Abstract:
Fine particulate matter (PM2.5) air pollution is associated with excess morbidity and mortality and estimated to be the most important environmental risk factor for mortality globally. Ultrafine particles (UFPs) and PM2.5 from wildfire smoke (WFS), are both thought to be more toxic than ambient PM2.5. The goal of this dissertation is to deepen our understanding of these understudied pollutants and to estimate health effects of their exposures in distinct applications .In the first aim, I developed an approach to use ultrafine particle mobile measurements from traffic-related air pollution mobile monitoring campaigns for epidemiology. I developed an exposure surface that is more representative of residential locations by lessening the contribution of localized on-road UFP plumes. Compared to an exposure surface that uses unadjusted mobile data, the exposure surface with predictions derived from on-road plume adjusted data performs better, with an MSE-based R2 of 0.71 compared to 0.60. Further, I found that UFP predictions at residential locations that use mobile measurements versus stationary measurements are more spatially resolved and pick up more information at known hotspot locations. We recommend correcting mobile measurements for use in epidemiology. Future research is needed to understand the impact of using predictions from on-road plume adjusted measurements in epidemiologic inference. In the second aim, I examined the impact of UFP, among several other traffic-related air pollutants, on physical functioning in the Adult Changes in Thought study, a cohort of older adults in the Puget Sound region of Washington state. Physical functioning is an important measure of aging and is measured by three assessments that make up the modified Performance-Based Physical Function Assessment, giving a score from 0-12: gait speed, chair-stand time, and grip strength. We used exposure predictions derived from the mobile monitoring campaign discussed in Aim 1, and examined UFP (several size bins), nitrogen dioxide (NO2), black carbon (BC), and PM2.5. We used linear mixed effect models to evaluate the association at baseline and longitudinally. There was no evidence of a longitudinal effect of UFPs on physical performance. Exposure to higher levels of BC, UFP in the range of 56-178nm, and non-time varying PM2.5 resulted in faster decline in physical performance score compared to exposure at lower levels, equivalent to an additional 3-5 months of aging over 5 years. In the third aim, I examined the association between wildfire smoke and emergency department (ED) visits in Washington state during June-September, 2017-2020. I examined the impact of wildfire smoke presence versus absence on ED visits, and the impact of each 10 µg/m3 increase in PM2.5 from wildfire smoke. I used a time-stratified case crossover design with conditional logistic regression. I observed an increased odds of asthma ED visits immediately after and in five days following initial wildfire smoke exposure (lag 0 OR: 1.13; 95% CI: 1.10, 1.17; lag 1-5 ORs all 1.05 or greater with a lower CI of 1.02 or higher), and an increased odds of respiratory ED visits across all five days following exposure. I observed mixed results for cardiovascular ED visits, with evidence of increased odds several days following initial exposure. I also found increased odds across all visit categories for each 10 µg/m3 increase in wildfire smoke impacted PM2.5. Finally, I observed increased odds of respiratory visits among ages 19-64, and among ages 5-64 for respiratory visits. The results of this study provide evidence for an increase in respiratory ED visits immediately following exposure and in the several days following, and a lagged effect on cardiovascular ED visits. These increased risks are observed particularly among children and younger to middle-aged adults. This dissertation advances our understanding of exposure characterization and health impacts of two toxic components of fine particulate matter, ultrafine particles and wildfire smoke, and provides support for further investigation.