In plain words. Losing sleep makes the days that follow feel flatter and more anxious. In trials where treatment, mostly a structured treatment called CBT-I (cognitive behavioural therapy for insomnia), improved people’s sleep, their symptoms of depression and anxiety eased too. Sleep-hygiene tips on their own help less than CBT-I. This article is for anyone losing sleep, whether over a hard stretch at work or anything else, and it suggests one small experiment that none of the studies cited here has tested on its own: for seven days, getting up at the same time every morning (the latest time you can keep on all seven days) and writing one line each evening. This is general education, not treatment or personal advice.
The night
What follows is not one real person’s night; it is pieced together from very common ones.
It is ten past three in the morning, and you are giving a presentation to an audience of one: the bedroom ceiling, which has now heard it four times and still offers no feedback. The second slide gets a little better with every run-through. You check your phone, as if it might know something you don’t. It tells you the time. Then you replay Friday’s email to your manager, weighing one sentence for bluntness, as though it might have grown sharper since Friday. And underneath it all sits the fear that does most of the keeping-awake: you will be tired tomorrow, and tomorrow is the one day you cannot afford to be.
You have already done everything the sleep-hygiene list asks: screens off at nine, a cool, dark room, no coffee after lunch. You have been a model sleeper in every respect except the sleeping. It has not helped. And you can feel the loop tightening: anxious because you cannot sleep, unable to sleep because you are anxious. It feels like a personal failing. It is not.
What the research says
The good news is the unglamorous kind: much of what happens on a night like this has been studied carefully. Here is the short version. The studies, the numbers and the finer points behind each statement are in The working, at the end of the article.
Losing sleep does make the days that follow harder. In experiments where healthy volunteers had their sleep taken away, cut short or broken on purpose, they felt less positive and more anxious than after normal sleep. Studies in which people recorded their sleep and mood every day found a similar back-and-forth in ordinary life: a poor night tended to be followed by a worse day, and a hard day by a worse night. The loop you feel at ten past three plays out within one night; these studies saw a similar pattern across days. It is a pattern researchers have documented, not a flaw in you. What no study can tell you is how much harder your own Thursday will be, because these results are averages across many people.
When treatment improves sleep, depression and anxiety tend to ease too. In trials, when people’s sleep improved, mostly through CBT-I, their depression and anxiety symptoms, measured by questionnaire, improved more than those of people in the comparison groups. On average the change was medium-sized: noticeable, not dramatic.
The sleep-hygiene list is not enough on its own. The main American sleep-medicine guideline advises doctors not to rely on sleep hygiene alone to treat long-lasting insomnia. Two trials show the gap. In the first, older adults with insomnia took either a CBT-I course or a sleep-education course of the same length; over the following years, the CBT-I group was about half as likely to develop depression, whether for the first time or again. In the second, adults with insomnia from family doctors’ practices all received a sleep-hygiene booklet, and half also had a few sessions of sleep restriction therapy, a supervised part of CBT-I that shortens time in bed; six months later, that half reported milder insomnia than those with the booklet alone.
That is not you failing the list. On its own, sleep-hygiene advice simply does less than CBT-I: a checklist, not a treatment.
Some of it may not be yours to fix. Heavy demands at work, job strain, bullying and effort that goes unrewarded are all linked with more sleep problems later; support at work, some control over your work and fair treatment are linked with fewer. These are links, not proof, although the researchers who reviewed them conclude that work does affect sleep. A restructuring, a new manager or an impossible deadline can make several of those conditions worse at once. Outside work, a small child, a parent who needs you at night or a contract about to end can cut into sleep too. And the email you keep replaying is a work worry: no sleep tip will settle it.
So this article will not tell you that your sleep is entirely yours to fix. Some of it may not be yours at all.
The experiment
You cannot order sleep to arrive; it has never once responded to being told. In most weeks, though, you can decide when the day begins. So the experiment is this: for seven days, you get up at the same time every morning, whatever the night was like.
A fixed getting-up time is one of the standard instructions in CBT-I. None of the studies cited here tested it on its own, and many sleep-hygiene programmes already include a regular sleep schedule as one tip among many. Here it is the only change to your sleep, kept every day and written down, so that you can watch how your nights and days go while you keep it.
Which time? After a bad night, a fixed getting-up time gives you less sleep than a lie-in would, and shorter nights tend to be followed by flatter days. So the time to choose is the latest one you can keep on all seven days: steady every morning, but with as much room for sleep as your week allows. Your week sets that limit, whether it is a rota, the school run or a child who wakes at six whatever you decide.
The studies cited here also give no reason to get up early. In a study of junior doctors, mood was a little better on days after they had got up later than usual, and also after days when their sleep was steadier than usual. For this experiment, those two findings pull in opposite directions: a fixed time makes your mornings steadier but rules out getting up later than usual. That is one more reason to choose the latest time you can keep. This choice is our reasoning, not a tested result.
