Signal Study finds consumer wearable heart-rate data tracks state-level economic hardship in the US

Summary

A study examines whether passively sensed, consumer-wearable physiological data can recover a socioeconomic gradient previously established mainly through clinical and survey research. Using 19.1 million quality-filtered photoplethysmography (PPG) readings from 18,734 opt-in users of the Welltory app, the researchers computed a cohort-adjusted mean daytime resting heart rate for each US state and related it to a four-component state-level material-hardship index covering uninsurance, food insecurity, utility shutoffs, and housing insecurity, using data from 41 states and 12,497 contributing users. After adjusting for six state health and behavior indicators, latitude, median age, and population density, state-level resting heart rate tracked hardship at a partial Spearman correlation of +0.74 (bootstrap 95% confidence interval of +0.31 to +0.87). The positive gradient held under every robustness check performed, including leave-one-state-out analysis, demographic re-weighting, census-region fixed effects, split-sample resampling, and a heavy-wear-only subsample. Users in the five highest-hardship states averaged 1.33 beats per minute higher resting heart rate than users in the five lowest-hardship states, and the relationship was not matched by income or inequality measures under the same statistical adjustment.