Short answer: most sleep advice is a list of habits: dark room, no screens, consistent bedtime. Those habits are reasonable, but as a standalone intervention they have surprisingly thin evidence behind them, which is why following them often changes very little. What does have strong evidence is cognitive behavioural therapy for insomnia, which is first-line treatment for persistent sleep problems and produces clinically meaningful improvements. For everyone else, the highest-yield changes are a consistent wake time, daylight early, regular exercise, and being honest about caffeine, since a large dose within twelve hours of bed measurably disrupts sleep even when you cannot feel it.
If you have already tried the standard list and your sleep has not improved, the problem may not be that you failed to follow it properly. It may be that the list was never the strongest thing available.
This article separates the advice that has solid evidence behind it from the advice that mainly sounds sensible, and covers what to change first if you train and want your sleep to support that.
How much sleep do you actually need?
Seven or more hours per night is the consensus figure for healthy adults. The American Academy of Sleep Medicine and the Sleep Research Society reached that recommendation jointly, using a structured consensus process, as the amount needed to promote optimal health.
Seven is a floor rather than a target. Individual needs vary, and plenty of people function better nearer eight or nine, particularly during heavy training blocks.
Chasing a specific number can backfire, though. Anxiety about hitting a target is itself a common cause of poor sleep, and the useful measure is how you feel and function during the day rather than what a tracker reports.
If you are consistently below seven and consistently tired, the gap is worth closing. If you sleep six and a half hours and feel fine, that is a different conversation.
Why do the usual sleep tips underdeliver?
Because sleep hygiene advice was never designed to work on its own. It is a supporting component of a larger treatment, and it gets published as though it were the treatment.
Sleep hygiene means the set of behavioural and environmental habits around sleep: a dark, cool room, a consistent schedule, limiting caffeine and alcohol, and a wind-down routine. None of that is wrong.
But a review of evidence-based psychological treatments for insomnia by McCurry and colleagues found that sleep hygiene education did not meet the criteria for an effective standalone intervention. The treatments that did meet those criteria were sleep restriction and multicomponent cognitive behavioural therapy.
That review looked specifically at older adults, so the conclusion should not be stretched further than it goes. But it points at something the listicles never mention: these habits are the easiest part of the intervention to write about and the weakest part to rely on.
What works when sleep problems persist?
Cognitive behavioural therapy for insomnia, usually shortened to CBT-I, is the first-line treatment for chronic insomnia, and the evidence behind it is considerably stronger than for anything in the habits list.
A systematic review and meta-analysis by Trauer and colleagues pooled 20 randomised trials covering 1,162 participants. Time to fall asleep improved by around 19 minutes, time awake during the night by around 26 minutes, and sleep efficiency by nearly 10 percentage points. The improvements appeared to hold at later follow-up, and no adverse outcomes were reported.
CBT-I combines several components: stimulus control, which rebuilds the association between your bed and sleeping, sleep restriction, which temporarily limits time in bed to consolidate sleep, cognitive work on unhelpful beliefs about sleep, relaxation, and sleep hygiene as one part among several.
Note where sleep hygiene sits in that list. It is a component, not the programme. If your sleep problems have lasted months, ask your GP about CBT-I rather than working through another article of tips. It is delivered in person, through apps and online programmes, and it is not something to improvise from a blog post.
How does your body clock actually work?
Your sleep is driven by a central clock in your brain, which in turn sets a series of clocks throughout the rest of your body. Understanding that split explains why timing matters more than most of the habits on the standard list.
The central clock sits in the suprachiasmatic nucleus, a small region of the hypothalamus. It is set primarily by light arriving through your retina. Alongside it, almost every tissue runs its own clock: liver, muscle, fat tissue, gut. Those peripheral clocks only partly follow the central one.
How separately those two operate has been demonstrated cleanly. Wehrens and colleagues delayed meals by five hours for six days in ten healthy men. The melatonin and cortisol rhythms, which mark the central clock, did not shift at all. But the plasma glucose rhythm shifted by nearly six hours, and clock gene expression in fat tissue moved by about an hour.
The practical version: light sets your master clock, food sets part of your body clocks. Eat consistently out of step with when you sleep and the two drift apart, even when your sleep duration looks fine on paper.
How much does caffeine really matter?
