Expressive suppression is a widely used emotion regulation strategy, but studies have reported mixed findings regarding its association with sleep quality in healthy adults. This meta-analysis aimed to synthesize the evidence on the association between expressive suppression and sleep quality in non-clinical adult samples and examine the moderating effect of cultural background and pandemic context. Electronic databases including PubMed, EBSCO, and Web of Science were systematically searched from database inception to May 2025 using predefined search terms related to emotion regulation and sleep quality. Studies were included if they: (1) involved non-clinical adult samples; (2) assessed expressive suppression with the Emotion Regulation Questionnaire (ERQ); (3) used standardized self-report measures of sleep quality (e.g., PSQI, ISI); and (4) reported zero-order Pearson correlations. A total of twenty-three independent samples (N = 13,636) met the inclusion criteria. A random-effects model was used to estimate the pooled effect size, and pre-specified subgroup and moderation analyses were conducted. Expressive suppression showed a small but significant positive association with poorer sleep quality (r = .14, 95% CI 0.12, 0.16, p < .0001). Between-study heterogeneity was low (I² = 28.21%). The association was stronger in studies explicitly focused on COVID-19–related stress than in studies conducted in general contexts, whereas cultural background, type of sleep measure, study design, and publication period did not significantly moderate the effect. Habitual use of expressive suppression appears to be associated with poorer sleep quality in healthy adults, and this association may be stronger under conditions of major psychosocial stress such as the COVID-19 pandemic. Interventions that reduce reliance on suppression or promote more flexible emotion regulation may help support sleep health, particularly during periods of widespread stress. This study has been registered with the PROSPERO (registration number: CRD420251059053).
Wang et al. (Thu,) studied this question.