UCSF

Objectives

Animal studies suggest that air pollution is neurotoxic to a developing fetus, but evidence in humans is limited. We tested the hypothesis that higher air pollution is associated with lower child IQ and that effects vary by maternal and child characteristics, including prenatal nutrition.

Methods

We used prospective data collected from the Conditions Affecting Neurocognitive Development and Learning in Early Childhood study. Outdoor pollutant exposure during pregnancy was predicted at geocoded home addresses using a validated national universal kriging model that combines ground-based monitoring data with an extensive database of land-use covariates. Distance to nearest major roadway was also used as a proxy for traffic-related pollution. Our primary outcome was full-scale IQ measured at age 4-6. In regression models, we adjusted for multiple determinants of child neurodevelopment and assessed interactions between air pollutants and child sex, race, socioeconomic status, reported nutrition, and maternal plasma folate in second trimester.

Results

In our analytic sample (N = 1005) full-scale IQ averaged 2.5 points (95% CI: 0.1, 4.8) lower per 5 μg/m³ higher prenatal PM₁₀, while no associations with nitrogen dioxide or road proximity were observed. Associations between PM₁₀ and IQ were modified by maternal plasma folate (p_interaction = 0.07). In the lowest folate quartile, IQ decreased 6.8 points (95% CI: 1.4, 12.3) per 5-unit increase in PM₁₀; no associations were observed in higher quartiles.

Conclusions

Our findings strengthen evidence that air pollution impairs fetal neurodevelopment and suggest a potentially important role of maternal folate in modifying these effects.