Efficacy of lifestyle medicine on sleep quality: A meta-analysis of randomized controlled trials.

Objectives: Randomized controlled trials (RCTs) on the efficacy of multicomponent lifestyle medicine (LM) interventions for improving sleep quality have yielded inconsistent findings. This study marks the first meta-analysis to evaluate the efficacy of multicomponent LM interventions in improving sleep quality.
Methods: We searched six online databases for RCTs that compared multicomponent LM interventions to an active or inactive control group in an adult population and assessed subjective sleep quality as a primary or secondary outcome using validated sleep measures at any post-intervention time-point.
Results: A total of 23 RCTs with 26 comparisons involving 2534 participants were included in the meta-analysis. After excluding outliers, the analysis revealed that multicomponent LM interventions significantly improved sleep quality at immediate post-intervention (d = 0.45) and at short-term follow-up (i.e., <three months) (d = 0.50) relative to an inactive control group. Regarding the comparison with active control, no significant between-group difference was found at any time-point. No meta-analysis was conducted at the medium- and long-term follow-up due to insufficient data. Subgroup analyses supported that multicomponent LM interventions had a more clinically relevant effect on improving sleep quality in participants with clinical levels of sleep disturbance (d = 1.02) relative to an inactive control at immediate post-intervention assessment. There was no evidence of publication bias.
Conclusion: Our findings provided preliminary evidence that multicomponent LM interventions were efficacious in improving sleep quality relative to an inactive control at immediate post-intervention and at short-term follow-up. Additional high-quality RCTs targeting individuals with clinically significant sleep disturbance and long-term follow-up are warranted.
Competing Interests: Conflict of interest The authors have no competing interests to report.
(Copyright © 2023 Elsevier B.V. All rights reserved.)