Digital behavior therapy for insomnia in the general population: secondary analysis of the multi-arm RESiLIENT trial...

Insomnia is common and often coexists with depression and anxiety. Digital delivery can expand access, yet few trials have compared insomnia-focused behavioral interventions with strong active controls under standardized conditions.

To examine whether a smartphone-delivered behavior therapy for insomnia (BI) improves insomnia more than a mood-focused behavioral activation (BA) module or a health-information (HI) comparator, and whether combining BI with BA (BA+BI) adds benefit.

We conducted a secondary analysis of the multi-arm RESiLIENT randomized controlled trial among adults from the general population with elevated insomnia (Insomnia Severity Index (ISI) ≥8). Participants were assigned to BI, BA+BI, BA, or HI, each self-guided over six weeks. The primary outcome was ISI change at weeks 6 and 26, analyzed using mixed-effects models. Depressive (PHQ-9) and anxiety (GAD-7) symptoms were secondary outcomes. Analyses followed a modified intention-to-treat approach (≥1 post-baseline assessment).

At week 6, BI reduced insomnia severity more than HI (ES: -0.43, CI: -0.58 to -0.28) and BA (ES: -0.26, CI: -0.42 to -0.11); BA also exceeded HI (ES: -0.16, CI: -0.32 to -0.01). Adding BA to BI gave no benefit over BI alone (vs HI, ES: -0.30, CI: -0.45 to -0.14). BI reduced depressive (ES: -0.44, CI: -0.59 to -0.29) and anxiety symptoms (ES: -0.27, CI: -0.42 to -0.12) more than HI, and matched BA (each ES: 0.05, CI: -0.10 to 0.20). At week 26, ISI differences persisted but PHQ-9 and GAD-7 differences attenuated.

This exploratory, hypothesis-generating analysis associated a focused smartphone-delivered sleep intervention with durable improvements in insomnia severity and mood. Replication in independent samples, and confirmation in trials recruiting participants with elevated insomnia, is required before recommendation at scale.