Long-term Fitbit data from 23,517 All of Us participants shows physical activity attenuates polygenic depression risk effects on symptom severity

Researchers calculated polygenic risk scores from genomic data and paired them with months-to-years of Fitbit step and heart-rate records plus depression-related ICD codes and PHQ scores. Higher average daily activity correlated with lower depression severity overall; the slope of this relationship steepened among participants in the top PRS decile, indicating a statistically significant interaction after adjustment for age, sex, BMI and socioeconomic covariates. The design improves on prior short-term or self-reported activity studies by capturing habitual behavior across diverse U.S. adults.

Prior observational work, including a 2023 JAMA Psychiatry meta-analysis of 1.2 million individuals, reported similar gene-by-exercise interactions using retrospective questionnaires, yet suffered from recall bias and limited temporal resolution. The current study’s objective, longitudinal accelerometry reduces measurement error and supports a plausible behavioral mechanism whereby sustained activity may blunt downstream expression of genetic liability through anti-inflammatory and neuroplastic pathways. Funding from NIH All of Us was disclosed; no industry conflicts were noted.

Because the analysis remains observational, residual confounding by unmeasured psychiatric treatment or reverse causation cannot be excluded. Next steps require randomized trials that stratify participants by PRS and prescribe stepped activity doses while tracking both symptom trajectories and objective biomarkers. Such designs would clarify whether the observed interaction is causal and whether personalized activity thresholds yield clinically meaningful prevention gains.