Association Between Disturbed Sleep and Depression in Children and Youths

<h3>Importance</h3> Disturbed sleep represents a potentially modifiable risk factor for depression in children and youths that can be targeted in prevention programs. <h3>Objective</h3> To evaluate the association between disturbed sleep and depression in children and youths using meta-analytic methods. <h3>Data Sources</h3> Embase, MEDLINE, PsycINFO, Scopus, Web of Science, and ProQuest Dissertations &amp; Theses Global were searched for articles published from 1980 to August 2019. <h3>Study Selection</h3> Prospective cohort studies reporting estimates, adjusted for baseline depression, of the association between disturbed sleep and depression in 5- to 24-year-old participants from community and clinical-based samples with any comorbid diagnosis. Case series and reports, systematic reviews, meta-analyses, and treatment, theoretical, and position studies were excluded. A total of 8700 studies met the selection criteria. This study adhered to the guidelines of the Preferred Reporting Items for Systematic Reviews (PRISMA) and Meta-analyses and the Meta-analysis of Observational Studies in Epidemiology (MOOSE) statements. <h3>Data Extraction and Synthesis</h3> Study screening and data extraction were conducted by 2 authors at all stages. To pool effect estimates, a fixed-effect model was used if<i>I</i>² &lt; 50%; otherwise, a random-effects model was used. The<i>I</i>²statistic was used to assess heterogeneity. The risk of bias was assessed using the Research Triangle Institute Item Bank tool. Metaregression analyses were used to explore the heterogeneity associated with type of ascertainment, type of and assessment tool for disturbed sleep and depression, follow-up duration, disturbed sleep at follow-up, and age at baseline. <h3>Main Outcome and Measures</h3> Disturbed sleep included sleep disturbances or insomnia. Depression included depressive disorders or dimensional constructs of depression. Covariates included age, sex, and sociodemographic variables. <h3>Results</h3> A total of 22 studies (including 28 895 patients) were included in the study, of which 16 (including 27 073 patients) were included in the meta-analysis. The pooled β coefficient of the association between disturbed sleep and depression was 0.11 (95% CI, 0.06-0.15;<i>P</i> &lt; .001; n = 14 067;<i>I</i>² = 50.8%), and the pooled odds ratio of depression in those with vs without disturbed sleep was 1.50 (95% CI, 1.13-2.00;<i>P</i> = .005; n = 13 006;<i>I</i>² = 87.7%). Metaregression and sensitivity analyses showed no evidence that pooled estimates differed across any covariate. Substantial publication bias was found. <h3>Conclusions and Relevance</h3> This meta-analysis found a small but statistically significant effect size indicating an association between sleep disruption and depressive symptoms in children and youths. The high prevalence of disturbed sleep implies a large cohort of vulnerable children and youths who could develop depression. Disrupted sleep should be included in multifaceted prevention programs starting in childhood.