Isotemporal Substitution Modeling on Sedentary Behaviors and Physical Activity With Mortality Among People With Different Diabetes Statuses: A Prospective Cohort Study From the National Health and Nutrition Examination Survey Analysis 2007–2018

Abstract

Background: To assess the associations of replacing sedentary behavior with different types of physical activity with mortality among the US adults of varying diabetes statuses. Methods: This prospective cohort study included 21,637 participants (mean age, 48.5 y) from the National Health and Nutrition Examination Survey 2007–2018. Physical activity including leisure-time moderate-vigorous-intensity activity (MVPA), walking/bicycling, worktime MVPA, and sedentary behavior. We conducted an isotemporal substitution analysis using Cox regression to estimate the associations between replacements and mortality risks. Results: We found significant protective associations between replacing 30 minutes per day sedentary behavior with 3 types of physical activity and all-cause, cardiovascular disease (CVD) mortality risk (except worktime MVPA for CVD mortality) among total participants, with hazard ratio (HR; 95% confidence interval [CI]) ranging from 0.86 (0.77–0.95) to 0.96 (0.94–0.98). Among participants with diagnosed diabetes, replacing sedentary behavior with leisure-time MVPA was associated with a lower all-cause mortality risk (HR 0.81, 95% CI, 0.70–0.94), which was also observed in other subgroups, with HRs (95% CI) ranging from 0.87 (0.80–0.94) to 0.89 (0.81–0.99). Among those with prediabetes/undiagnosed diabetes, replacing sedentary behavior with walking/bicycling was associated with lower CVD mortality risk, and replacement to work-time MVPA was associated with lower all-cause and CVD mortality risk, with HRs (95% CI) ranging from 0.72 (0.63–0.83) to 0.96 (0.92–0.99). Conclusions: Replacing sedentary behaviors with 30 minutes per day leisure-time MVPA was associated with lower all-cause mortality, regardless of diabetes statuses. Among people with prediabetes/undiagnosed diabetes, walking/bicycling was additionally associated with lower CVD mortality, and worktime MVPA was associated with lower all-cause and CVD mortality.