What stays the same. Your bedtime. The only change to your sleep is the alarm. This is not sleep restriction therapy, which deliberately shortens time in bed under supervision; you simply give up the lie-in. That can still mean a shorter night after a bad one, so some tiredness is possible. If daytime sleepiness starts to affect your driving or safety, that is a reason to stop and see your doctor. If a doctor has told you that losing sleep is a risk for you, or you work nights or rotating shifts, talk to them before you start.
Running the experiment
The one rule: the alarm goes off at the same minute after a good night and after a terrible one.
Tonight, you choose your time and set your phone’s alarm for all seven days. From tomorrow evening, you write one line each evening, before bed rather than at three in the morning. Each line has three parts: last night (roughly when you fell asleep, and how it went, in your own words rather than a sleep tracker’s score); today (how it went, and whether it was better or worse than you had feared); and anything unusual, such as a late coffee or a hard meeting. The practice sheet below has space for all seven lines.
After seven evenings, the lines can be read with two questions in mind. Were your hardest days the ones after your shortest nights? And what did the week show you about your own sleep and mood? The notes cannot prove what the fixed time itself did, because there is no ordinary week to compare with; what they can show is how your nights and days moved together. Then the choice is yours: if the week felt worth it, you can keep the time; if it did not, you have lost a few lie-ins and learned something about your own sleep.
When to get help. A week of notes cannot show everything, and this article is education, not treatment. Your doctor is the first person to talk to if your sleep has been poor for more than a few weeks, or at any point if it worries you; if you snore loudly, or someone has noticed pauses in your breathing or gasping; if you wake unrefreshed despite enough time in bed; if daytime sleepiness affects your driving or safety; or if low mood does not lift when your sleep improves. For long-lasting insomnia, sleep-medicine guidelines strongly recommend CBT-I. And if what you feel is more than tiredness, if it feels like you cannot cope, talk to your doctor about that too. If you need support before you can see them, the Start Here page on this site lists helplines you can reach straight away.
One limit: the people in these studies were healthy volunteers, adults with insomnia, junior doctors in their demanding first year and workers in general, not people in exactly your situation. So the findings are useful for you, but not certain.
So, back to ten past three, on a night during your seven days. The ceiling may still get its presentation, and the email may still get weighed. But the alarm is set, you know you will get up when it rings, and you know the loop is not your fault. None of this promises a better night. What seven evenings give you is something a sleep score cannot: a record, in your own words, of how your nights and days went together. If you want to try it, it starts tonight: choosing the latest getting-up time you can keep all week, setting the alarm, and writing your first line tomorrow evening.
Practice sheet
My getting-up time for the seven days: ________ (the latest time I can keep every day)
One line each evening, starting tomorrow.
| Day | Last night: roughly when I fell asleep, and how it went | Today: how it went, better or worse than I feared | Anything unusual |
|---|---|---|---|
| 1 | |||
| 2 | |||
| 3 | |||
| 4 | |||
| 5 | |||
| 6 | |||
| 7 |
After day 7: Were my hardest days the ones after my shortest nights? What did this week show me about my own sleep and mood?
Download the practice sheet — A4, for printing (PDF)
Download the practice sheet — phone version (PDF)
The working
This section is for readers who want the studies and the numbers behind the article. Everything above stands without it. Every figure here describes a group of people, not a forecast for one person.
Losing sleep makes the days that follow harder
The main source is a 2024 review that combined 154 experiments involving 5,717 healthy people aged 7 to 79, so children were included as well as adults (Palmer and colleagues). In these experiments, researchers kept people awake all night, shortened their nights, or woke them repeatedly, and compared their feelings with feelings after normal sleep. The review reports results separately for each kind of sleep loss. Three effects were clear: fewer positive feelings, more anxiety, and less emotional arousal (things that would normally stir people up stirred them less). For negative feelings in general, emotional tone and depression symptoms, the effect depended on the kind of sleep loss: it appeared with some kinds and not with others, and the review’s published summary does not say which kind did what. The amount of sleep lost mattered, but not in simple proportion. For readers who know the statistics: positive feelings fell by standardised mean differences of 0.27 to 1.14 and anxiety rose by 0.57 to 0.63, depending on the kind of sleep loss. The full text of the review was not openly accessible, so details such as the length of each experiment, and exactly when feelings were measured, were not checked; some experiments may have lasted several nights, so the review may not show the effect of one bad night on its own.
Because these experiments change sleep on purpose, they show which comes first: the lost sleep, then the harder mood. Exactly how the brain turns one into the other is still being worked out, which is why the Ledger grades this mechanism as Plausible rather than Established (Ledger entry 2).
