Satellite imagery encodes features predictive of regional mortality and life expectancy

Background Increasingly accessible satellite imagery provides scalable measures of the built and natural environment relevant to population health. However, whether such imagery can capture subnational variation in mortality and life expectancy remains unclear. We therefore assessed its predictive value for regional mortality and life expectancy across OECD regions. Methods We conducted an ecological, cross-sectional prediction study using 2023 data from OECD Territorial Level 3 (TL3) regions. Annual cloud-masked composites from the Harmonized Landsat and Sentinel-2 collection were processed in the Google Earth Engine, tiled at 224 x 224 pixels, and encoded with the pretrained Prithvi foundation model to derive region-level satellite embeddings. For each outcome, we trained LightGBM regressors for a country-only baseline, a satellite-only model, a combined model (country + satellite), and a final contextual model that additionally included prespecified socioeconomic and environmental covariates. Performance was evaluated using 10-fold outer cross-validation with held-out test folds; R2 was the primary metric. Results The analytic sample comprised 2,414 OECD TL3 regions across 38 countries, for which 939,959 satellite image tiles were processed. In paired bootstrap comparisons, adding satellite features to country indicators improved predictive performance for all outcomes, with incremental R2 ranging from 0.097 to 0.233. The final contextual model achieved R2 values of 0.78 (95% CI, 0.74-0.81) for crude mortality, 0.87 (0.84-0.89) for age-adjusted mortality, 0.86 (0.82-0.88) for infant mortality, and 0.76 (0.69-0.84) for life expectancy. In SHAP analyses, the aggregated satellite image effect consistently ranked among the top predictors across outcomes. Conclusion Satellite imagery captures subnational environmental heterogeneity relevant to regional mortality and life expectancy beyond country identity alone. Earth observation may therefore provide a scalable, complementary data source for characterizing geographic disparities in population health.