Temporal and bidirectional associations between physical activity and sleep in adolescents

The purpose of this study was to assess the temporal day-to-day associations between physical activity and sleep in adolescents during a complete week. A total of 236 secondary education students (127 males, 109 females), age range 12-15 years, enrolled in 10 different high schools agreed to participate. They were asked to wear a GT3x accelerometer and complete a diary for a full week (24 hours / sev-en days). Participants' Body Mass Index and socio-economic status were assessed. Multilevel models were used to test the associations of nighttime sleep (onset, off-set, duration and efficiency) and wakefulness movement variables (MVPA, sed-entary behavior). MVPA was a significant predictor of sleep duration (B = .004, p < .001) and sleep offset (B = .006, p <.001). Adolescents who were more seden-tary during the day fell asleep later (B = .005, p < .001) and woke up later (B = .002, p < .001). Sedentary behavior during the day was inversely related to sleep duration that night (B =-.003, p < .001). Sleep offset was inversely related to MVPA (B =-.064, p < .001) and sedentary behavior (B =-.120, p < .001). Sleep duration was inversely associated with MVPA (B =-.27, p < .05) and sedentary behavior (B =-.41 p < .001) the following day. It seems physical activity could improve adolescents' sleep, while sedentary time could have a negative impact on it. The predictive role of sleep in physical activity is even less conclusive, indicating that the physiological relationship could be influenced by other factors such as time conflicts between individuals' daily activities. More research in this field is needed.