A fully automated online insomnia programme still helped two years later — but the gap narrowed

A programme that teaches the core skills of CBT-I — the cognitive-behavioural approach to insomnia — entirely online, with no human contact, still left people sleeping better than a sleep-advice website two years later. The advantage was real but roughly halved over time, and only about a third of participants answered the final questionnaire, so the long-term result is less certain than the headline suggests.

What they found

Researchers in Norway randomly assigned 1,720 adults with significant insomnia to one of two online options. Half got SHUTi, a fully automated programme delivered in six weekly "cores" covering sleep restriction, stimulus control, thinking about sleep, sleep habits and relapse prevention, with feedback based on a daily sleep diary. The other half got a static website with information about insomnia and standard advice, including stimulus control and sleep hygiene.

Insomnia was measured with the Insomnia Severity Index (higher is worse; both groups started at about 19). After a median of 28 months, the programme group averaged 10.7 and the website group 13.4. The difference was 1.8 points (95 % confidence interval 0.9 to 2.7) — a small-to-moderate effect (d = 0.43). At six months the effect had been larger (d = 0.83), and at nine weeks larger still (d = 1.21). People in the programme group also reported less fatigue, but not less psychological distress. The authors estimate the programme was likely cost-effective over two years.

How much to trust this

Evidence: Moderate evidence. This is a large, pre-registered randomised trial with a realistic comparison — an advice website that already contained some of the same rules — which makes the result more convincing than a comparison with "nothing".

The weak spot is follow-up. Only 587 of 1,720 people (34 %) provided two-year data. When the authors tested how sensitive the result was to missing answers, the two-year difference stopped being statistically clear if the non-responders in the programme group were assumed to be only about 1.4 points worse than predicted. Only 46 % of the programme group completed all six cores. Participants were mostly women (68 %), on average 44 years old, self-referred and well educated — not everyone. Side effects were not systematically recorded.

The trial was publicly funded (Research Council of Norway). One author developed SHUTi and has a financial stake in companies that license it; the paper states those companies had no role in the manuscript.

Disclosure: Resteady also makes a digital sleep programme. We have no ties to the authors or to SHUTi.

What this means for you

If poor sleep has dragged on for months, a structured, step-by-step programme is worth more than another list of sleep tips: the website group got the tips and still did worse. The core of these programmes is behavioural, and the hardest part is doing it consistently for several weeks.

• Start with the rules that re-teach your bed to mean sleep: The six rules that re-teach the bed.
• Keep a simple morning diary for a week before changing anything: Why we start with a diary, not a fix.
• Read how a sleep window works before trying one — and check the safety screen first; it is not for everyone (for example, people who drive for work or have a history of bipolar-type mood swings): The sleep window.

If insomnia is severe, keeps you from functioning, or comes with loud snoring and gasping at night, talk to a doctor or a sleep specialist rather than relying on a self-guided programme.

What this does NOT mean

• It does not mean every app works: this is one well-studied programme with a fixed structure.
• It does not mean the benefit stays the same size — it shrank over two years, and many people stopped responding.
• It does not mean self-guided is as good as working with a specialist; the trial did not compare the two.
• It does not mean total sleep time went up — it did not differ between groups; the gains were in sleep continuity and how people rated their insomnia.

Source

Vethe D, Khan ZA, Vestergaard CL, et al. Long-term clinical and cost-effectiveness of a fully automated digital cognitive behavioural therapy for insomnia: 2-year follow-up of a single-blind, superiority, randomised controlled trial. The Lancet Regional Health – Europe, 2026;66:101691 (online 8 May 2026). DOI: 10.1016/j.lanepe.2026.101691 · PMID 42164344 · Open access: PMC13186012 · Trial registration NCT02558647.