The Twin Crises of Climate Change and Air Pollution: Characterizing the Acute Cardiovascular Effects of Temperature and Uncertainties of Fine Particulate Matter Concentrations

Climate change is already harming public health through warmer, more erratic weather and many downstream consequences. Research can support climate change adaptation by characterizing climate-related exposures, identifying vulnerable populations, and identifying effective interventions. Furthermore, the main source of greenhouse gas emissions, fossil fuel combustion, also produces air pollutants such as fine particulate matter (PM2.5) that directly harm human health. In this dissertation my colleagues and I have examined the effects of short-term temperature exposure on myocardial infarction and stroke to inform adaption (Chapters 1 to 3) and leveraged multiple exposure models to estimate annual PM2.5 concentrations and quantify uncertainty (Chapter 4). To examine the effects of short-term temperature exposure, we conducted case-crossover analyses using an administrative dataset of hospitalizations in New York State. For PM2.5 prediction and uncertainty characterization, we applied a flexible ensemble approach to leverage seven already-developed PM2.5 models. Overall, we found that warmer hourly temperatures and greater daily temperature variability increased the risk of ischemic events, and lower hourly temperatures increased the risk of hemorrhagic stroke. The ensemble model showed high predictive accuracy, demonstrating the strength of this approach, and we observed greatest uncertainty in the Pacific Northwest and southeast coast. The methods applied in this dissertation can be applied to other exposures and in different settings to further quantify the risks of climate impacts and improve air pollution assessment. Future research should examine the joint impacts of multiple weather factors, strategies to protect people in group housing from extreme weather, measurements and models to reduce uncertainty of air pollution exposures, and propagate exposure uncertainty into health models. However, the need for further research should not delay climate action today