Investigation of the Short-term Effects of Heatwaves and Air Pollution on Birth Outcomes and Child Health in the U.S.
AdvisorStrickland, Matthew J.
AltmetricsView Usage Statistics
Air pollution and climate change have been global public health issues for decades. However, methodological challenges and the quality of exposure measurement have limited the ability to estimate associations between temperature extremes and source-specific air pollution and child health and birth outcomes. This dissertation aims to investigate the associations between heatwaves and preterm and early-term births in the U.S.; to evaluate seasonality adjustment methods of three commonly used study designs (time-stratified case-crossover, time-series, and case-control) in studies of temperature extremes and birth outcomes; and use data from source apportionment modeling to estimate associations of source-specific outdoor PM2.5 concentrations with pediatric respiratory outcomes. We performed a matched (on location, race/ethnicity, and maternal education) case-control study to examine the acute associations between heatwaves and early-term and preterm births; a simulation study to evaluate bias in seasonality adjustment methods; and a time-stratified case-crossover analysis to estimate acute associations between source-specific PM2.5 and pediatric respiratory emergency department (ED) visits. In the first study, we observed positive associations between heatwaves and early-term birth. Exposure to more high-heat days over the previous week yielded higher odds ratios (OR). ORs between heatwaves and preterm and early-term birth were higher among black and Hispanic mothers than among white mothers. In the presence of seasonality of conception, the simulation study (aim 2) found a slight positive bias in the coefficient for mean temperature in the warm season for the case-crossover approach. For the time-series design, a “pregnancy-at-risk” approach completely adjusted for the seasonal bias. No bias was observed in the case-control design with or without adjusting for seasonality. The variance of the coefficients increased in the adjusted models and was highest in the case-control design. In the third dissertation aim, we observed metals, natural gas, and all other sources (a combined category which includes sources that the source apportionment method cannot categorize) of PM2.5 were positively associated with pediatric emergency department visits for asthma, pneumonia, and acute upper respiratory infection. Heterogeneity was observed in ORs across sources and outcomes. Collectively, this dissertation’s findings have implications for intervention programs to mitigate heatwave impacts on birth outcomes for vulnerable populations. The simulation results can inform method choice for seasonality adjustment when studying temporally varying exposures and birth outcomes. The aim 3 findings have implications for identifying the relative importance of PM2.5 sources in triggering acute pediatric respiratory events.