Association of Domain-Specific Physical Activity with Sleep Disorders in U.S. Adults: A Cross-Sectional Study Based on NHANES

Background: The relationships of different physical activity domains [e.g., Occupation-related Physical Activity (OPA), Transportation-related Physical Activity (TPA), and Leisure-time Physical Activity (LTPA)] with sleep disorders (SD) remain poorly elucidated. This study investigated the associations of OPA, TPA, and LTPA with SD in American adults. Methods: Data were derived from 25,732 adults with a mean (standard deviation) age of 49.72 (17.53) years (12,692 male and 13,040 female) participating in the 2007–2018 cycles of the National Health and Nutrition Examination Survey (NHANES). The PA domains were assessed using the Global Physical Activity Questionnaire (GPAQ), and adherence to PA recommendations was determined according to the 2018 PA guidelines. SD were identified by self-reported or doctor-diagnosed questionnaire items. Survey-weighted logistic regression models were applied to evaluate the associations between PA domains and SD, and restricted cubic spline (RCS) analyses were conducted to examine potential nonlinear associations. Results: In the fully adjusted model, total PA was not significantly associated with SD (OR 1.01, 95% CI 0.89–1.13, p = 0.912). Regarding specific PA domains, TPA was associated with lower odds of SD (OR 0.79, 95% CI 0.67–0.93, p = 0.006), whereas OPA was associated with higher odds of SD (OR 1.15, 95% CI 1.02–1.30, p = 0.023). No significant association was observed between LTPA and SD (OR 0.96, 95% CI 0.85–1.09, p = 0.511). The multivariable-adjusted RCS analysis further indicated significant nonlinear associations between PA domains and SD (all p for non-linear <0.001). Conclusion: Our findings underscore a prominent “PA paradox” in sleep health among US adults, suggesting that active commuting (TPA) exerts a protective role against SD, whereas OPA serves as a potential behavioral risk factor. Future longitudinal studies and objective PA measurements are needed to validate these cross-sectional associations.