Ten Years of Exposure: Investigating Links Between Air Pollution and Child Asthma Across the Contiguous U.S. Using Spatiotemporal Interpolation

Document Type

Conference Proceeding

Publication Date

11-6-2017

Publication Title

Proceedings of the American Public Health Association Annual Meeting and Exposition

Abstract

Asthma is the most common chronic disease among youth, affecting 8.6% of children in the U.S. Childhood asthma is associated with air pollutants such as fine particulate matter (PM2.5) and ozone (O3). Environmental exposure assessment is a critical analytical tool to spatially and temporally link varying exposures and health outcomes. Objectives: This study 1) applied advanced spatiotemporal interpolation methods (cross validation, support vector machine, random forest, bagging, inverse distance weighting) to national-level data to estimate local neighborhood air pollution exposure at the census block group level and 2) evaluated associations between neighborhood PM2.5 and O3 pollution and childhood asthma outcomes across the contiguous U.S. over the last decade. Methods: Using air pollution data from the Environmental Protection Agency (EPA) Air Quality System, we spatiotemporally interpolated monthly PM2.5 and O3 concentration values at U.S. census block group centroids from 2006 to 2015. Mean and max PM2.5 and O3 monthly values were aggregated to create lifetime air quality exposure variables. Child asthma data were obtained from National Health Interview Survey (NHIS) and included the question: “Has a doctor or other health professional EVER told you that (child) had asthma?” Geographic variables including state, county, census tract, and block group used to merge NHIS data with EPA data were restricted variables and were accessed through the CDC Restricted Data Center. Regression models estimated the influence of PM2.5 and O3 on childhood asthma controlling for sex, age, race, poverty ratio, adult smoking status, household type, and urbanicity. Results: A total of 98,223 NHIS respondents indicated they had a child living at home. Approximately 13.9% (n=13,649) indicated that their child had ever had asthma. The majority of children were white (75.2%), male (51.1%), and had a mean age of 8.5. Children mostly lived in a house, apartment, flat, or condo (94.9%), had a family member with health insurance (42.8%), and lived with an adult that never smoked (52.0%). Results indicated that children exposed to higher mean and maximum O3 levels were significantly more likely to ever have asthma (OR=7340.9, CI: 23.2, 2323447; OR= 36.8, 95% CI: 2.5, 548.2). Max PM2.5 levels were not significance. However, contrary to previous research children exposed to higher mean PM2.5 were less likely of ever having asthma (OR= 0.9, CI: 0.9, 1.0). As expected, covariates: race, income-to-poverty ratio, smoking in the household, and living quarters showed significant relationships with children ever having asthma. Conclusion: This study provided evidence of spatiotemporal relationships and disparities relating to air pollution and childhood asthma at a national level over multiple years. PM2.5 results were not as expected and further adaptation and validation of big data modeling techniques for air pollution values across space and time is warranted. Identifying these relationships will aid healthcare decision-making capabilities and resource allocation, guide health promotion and education efforts, lay foundation for intervention research, and inform environmental and policy changes to improve air quality and improve child asthma outcomes.

Link to Full Text

Share

COinS