Abstract 11547: Physical Activity Modifies the Association Between Sleep Duration and Incident Cardiometabolic Risk: Findings from the All of Us Research Program
Abstract
Background: Sleep duration and physical activity are important behavioral determinants of cardiometabolic risk, yet their joint effects on incident hypertension, obesity, and type 2 diabetes remain incompletely understood. We evaluated whether objectively measured physical activity modifies associations between sleep duration and cardiometabolic outcomes and explored whether genetic predisposition to sleep disorder modifies the association between physical activity and cardiometabolic risk. Methods: Participants from the All of Us Research Program with longitudinal wearable sleep and activity data (Fitbit) were analyzed. Monthly retained mean sleep duration and physical activity metrics were derived from these participants. Outcome-specific Cox proportional hazards models assessed incident hypertension, obesity, and type 2 diabetes among participants free of disease at baseline. Models were adjusted for age, sex, race, social deprivation index, smoking status, antihypertensive medication use, antidiabetic medication use, and outcome-specific covariates. Restricted cubic splines and interaction analyses evaluated the modification of sleep-outcome associations by physical activity. Exploratory two-stage residual inclusion Mendelian randomization analyses leveraged sleep disorder polygenic risk scores as instrumental variables for objectively measured sleep duration to evaluate potential causal modification of physical activity associations with outcome incidence. Results: Among 20,461 participants (mean age 45.8 ± 15.7 years; 68% female), monthly retained mean sleep duration was 6.8 ± 0.8 hours/night and daily step count was 7,228 ± 3,302 steps/day. Median follow-up was 8.6 years (IQR 4.7-13.4), during which 5,033 hypertension, 4,330 obesity, and 1,982 type 2 diabetes events occurred. Higher daily step count was associated with lower incident hypertension (HR per 1,000 steps/day 0.98, 95% CI 0.97-0.98), obesity (HR 0.89, 95% CI 0.89-0.90), and type 2 diabetes (HR 0.95, 95% CI 0.93-0.96). Physical activity significantly modified the association between sleep duration and incident obesity (interaction p=4.03х10 − 8 ) and type 2 diabetes (p=0.0059), but not hypertension (p=0.60). Exploratory Mendelian randomization analyses demonstrated evidence consistent with a potential causal sleep-physical activity interaction for obesity (p=0.04). Conclusions: Higher physical activity was associated with lower risks of hypertension, obesity, and type 2 diabetes and significantly modified the relationship between sleep duration and incident obesity and type 2 diabetes. Exploratory Mendelian randomization findings support a potential causal sleep-physical activity interaction for obesity, highlighting the importance of jointly targeting sleep and physical activity in cardiometabolic disease prevention.