Early-life exposure to air pollution and greater use of academic support services in childhood: a population-based cohort study of urban children

Abstract

Background: There is a growing literature showing associations between prenatal and early-life exposure to air pollution and children's neurodevelopment. However, it is unclear if decrements in neurodevelopment observed in epidemiologic research translate into observable functional outcomes in the broader pediatric population. The objective of this study was to examine the association between early-life exposures to common urban air toxics and the use of academic support services, such as early intervention and special education within a population-based cohort of urban children.

Methods: Data for 201,559 children born between 1994 and 1998 in New York City were obtained through administrative data linkages between birth, early intervention and educational records. Use of academic support services was ascertained from birth through attendance in 3^rd grade. Census tract at birth was used to assign estimates of annual average ambient concentrations of benzene, toluene, ethylbenzene and xylenes (BTEX) using the 1996 National Air Toxics Assessment. Discrete-time hazard models were fit to the data and adjusted for confounders including maternal, childhood and neighborhood factors.

Results: Children with higher exposures to BTEX compounds were more likely to receive academic support services later in childhood. For example, the adjusted hazard ratio comparing children exposed to the highest decile of benzene to those with lower exposure was 1.09 (95% confidence interval 1.05, 1.13). Results were consistent across individual BTEX compounds, for exposure metrics which summarized exposure to all four BTEX pollutants and for multiple sensitivity analyses.

Conclusions: These findings suggest urban air pollution may affect children's neurodevelopment and educational trajectories. They also demonstrate the use of public health data systems to advance children's environmental health research.

Keywords: Benzene; Early intervention; School outcomes; Special education; Traffic-related air pollution.

Fig. 1

Proportion of children not using academic support services, stratified by high exposure to at least one BTEX pollutant. New York City Longitudinal Study of Early Development 1994–1998 birth cohorts. Predicted probability curves resulting from adjusted discrete hazard models of time to using academic support services. Covariates fixed at a married White, non-Latina mother, who was born in the US, completed high school, delivered the child at 25, had Medicaid at the time of delivery, who was living in a community with the mean neighborhood deprivation index and whose child had a childhood blood lead less than 3 micrograms/deciliter

Fig. 2

Proportion of children not using academic support services, stratified by high exposure to all BTEX pollutants. New York City Longitudinal Study of Early Development 1994–1998 birth cohorts. Predicted probability curves resulting from adjusted discrete hazard models of time to using academic support services. Covariates fixed at a married White, non-Latina mother who was born in the US, completed high school, delivered the child at 25, had Medicaid at the time of delivery, who was living in a community with the mean neighborhood deprivation index and whose child had a childhood blood lead less than 3 micrograms/deciliter