The everyday-life pattern comes from a review of diary studies, in healthy people and in people with health conditions, who recorded their own sleep and mood each day (Konjarski and colleagues, 2018). It found a two-way link over short periods: poorer self-reported nights with worse following days, and harder days with worse following nights. It gives no single combined figure. It describes links between nights and days; it did not study worry and sleeplessness within a single night.
When treatment improves sleep, depression and anxiety tend to ease too
Two large reviews of randomised trials support this. The first (Scott and colleagues, 2021) included 65 trials with 8,608 people, all trials in which the treatment improved sleep quality compared with a comparison group; most treatments were CBT-I, and about half of the comparisons involved people who also had a physical or mental health problem. Depression, anxiety and rumination (going over and over the same worries) improved more than in the comparison groups, by medium amounts (effect sizes of 0.63, 0.51 and 0.49), and trials that improved sleep more saw larger improvements. Two cautions: the review included only trials where sleep improved, and the link between bigger sleep gains and bigger improvements compares trials, not people. So it shows that treatments which improved sleep also eased symptoms; on its own it does not prove that the change in sleep caused the change in symptoms, although the authors interpret it that way.
The second review (Gee and colleagues, 2019) included 49 trials with 5,908 people and found a fall in depression symptoms compared with control conditions (effect size 0.45 overall). The fall was larger in the 7 trials of people with a mental-health problem (0.81) than in the other 42 (0.41, which the authors call small). The first review, by contrast, found benefits whether or not people had other health problems; the two reviews grouped people differently, so their answers to this question are not directly comparable. These are questionnaire scores, not diagnoses. Neither review explains why these symptoms ease when sleep improves. “Noticeable, not dramatic” in the article is our plain-language gloss on “medium”.
Why does this not clash with the mixed result for depression symptoms after sleep loss? That result was “sometimes”: the effect depended on the kind of sleep loss. The two sets of studies also asked different questions of different people: one took sleep away from healthy volunteers in experiments; the other improved sleep through treatment, often in people who were already struggling.
The sleep-hygiene list is not enough on its own
The American Academy of Sleep Medicine’s 2021 clinical practice guideline (Edinger and colleagues) strongly recommends CBT-I for long-lasting (chronic) insomnia and advises clinicians not to use sleep hygiene as a treatment on its own; that second recommendation is graded “conditional”, a weaker grade than “strong”. A review of 15 studies (Chung and colleagues, 2018) found that sleep-hygiene education improved people’s own reports of their sleep less than CBT-I did. About half of those sleep-hygiene programmes already covered regular sleep and waking times. None of those 15 studies compared sleep hygiene with no treatment at all. The American Family Physician summary of the guideline mentions one study in which sleep-hygiene education did better than a waiting list, and it also describes typical sleep-hygiene education as minimally effective. These studies tested structured education programmes, which may not match the three tips in the article’s opening.
The first trial (Irwin and colleagues, 2022) involved 291 adults aged 60 or over with insomnia and no depression in the previous year. For two months, one group attended weekly CBT-I sessions led by a psychologist; the other attended a sleep-education course of the same length and format, led by a health educator, covering sleep hygiene, sleep biology and stress. Over up to three years, 12% of the CBT-I group and 26% of the sleep-education group were diagnosed with depression in a clinical interview (a first episode or a return of an earlier one), which makes the CBT-I group about half as likely to develop it (hazard ratio 0.51; the authors’ pre-planned adjusted figure is 0.45). About half of each group completed the full three years, and the share who stayed differed between the groups, so the raw percentages should be read with care; the hazard ratio takes follow-up time into account.
The second trial (the HABIT trial; Kyle and colleagues, 2023) involved 642 adults with insomnia disorder from 35 family doctors’ practices in England. All received a sleep-hygiene booklet; half also had four sessions of sleep restriction therapy, delivered by nurses. Sleep restriction therapy limits the time spent in bed, starting from a person’s average sleep in a sleep diary, and can bring short-lived tiredness and poor concentration. After six months, the therapy group’s scores on the Insomnia Severity Index, a questionnaire, were about three points lower (better). The first trial tested a whole treatment, so it cannot show which part of CBT-I made the difference; its two groups also had different leaders, so the result may reflect who taught the sessions as well as what was taught. The second trial added one part of CBT-I, sleep restriction therapy, to the booklet that everyone received. What both show is that sleep advice alone did less well: on depression diagnoses in the first trial and on insomnia scores in the second. The sleep advice itself differed: a two-month taught course in the first trial, a booklet in the second.
Work and sleep
A review of 24 studies that followed workers over time (Linton and colleagues, 2015) linked high demands, job strain, bullying and effort–reward imbalance with more sleep problems later, and social support at work, control over work and fair treatment at work with fewer. These are associations, not proof of cause, although the review’s authors themselves conclude that work factors have an impact on sleep. The review’s published summary does not mention job insecurity; the point about restructuring is our description, not one of its findings.
