Linking "Big" Geospatial and Health Data: Implications for Research in Environmental Epidemiology

Environmental epidemiology studies increasingly integrate "big" geospatial and health data sets to examine associations between environmental factors and health outcomes. Using such data sets ─ and linking between them ─ presents a number of complexities with regard to study design and analytic approaches. These complexities are often magnified with the integration of additional contextual data representing other neighborhood characteristics including socioeconomic factors. Guidance regarding the design of environmental health studies that leverage "big" geospatial and health outcome data is limited and fragmented.

Drawing on methodological literature and case studies, this commentary outlines common challenges related to geospatial and health data linkages, posing a series of guiding questions and considerations for investigators conducting environmental health studies, particularly analyses with an etiological focus.

Recommendations include: 1) using a target trial approach to guide causal analysis, 2) aligning measures with hypothesized causal mechanisms, 3) exploring opportunities to "groundtruth" and validate data, and 4) prioritizing interdisciplinary science. The goal of the commentary is to consolidate insights from multiple disciplines ─ including exposure science, epidemiology, and sociology ─ to provide a foundation for etiologic research focused on advancing environmental health for all populations.