PhD defence: Rethinking environmental exposure assessments through human daily mobility

Most epidemiological studies attempt to assess how the environment impacts health by estimating exposures based solely on individuals' home locations. This approach is limited because people are mobile and spend significant time outside their homes, which can lead to a potential misclassification of environmental exposures. This project aimed to address this gap by advancing mobility-based exposure assessments, focusing on making them more accurate, valid, and efficient. A crucial finding is that while incorporating mobility data offers theoretical advantages over home-based approaches, mobility data alone does not substantially improve the accuracy of exposure assessments. Instead, significant validity gains are achieved only when relevant contextual information is integrated, such as the exposure differences between indoor and outdoor environments. Furthermore, we demonstrated that highly efficient assessment strategies are practical: short-term (7 to 14 days), low-frequency GPS data collected via smartphones is sufficient for capturing long-term exposure. Since the majority (over 80%) of passive exposure occurs at key activity locations like home and work, this thesis supports simplifying mobility-based assessments into activity location-based approaches, using key activity locations where people spend most of their time (e.g., home and workplace), thereby providing a scalable and efficient framework for population-level exposure assessment.