiety, and depression. In this study, we examined associations of early-life air pollution exposure with eye-tracking-derived measures of cognitive workload and emotion-related attentional bias in school-aged children, while also assessing whether recent exposure was associated with these outcomes. Methods This prospective cohort study was conducted within the ongoing ENVIR ON AGE birth cohort, which has enrolled over 2400 mother–child pairs. The present study included 196 children aged 9–11 years who had reached the eligible age for follow-up and completed eye-tracking paradigms assessing cognitive workload and subconscious attention bias between September 2021 and November 2023. Residential black carbon (BC), nitrogen dioxide (NO 2 ), and fine particulate matter (PM 2.5 ) exposures were estimated using a high-resolution spatiotemporal modelling framework based on fixed-site monitoring data, land-use/land-cover information, and dispersion modelling. Associations between predefined exposure windows and eye-tracking outcomes were first assessed using multiple linear regression models, followed by Bayesian kernel machine regression distributed lag models (BKMR-DLM) to evaluate multi-pollutant and multi-window exposure-response patterns. Findings Higher recent residential air pollution exposure was most consistently associated with saccadic outcomes. After correction for multiple testing, an interquartile range (IQR) increase in last-month BC and NO 2 exposure was associated with lower saccade velocity (BC: −10.97°/s; 95% CI −18.67, −3.26; NO 2 : −10.81°/s; 95% CI −18.13, −3.49), and last-month NO 2 and PM 2.5 exposure were associated with fewer saccades. Early-life exposure windows also showed associations mainly with saccade velocity, including first-trimester BC and PM 2.5 , whole-pregnancy PM 2.5 , and first-1000-days BC exposure. Associations with pupil diameter, fixation count, urinary BC, and emotion-related attention-bias outcomes did not remain statistically significant after correction for multiple testing. Interpretation Using objective eye-tracking outcomes within a prospective birth cohort, our findings suggest that residential air pollution exposure is associated with subtle oculomotor differences in school-aged children, particularly patterns consistent with increased cognitive workload. Emotion-related attention-bias findings were exploratory after correction for multiple testing. Funding Special Research Fund (BOF), Flemish Scientific Research Fund (FWO), Methusalem, and Horizon Europe project MISTRAL.