The experiment
A fixed getting-up time is one of the standard instructions in stimulus control, one of the methods in CBT-I (the AASM guideline, as summarised in American Family Physician, 2022). Stimulus control also includes other instructions, such as going to bed only when sleepy; the experiment uses only the fixed getting-up time. It has not been tested on its own as a way to improve mood in any study cited here.
The timing evidence comes from one observational study (Fang and colleagues, 2021) that followed 2,115 first-year junior doctors in the United States with wearable sleep trackers, a daily mood rating from 1 to 10, and a depression questionnaire every three months. Within each doctor, mood was a little better after nights with more sleep than that doctor’s usual, after earlier bedtimes, after later getting-up times than usual, and on days when sleep length and getting-up time were closer to the previous day’s. Across doctors, less sleep, later bedtimes and more irregular sleep went with slightly higher depression scores; the authors call the effect of irregular sleep substantial, but their own figures show it explained only a small part of the differences between doctors. Because the study only observed, it cannot show that steadier sleep causes better mood; the authors hope future trials will test whether steadier sleep reduces depression.
Keeping the getting-up time fixed changes two things at once: it makes timing steadier, and after a bad night it leaves less time for sleep. In the doctors’ study these pulled in opposite directions, so the evening notes cannot separate them. Choosing the latest getting-up time you can keep every day is our reasoning from these findings, not a tested result.
What does not come from the cited studies
The list of reasons to see a doctor is general guidance, not a finding of these studies. The points about restructuring, new managers, deadlines, caring duties and contracts, and the remark that no sleep tip will settle a work worry, are our description. That giving up a lie-in can bring some tiredness is our reasoning. The advice to check with a doctor before starting, if losing sleep is a known risk for you or you work nights or rotating shifts, is general caution, not a finding of these studies. Keeping bedtime unchanged, so that only one thing about your sleep changes, is a design choice. That the loop is not a personal failing or a flaw in you is our view, not a finding of these studies; so is the point that you cannot choose when you fall asleep but, in most weeks, can choose when you get up.
Sources
- Palmer CA et al. Sleep loss and emotion: A systematic review and meta-analysis of over 50 years of experimental research. Psychol Bull. 2024;150(4):440–463. PMID 38127505 · doi:10.1037/bul0000410
- Konjarski M et al. Reciprocal relationships between daily sleep and mood: A systematic review of naturalistic prospective studies. Sleep Med Rev. 2018;42:47–58. PMID 30404728 · doi:10.1016/j.smrv.2018.05.005
- Scott AJ et al. Improving sleep quality leads to better mental health: A meta-analysis of randomised controlled trials. Sleep Med Rev. 2021;60:101556. PMID 34607184 · doi:10.1016/j.smrv.2021.101556
- Gee B et al. The effect of non-pharmacological sleep interventions on depression symptoms: A meta-analysis of randomised controlled trials. Sleep Med Rev. 2019;43:118–128. PMID 30579141 · doi:10.1016/j.smrv.2018.09.004
- Edinger JD et al. Behavioral and psychological treatments for chronic insomnia disorder in adults: an American Academy of Sleep Medicine clinical practice guideline. J Clin Sleep Med. 2021;17(2):255–262. PMID 33164742 · doi:10.5664/jcsm.8986
- Behavioral and Psychological Treatments for Chronic Insomnia Disorder: Updated Guidelines From the American Academy of Sleep Medicine. Am Fam Physician. 2022;105(1):97–98. https://www.aafp.org/afp/2022/0100/p97
- Chung KF et al. Sleep hygiene education as a treatment of insomnia: a systematic review and meta-analysis. Fam Pract. 2018;35(4):365–375. PMID 29194467 · doi:10.1093/fampra/cmx122
- Irwin MR et al. Prevention of Incident and Recurrent Major Depression in Older Adults With Insomnia: A Randomized Clinical Trial. JAMA Psychiatry. 2022;79(1):33–41. PMID 34817561 · doi:10.1001/jamapsychiatry.2021.3422
- Kyle SD et al. Clinical and cost-effectiveness of nurse-delivered sleep restriction therapy for insomnia in primary care (HABIT). Lancet. 2023;402(10406):975–987. PMID 37573859 · doi:10.1016/S0140-6736(23)00683-9
- Linton SJ et al. The effect of the work environment on future sleep disturbances: a systematic review. Sleep Med Rev. 2015;23:10–19. PMID 25645126 · doi:10.1016/j.smrv.2014.10.010
- Fang Y et al. Day-to-day variability in sleep parameters and depression risk: a prospective cohort study of training physicians. NPJ Digit Med. 2021;4(1):28. PMID 33603132 · doi:10.1038/s41746-021-00400-z