0245 When do Emotions Impact Sleep? A State of the Evidence

Abstract

Abstract Introduction While theories often highlight the bi-directional nature of sleep and emotional processes, the impact of emotions on sleep has been comparatively neglected. To appraise the state of evidence for the causal influence of emotions on sleep, a meta-analysis of the existing literature was conducted, alongside a diary study to estimate naturalistic effects. Methods First, a pre-registered meta-analysis using PRISMA guidelines evaluated the strength, form, and context of experimental effects of emotion inductions on sleep parameters (k=27). Quality of primary experiments was evaluated by independent raters, and theoretically-relevant features were extracted and examined as moderating factors of observed effects (i.e., sleep parameter, design, sleep context, types of emotion inductions and emotions). Random-effect models were used to aggregate effects for each parameter. Second, a complementary pre-registered diary study of young adults (N=89) tracked the links of their global emotions (reported separately in the evening and morning) with actiwatch-assessed sleep across two weeks (Nobs=1,188). Results First, across the meta-analyzed experiments, there was a significant impact of emotion inductions on delayed sleep onset latency (D=2.80 min, 95%CI 1.01, 4.52, g = .47), but no significant effects on other sleep parameters. While there was little evidence of publication bias, the studies overall were often of weak methodological quality and the typical study could only detect moderate-to-large impacts. There was also large heterogeneity pointing to substantive differences in effects. Second, multi-level regressions of sleep parameters on emotions reported in the evening from the diary study provided some evidence for delayed sleep on evenings with higher negative affect (b = .05, p < .10), with again no changes in other sleep parameters. Also, higher positive emotions predicted earlier and shorter sleep. The estimates were robust to accounting for emotions at the previous point. Conclusion These pre-registered investigations support the hypothesis that negative emotions delay sleep onset, but evidence regarding other sleep parameters was not conclusive. A diary study of real-life functioning partially replicated delayed sleep onset following more negative emotions, but the effect was modest. The results call for more targeted investigation that disambiguate distinct features of emotions and sleep. Support (If Any)