Sleeppreliminary · human dataAdded 30 July 2026

Sleep diaries alone matched online sleep coaching

In a randomised trial of German shift workers, four personalised online counselling messages did not improve sleep efficiency any more than simply completing weekly sleep diaries. Sleep efficiency, insomnia and depression symptoms improved in both groups, while daytime sleepiness improved in neither.

Why it matters

Shift work is associated with an increased risk of sleep disorders, yet tailored prevention and treatment programmes for shift workers remain rare and understudied. Cognitive behavioural therapy for insomnia is the first-line treatment in the general population, but it is usually delivered face to face over several sessions, which is awkward for people working rotating nights. An anonymous, low-cost online service built on the same cognitive behavioural principles could in theory reach workers who would never book a clinic appointment. The question this trial set out to answer was whether such a service actually outperforms doing nothing much at all.

What they did

The researchers developed an anonymous online counselling service based on cognitive behavioural principles, adapted it to shift workers' needs, and tested it in German shift workers drawn from various industries. This was a prospective, randomised, controlled superiority trial with two parallel groups. The intervention group received four personalised counselling messages and completed four weekly sleep diaries; the waiting-list control group completed sleep diaries only across the same four-week waiting period. The primary outcome was sleep efficiency, with symptoms of insomnia and depression plus daytime sleepiness as secondary outcomes. The intention-to-treat sample comprised 66 participants, and the trial was pre-registered in the German Clinical Trials Register.

What they found

On the primary outcome, sleep efficiency, significant improvements were observed in both groups over the four weeks. The same pattern held for the secondary outcomes of insomnia and depression symptoms, which also improved in both groups. Crucially, the improvements were not significantly greater in the counselling group than in the diary-only control group, so superiority could not be demonstrated. Daytime sleepiness did not improve in either group. The authors note that repeated sleep diary completion and study participation may themselves have exerted therapeutic effects, and flag the short time frame and modest sample size as limitations.

What it actually shows

One small superiority RCT, intention-to-treat n=66 German shift workers, four weeks, waiting-list control that also kept sleep diaries; self-reported outcomes and the authors flag limited sample size and time frame.

RCT · Sci Rep

Where it fits

Cognitive behavioural therapy for insomnia is already the first-line treatment for insomnia in the general population, and this trial does not challenge that. What it complicates is the assumption that a stripped-down, message-based adaptation adds measurable value on top of structured self-monitoring. Because the control condition involved four weeks of sleep diaries rather than genuinely doing nothing, the trial cannot separate an ineffective intervention from an unexpectedly active comparator. Open questions include whether more counselling contacts, a longer follow-up or a larger sample would reveal a difference, and how much of the benefit seen here is simply attention and structure.

What it means for you

The honest read is that a light-touch, four-message online counselling package was not shown to beat weekly sleep diary keeping in shift workers. That is a reason to be sceptical of short digital sleep programmes marketed as clinically superior, and a reason to take structured self-monitoring more seriously than it usually gets taken. It is not evidence that cognitive behavioural approaches to insomnia do not work, since both arms improved and neither was compared with genuine inaction. If your sleep is disrupted by shift patterns, this study says the format and intensity of support probably matter, and that a single small trial cannot tell you which format wins.

Try the calculator →

More new research, with the reality check: This week in the science → · or open the full Pulse feed →