Location-weighted traffic-related air pollution and asthma symptoms in urban adolescents

Epidemiologic studies that implement area-based ambient air pollution measures to understand population-level health outcomes primarily use home address to assign exposure values to participants. However, as children spend a considerable amount of time at school, ambient air pollution levels at school locations may also be associated with health outcomes, particularly for pediatric asthma. We evaluated the impact of assigning a location-weighted average of nitrogen dioxide (NO2), a marker of traffic-related air pollution, on self-reported asthma symptoms in an urban cohort of middle school children based on combined school and home locations compared to home and school location alone. Middle school students in Oakland, CA (n = 4017), were administered a survey based on a modified International Study of Asthma and Allergy in Childhood (ISAAC) questionnaire. Annual average NO2 values were assigned to students' home address, school address, and a time-weighted average of the two locations from an annual land use regression model for the county. We implemented logistic regression to compare quartiles of home, school, and location-weighted average NO2 exposure for the eleven symptoms collected in the survey. Based on survey responses, for children classified as "current asthma," significant associations between NO2 exposure and asthma symptoms were found at home locations only. Students in the 2nd quartile of exposure were more likely to report wheezing and trouble sleeping compared to students in the 1st quartile of exposure at home (adjusted odds ratio [OR] = 1.152, 95% CI: 1.042, 2.310, and OR = 1.531, 95% CI: 1.020, 2.296, respectively). School-based exposures maintained lower, non-significant OR with wheezing and trouble sleeping for students in the 2nd quartile of NO2 (OR = 1.125, 95% CI: 0.748, 1.694 and OR = 1.262, 95% CI: 0.832, 1.916, respectively), while location-weighted average OR for both symptoms were attenuated compared to both locations and not statistically significant. Studies that assign area-based pollution measures at home only may not account for total pollution exposures, especially for children who spend a considerable portion of time in school locations.