For tens of millions of people whose jobs begin when most of the world sleeps, the body’s internal clock is a persistent adversary. Shift work is linked to a long catalog of harms, from insomnia to depression and serious long-term illness, yet treatment options that fit around rotating schedules are scarce. A new randomized trial tested one: anonymous, text-based counseling grounded in cognitive behavioral therapy for insomnia, offered preventively to German shift workers over four weeks. The result was null, and a quiet endorsement of something far simpler.
Sleep quality improved markedly among workers who received the personalized online counseling. But it improved just as much among the waiting-list control group, whose members only filled in four weekly sleep diaries; statistically, the two groups were indistinguishable. The study, published in Scientific Reports, could not demonstrate that expert counseling outperformed a diary and a bit of attention. That lands on a question sleep researchers have circled for years: how much of the benefit of guided sleep therapy comes from the therapy itself, and how much from simply paying attention to your own sleep?
What they did
This was a prospective, randomized, controlled superiority trial with two parallel groups, conducted in Germany from January 2020 to June 2022. Sixty-six shift workers completed the baseline assessment: 44 came through six industry partners (two hospitals, one eldercare chain, and three engineering and manufacturing firms), plus 22 who applied after recruitment opened to the public in December 2021. The sample was 59% male, average age 43.8 years (range 23 to 60). A planned face-to-face arm at a sleep clinic was abandoned when the pandemic overwhelmed the health system.
Anonymity was built into every layer of the design. Participants registered with nothing but an email address, instructed to use one free of names or other identifying details. Assignment was carried out by a staff member of a separate institute who had no other involvement in the trial. Two trained psychologist-counselors delivered the intervention asynchronously, entirely in text, investing about two hours per participant over the four-week program.
The counseling adapted the standard CBT-I toolkit to rotating-shift realities. Classic sleep restriction assumes regular bedtimes, impossible when work hours drift across the clock. Instead, the counselors built sleep windows around each worker’s schedule, emphasizing anchor sleep, the longest overlap between times in bed, and allowing sleep to be split into two windows per 24 hours around night shifts. Near shift changes, bedtimes could flex by up to three hours and 90 minutes. Participants below 85% sleep efficiency received sleep restriction instructions; the rest kept regular bedtimes without restriction. Weekly messages added personalized sleep hygiene tips, relaxation training, and tailored progress feedback.
The control group completed the same four weekly sleep diaries, received a brief motivational message after each, and were offered the counseling after the waiting period. The primary outcome was sleep efficiency, the ratio of sleep time to time in bed, averaged from the diaries. Secondary outcomes were insomnia, depression, and daytime sleepiness, measured by standard questionnaires before and after the four weeks.
Attrition was heavy, as is typical of online studies: 59% of participants dropped out overall, about 12% per week. Twenty-seven participants (15 counseling, 12 control) completed at least four diaries, and 22 provided complete data. The main analysis used intention-to-treat with all 66.
What they found
Both groups improved, and neither pulled ahead. Sleep efficiency rose by 3.6 percentage points in the counseling group and 4.9 in the control group, both significant gains. Insomnia and depression symptoms also fell significantly in both groups, with large effect sizes (Cohen’s d from 0.90 to 1.28). Daytime sleepiness did not change in either group.
The decisive test was the interaction between time and treatment group, whether the counseling group changed more than the control group. Every time-by-group interaction test came back non-significant (all p values above 0.10), so the counseling could not be shown to add anything beyond the diaries. The pattern survived every robustness check, holding in the per-protocol sample of 27 and when participants already above the clinical threshold were excluded to rule out a ceiling effect. Participants rated satisfaction at 3.6 out of 4 on average; perceived helpfulness was more subdued at 2.9.
The most plausible explanation is also the most interesting: the control condition may have been therapeutic in its own right. Four weeks of recording bedtimes, wake times, and sleep duration forces attention onto sleep, and people who notice suboptimal routines tend to adjust them. In that reading, the null result is not evidence that the counseling failed, but that the diary, by itself, may be an active ingredient.
Why it matters
Between 10% and 23% of shift workers meet diagnostic criteria for a sleep disorder, and many more hover below the threshold with chronically short, restless sleep. For insomnia, cognitive behavioral therapy is the standard first choice in most clinical guidelines, yet it has seldom been adapted and rigorously tested in rotating-shift populations, because its core tool, sleep restriction, clashes with irregular schedules. This trial shows a workable adaptation exists, that it can be delivered anonymously and preventively across industries, and that it is feasible in the field, even if its specific contribution could not be separated from the diaries themselves.
The larger lesson concerns the humble sleep diary. If self-monitoring carries much of the therapeutic weight, cheap, low-threshold diary programs, or even wearable trackers, may be a viable prevention tool for employers and occupational health services. The results also underscore why sleep research needs active control groups: without the diary-only arm, the counseling would have looked convincingly effective.
Limits
The trial carries important caveats. The sample was small, and the observed effect size for sleep efficiency (d = 0.66) came in below the pilot estimate (d = 0.76) that drove the power calculation, so a modest advantage of counseling could have been missed. There was no follow-up, so long-term effects are unknown; the waiting-list design made longer follow-up ethically impractical. All outcomes were subjective, since anonymity ruled out actigraphy. And recruitment ran through the worst of the COVID-19 pandemic, which strained hospital and care partners, shrank manufacturing hours, and shaped who enrolled and how they slept.
Bottom line
As null results go, this one is unusually constructive. It answers the question it asked: over four weeks, an anonymous, personalized, preventive online counseling program did not beat a sleep diary plus study participation for shift workers. Both groups slept better, and the gap between them was indistinguishable from zero. Rather than a verdict against the counseling, it is a reason to take the simplest instrument in sleep science seriously as a preventive tool, and to keep testing whether guidance, timing, or format adds measurable value.
Source
Retzer L, Graessel E, Feil M, Lehmann R, Stemmler M, Richter K. Efficacy of an anonymous, preventive, personalized online counselling to improve shift workers’ sleep quality: a randomized controlled trial. Scientific Reports. 2026;16:23559. DOI: 10.1038/s41598-026-64210-7. PMID: 42527422. Open access; trial registration: DRKS00017777.