Abstract Purpose of Review Spatial data and spatially enabled tools are now embedded in routine epidemiologic practice, yet their widespread availability has outpaced explicit attention to the conceptual reasoning that should guide their use. This review asks how epidemiologists can more deliberately bring spatial thinking — reasoning about place, scale, and spatial relationships — to bear before choosing analytic methods, and what conceptual frameworks best support that process. Recent Findings The geographic and spatial statistical literatures offer three foundational components of spatial thinking: concepts of space and place, representations of space and place, and processes of spatial reasoning. Epidemiologic conceptualizations of place have evolved across four frameworks — place as container, cause, modifier, and dynamic system — each of which foregrounds different questions and methods. Methodological work highlights persistent challenges including the modifiable areal unit problem, scale-data misalignment, and the ways that analytic choices can quietly redefine the estimand. Summary Spatial thinking should precede and structure spatial analysis rather than follow it. Organizing that thinking around three recurring epidemiologic questions — whether meaningful spatial heterogeneity exists, whether spatial dependence is substantively informative, and how places are relationally connected — clarifies the inferential target and improves analytic coherence. Spatial approaches yield their greatest epidemiologic value when grounded in explicit conceptual framing and when maps and models are understood as representations of ongoing population processes rather than causal endpoints.
Kramer et al. (Sat,) studied this question.