Title : Objectively measured physical activity and risk of incident psoriasis: A retrospective cohort study using Fitbit data
Abstract:
Background: Psoriasis is a chronic inflammatory skin disease with systemic immune and metabolic implications that impact skin health and overall wellness. Lower physical activity has been linked to worse psoriasis severity and cardiometabolic outcomes, however, limited data exist examining whether objectively measured physical activity patterns precede psoriasis onset.
Methods: We conducted a retrospective cohort study using Fitbit data linked to electronic health records (EHRs) from the National Institutes of Health All of Us Research Program. Fitbit devices are consumer-grade, wrist-worn accelerometer-based activity trackers that continuously measure step counts and classify movement intensity into sedentary, light, fairly active, and very active categories based on proprietary algorithms. The analytic cohort included 14,029 participants with sufficient Fitbit data, of whom 186 developed psoriasis. Physical activity was assessed over the two-year period preceding psoriasis diagnosis (cases) or the most recent EHR date (controls). Inclusion required ≥50 valid wear days (≥10 hours/day with 100–45,000 steps), ≥180 days of device span, and ≥365 days of EHR data. Physical activity was operationalized as the proportion of daily wear time spent sedentary, lightly active, fairly active, and very active. Descriptive analyses were performed, and group differences were assessed using Fisher’s exact tests, Pearson’s chi-squared tests, and Wilcoxon rank sum tests, as appropriate.
Results: Among the 14,029 participants (mean age 49.2 ± 15.4 years; 72.1% female), participants who developed psoriasis demonstrated a higher mean proportion of daily wear time spent sedentary compared with those without psoriasis (76.11% vs 74.82%, p=0.017). In contrast, the proportion of wear time spent lightly active (20.15% vs 21.43%, p=0.006) and very active (1.67% vs 1.88%, p=0.018) was lower among those later diagnosed with psoriasis. Fairly active time did not significantly differ between groups (2.06% vs 1.87%, p=0.524). Participants who developed psoriasis also had a higher mean BMI (30.68 vs 29.63, p=0.023). Mean daily step count was lower among participants who developed psoriasis (7.74 vs 7.93 thousand steps), though this difference was not statistically significant (p=0.444).
Conclusions: In this large EHR-linked wearable cohort, individuals who developed psoriasis exhibited a more sedentary activity profile in the two years preceding diagnosis, characterized by greater sedentary time and reduced very active time. These findings suggest that lower-intensity activity patterns may precede psoriasis onset and highlight the potential role of consumer digital health technologies in identifying modifiable behavioral risk patterns in inflammatory skin disease. Further studies are needed to clarify causality and underlying mechanisms linking activity patterns and psoriasis risk and to adjust for possible confounding variables.