Air pollution exposure during pregnancy and brain morphology in young children: a population-based prospective birth cohort study

Abstract

Background: Increasing evidence indicates that air pollution exposure during pregnancy is related to impaired child neuropsychological development. However, it is unclear if brain structural alterations underlie this association. Therefore, we aimed to assess whether air pollution exposure during pregnancy is related to brain morphology in young children. Methods: Population-based birth cohort set up in Rotterdam, The Netherlands (2002-2006). Air pollution levels at birth addresses were estimated by land-use or dispersion models based on monitoring campaigns done between 2008 and 2010. Levels were extrapolated back in time to the pregnancy period using routine background monitoring network. Structural neuroimaging, diffusion tensor imaging, and neuropsychological assessment using NEPSY-II were performed at age 8 years (n=783). Results: Air pollution levels were not associated with global brain volumes or white matter microstructure. Conversely, children exposed to PM2.5, PMcoarse, PM2.5absorbance, and benzo(a)pyrene (BaP) during pregnancy had thinner cortices in several brain regions particularly in the frontal lobe, after adjusting for child’s sex, age, IQ, and parental socioeconomic and life-style characteristics. Specifically, PM2.5 and BaP levels were associated with a reduced cortical thickness of the rostral middle frontal and praecuneus regions, and PM2.5 levels during pregnancy were related to the pars orbitalis region. PM2.5 and BaP levels were associated with an impaired worse inhibitory control of the children, while a reduced cortical thickness of the regions related to these pollutants was also associated with an impaired inhibitory control. Conclusions: Our results suggested that air pollution exposure during pregnancy is associated with a reduced cortical thickness in several brain regions, particularly in the frontal lobe. Both the exposure and the regions with reduced cortical thickness were related to an impaired inhibitory control in children.