0326 Daily Sleep and Affect in Adolescents: Differential Associations During Constrained Versus Unconstrained Sleep

Abstract

Abstract Introduction Few studies have examined bi-directional associations between daily sleep and affect in adolescents, and even fewer assessed both high and low arousal affect under naturalistically-occurring constrained (school) and unconstrained (vacation) sleep opportunities. Methods 205 adolescents (54.1% females, age M±SD=16.9±0.87 years) completed daily measures of sleep and affect over 28 continuous days (2-week school and 2-week vacation). Total sleep time (TST) and sleep efficiency (SE) were measured using actigraphy and sleep diary. High- and low-arousal positive and negative affect (PA, NA) were self-reported each afternoon. Cross-lagged, multilevel models were conducted: affect predicted same-night sleep controlling for previous-night sleep; sleep predicted next-day affect controlling for previous-day affect. Day of week, study day, and sociodemographics were controlled. Results During both school and vacation, adolescents with overall higher low-arousal PA also had greater self-reported SE. Other aspects of sleep-affect associations differ between school and vacation (all p<.05). During school, significant associations were between, not within individuals. Bi-directional associations were found between longer actigraphy-TST and greater high-arousal PA, and between higher self-reported SE and lower low-arousal NA. High-arousal PA and NA were associated with self-reported TST (positive and negative respectively). During vacation, between-individuals, higher NA was bi-directionally associated with lower self-reported SE, regardless of arousal. Longer self-reported TST was associated with lower low-arousal NA. On the within-person level, regardless of sleep measurement, nights with longer-than-average TST were associated with lower NA the next day (high- and low- arousal). Nights with higher-than-average SE predicted lower next-day low-arousal NA. Conclusion Sleep-affect associations differed based on sleep opportunity and arousal, suggesting potentially different mechanisms of action. When sleep is typically constrained, overall levels of sleep-affect associations were stronger than daily fluctuations. When sleep is typically unconstrained, significant associations were found both between- and within- persons. In particular, daily fluctuations in sleep were predictive of next-day NA, rather than the other direction. Support Monash International Postgraduate Research Scholarship and Monash Graduate Scholarship