Efficacy of Digital Health Interventions for Insomnia: Systematic Review and Meta-Analysis of Randomized Controlled Trials
Journal of Medical Internet Research ·
Background: Insomnia is common, and treatments are limited by tolerability and access. Digital health interventions (DHIs) are increasingly used for sleep management, but evidence comparing multiple outcomes is scarce. Objective: This study aimed to systematically evaluate the effects of DHIs on 8 sleep outcomes, including insomnia severity, sleep efficiency, total sleep time, and sleep quality. Methods: We systematically searched PubMed, Embase, Cochrane Library, Web of Science, Scopus, CINAHL, CNKI, CBM, VIP, Wanfang, ClinicalTrials.gov, and the American Academy of Sleep Medicine’s website from inception to July 2026. Two reviewers (SG and BZ) independently screened studies, extracted data, and assessed risk of bias using the Cochrane Risk of Bias 2.0 tool. We included randomized controlled trials or controlled clinical trials comparing DHIs with any control in insomnia, without language restriction, and excluded experimental sleep deprivation, conference abstracts, protocols, and nonrandomized or noncontrolled designs. Pooled standardized mean differences (SMDs) with 95% CIs used restricted maximum likelihood with Hartung-Knapp-Sidik-Jonkman correction; 95% prediction intervals (PIs) used the Nagashima method. We also performed subgroup analyses by intervention category, comparator type, diagnostic criteria, and country income, plus sensitivity analyses, Egger tests, and GRADE (Grading of Recommendations Assessment, Development and Evaluation) assessment. Results: We included 44 randomized controlled trials (n=10,521) published between 2004 and 2026 across 14 countries. DHIs significantly improved insomnia severity (SMD −0.91, 95% CI −1.12 to −0.70), sleep efficiency (SMD 0.59, 95% CI 0.45-0.72), total sleep time (SMD 0.20, 95% CI 0.11-0.29), sleep quality (SMD −0.61, 95% CI −0.78 to −0.45), and secondary outcomes (SMDs −0.37 to −0.64). However, 95% PIs crossed the null for insomnia severity (−2.04 to 0.22), total sleep time (−0.05 to 0.46), and all secondary outcomes. Subgroup analyses showed larger insomnia severity effects in high- versus upper-middle-income countries, modality differences for sleep efficiency, and comparator-type differences for total sleep time (all <.01). Egger tests suggested small-study effects; sensitivity analyses did not change direction or significance. GRADE certainty ranged from very low to moderate. Conclusions: Although DHIs showed statistically significant average effects on multiple sleep outcomes, wide PIs crossing or approaching the null, effects approaching or below Minimum Clinically Important Difference thresholds, and very low-to-moderate GRADE certainty indicate that the evidence remains exploratory. Subgroup analyses revealed effect modifications by country income, intervention modality, and comparator type, underscoring contextual factors. Unlike previous reviews focusing on single DHI formats or average statistical effects, this review provides stratified comparisons and integrates PIs and Minimum Clinically Important Difference conversion to distinguish statistical significance from clinical relevance. This paper provides an evidence synthesis framework showing that DHI benefits are context-dependent and insufficient for a standard clinical role. In practice, DHIs cannot be recommended as a standard substitute for face-to-face cognitive behavioral therapy for insomnia, although they may serve as a low-intensity option in stepped care. Large-scale, head-to-head implementation trials in diverse populations with extended follow-up are needed to define their optimal role. Trial Registration: PROSPERO CRD420251207504; https://www.crd.york.ac.uk/PROSPERO/view/CRD420251207504