Epidemiology and geospatial analysis of built environment determinants of healthy and resilient cities

Urbanisation and climate change are expected to introduce novel public health, environmental, and sustainability challenges. Epidemiological studies indicate that climate and public health vulnerabilities vary by neighbourhood. However, data at these spatial levels are largely unavailable despite studies demonstrating that geographically aggregated data mask disparities. To address this gap, this dissertation developed and applied high resolution geospatial vulnerability and health indicators across British Columbia (BC), Canada. The first dataset applied principal components analysis to more than 30 measures to map exposures, population sensitivities, adaptive capacities, and overall vulnerabilities of four climate hazards (extreme heat, inland flooding and sea level rise, wildfire smoke, and ground-level ozone) across 4188 dissemination areas in two health regions. A principal components analysis revealed varied opportunities for adaptive capacities across all hazards (16%-47% contribution to variation in overall vulnerability), with the greatest contribution found for flooding (47%). Overall, sensitivity explained the most variance, suggesting strategies targeting age and those with pre-existing health conditions in public health and emergency responses. Building on this result, the second dataset linked mortality data and sociodemographic information in a Bayesian small area model to estimate life expectancy (LE) at birth and 20 causes of mortality over 27 years across 368 Census Tracts (CTs) in Metro Vancouver, BC. The dataset identified spatial LE gaps of more than 10 years that widened in recent years. Absolute inequalities decreased for all diseases except for neoplasms, but relative inequalities increased for all causes. In the final study, difference-in-differences models were applied to the small area mortality data to evaluate relationships with population density and sociodemographic measures to assess optimum density levels, and the effects of density changes over time. At densities above ~9,400 persons per km2, LE began to decrease more rapidly. By cause, densification was linked to decreased mortality for major causes of mortality in the region, such as cardiovascular diseases, neoplasms, and diabetes. Through these three studies, this dissertation provided evidence for the importance of local-level indicators of health, vulnerability, and built environment variables for future and ongoing surveillance of healthy and resilient cities.Medicine, Faculty ofPopulation and Public Health (SPPH), School ofGraduat