Continuous sleep tracking in digital CBT-I: Efficacy and insights from a naturalistic-environment study

TL;DR

A digital CBT-I program with continuous sleep tracking improved insomnia and well-being in a real-world study over eight weeks.

Contribution

The study introduces a novel app-based CBT-I program with continuous sleep tracking in a naturalistic setting.

Findings

Insomnia prevalence decreased from 92% to 67% after the intervention.

Subjective sleep quality and psychological strain improved significantly.

Objective data showed trends toward reduced total sleep time and light sleep.

Abstract

Insomnia is highly prevalent, yet few receive cognitive behavioral therapy for insomnia (CBT-I) due to limited treatment availability. Unguided digital CBT-I offers an accessible alternative to traditional face-to-face therapy. Research in this area often relies on either subjective sleep measurements (e.g., sleep diaries) or controlled single-night lab studies. This study examines the effectiveness of a novel app-based CBT-I program combining therapy with continuous subjective and objective sleep tracking via a heart rate (HR) sensor in a naturalistic setting.

Eighty-eight participants (56.8 % female) aged 20–85 years (M = 49.9 ± 13.10) completed an 8-week app-based CBT-I intervention with continuous sleep tracking (sleep diaries and HR sensor), followed by a 2-week follow-up. Assessments at baseline, post-intervention, and follow-up included sleep related problems, psychological strain, quality of life, and dysfunctional beliefs.

Insomnia prevalence dropped from 92 % at pre-intervention to 67 % at follow-up. Improvements were observed in subjective sleep quality (p < .001, r = 0.59), dysfunctional beliefs (p < .001, r = 0.48), quality of life (p’s < .002, r’s > 0.33), psychological strain (p < .001, r = 0.43), depression (p = .010, r = 0.27), and anxiety (p = .003, r = 0.32). While sleep diary data showed improvements in various sleep parameters, objective data revealed statistical trends towards a reduced total sleep time (TST; p = .083, r = 0.19), driven by sleep restriction, and light sleep (p = .089, r = 0.18). Using continuous sleep monitoring we additionally found relevant changes during the intervention levels for subjective wake after sleep onset, sleep efficiency as well as objective TST.

Findings support the effectiveness of app-based CBT-I and suggest that continuous objective sleep tracking over weeks can reveal previously undetected sleep and well-being improvements in real-world settings.

Figures