ABSTRACT
OBJECTIVE: Exercise and sleep are two important health promoting behaviors, but large parts of the population do not exercise or are chronically sleep deprived. We investigated to what degree exercise and sleep compete with each other and with other waking activities for time. METHODS: Analyses are based on 2003-2016 American Time Use Survey data of a representative sample of N = 47,862 working Americans aged 18-65 years interviewed on weekdays. Models were adjusted for various sociodemographic variables and time spent working on the interview day. RESULTS: Long work hours and several sociodemographic variables were associated both with short sleep and low exercise rates. Respondents who exercised slept on average 15.5 min less (95% CI: -18.4 min; -12.7 min, P < .0001) than those who did not exercise. Sleep duration decreased with exercise duration in a dose-dependent manner. The strongest association between exercise and sleep duration was observed for exercise between 6 AM and 8 AM (ie, before work) and between 9 PM and 11 PM (ie, before bed). However, unless exercise was performed before work in the morning, exercise durations of up to 1 h were not associated with relevantly curtailed sleep. CONCLUSIONS: These results confirm that exercise and sleep compete with each other for time. However, exercise regimens that follow current guidelines are unlikely to curtail sleep substantially, especially since exercise has been shown to positively affect sleep structure and quality. Public health efforts should nevertheless emphasize the importance of both exercise and sleep, and of not sacrificing one activity for the other.