More than most people think, and the dose and timing both matter. A randomised crossover trial by Gardiner and colleagues gave participants 100 mg or 400 mg of caffeine at 12, 8 or 4 hours before bedtime, measuring sleep with in-home polysomnography.
The 100 mg dose, roughly one coffee, produced no significant effect on sleep even four hours before bed. The 400 mg dose was a different story: it delayed sleep onset and altered sleep architecture when taken within twelve hours of bedtime, and increased sleep fragmentation within eight hours.
The most useful finding is the one about perception. Objective and subjective sleep quality diverged, which led the authors to conclude that people may have difficulty accurately perceiving how caffeine affects their sleep. "Coffee doesn't affect me" is precisely the judgement the data says you are not well placed to make.
Caffeine also shifts your body clock. A study by Burke and colleagues found that a dose equivalent to a double espresso three hours before bed delayed the circadian melatonin rhythm by about 40 minutes, roughly half the delay produced by three hours of bright light in the evening.
| Amount | Roughly equal to | Effect on sleep in the trial |
|---|---|---|
| 100 mg, 4 hours before bed | 1 coffee | No significant effect measured |
| 400 mg, 12 hours before bed | 4 coffees at once | Delayed sleep onset, altered sleep architecture |
| 400 mg, 8 hours before bed | 4 coffees at once | The above, plus more fragmented sleep |
| 400 mg, 4 hours before bed | 4 coffees at once | The above, plus a 34 percent drop in perceived sleep quality |
Based on Gardiner et al. (2025), 23 male participants with moderate habitual caffeine intake. Individual sensitivity varies considerably.
Does exercise improve sleep?
Yes, though the honest version of that answer is narrower than the headline. A systematic review and meta-analysis by Xie and colleagues covering 22 randomised trials found that regular exercise improved self-reported sleep quality, insomnia severity and daytime sleepiness.
The effects on objectively measured sleep were much smaller. Across the actigraphy data, few sleep parameters improved significantly other than sleep efficiency.
Two details from that analysis are practically useful. Conventional exercise and mind-body exercise such as yoga produced similar improvements, and shorter interventions of up to three months produced greater reductions in sleep disturbance than longer ones.
Feeling that you sleep better is not nothing. It is the outcome most people actually care about, and it is what sleep questionnaires are designed to capture. Just do not expect a training programme to transform your sleep architecture.
When should you train if you sleep badly?
Train at whatever time you will reliably show up, and adjust only if evening sessions clearly cost you sleep. The blanket advice to avoid training in the evening is stronger than the evidence supports for most people.
What does seem to matter is what happens afterwards. A hard session late at night raises core temperature, heart rate and arousal, and those take time to settle. The practical fix is a longer gap between finishing and lying down, not moving the session to a time you will skip.
Caffeinated pre-workout is the more common culprit. A scoop taken before a 19:00 session lands squarely inside the window where a large dose measurably disrupts sleep.
If you train in the evening and sleep badly, test the caffeine before you change the schedule. It is the easier variable to move and the more likely offender. If you are unsure how to structure your week, that is something we work out together in English-speaking personal training in Amsterdam.
What is your chronotype, and what is social jetlag?
Your chronotype is your natural preference for when you sleep and when you are active, and it varies considerably between people. It is largely governed by your circadian clock, with contributions from both genetic variation in clock genes and your environment.
Most people sit somewhere between the extremes. School and work schedules make little allowance for that spread, and late types absorb most of the cost.
Wittmann and colleagues named the result social jetlag: the gap between your biological time and your social time. Late chronotypes show the largest differences in sleep timing between work days and free days, accumulate a sleep debt across the working week, and compensate at the weekend. In their study of 501 participants, those discrepancies were associated with sleep quality and wellbeing, most strongly in people up to around 25.
Two things follow. You cannot train your chronotype away, but you can narrow the gap: morning light and a more consistent wake time pull a late type slightly earlier. And if you are a late type with an early job, treat your tiredness as a timing problem rather than a discipline problem.
One caveat. Chronotype systems that sort people into four or five named animal types, of the sort found in popular books, are commercial simplifications. In research, chronotype is measured on a continuous scale, for example with the Munich ChronoType Questionnaire.
What does poor sleep do to your training?
It changes what your body does with the work you have done, and the clearest demonstration of that is in body composition. In a controlled crossover study by Nedeltcheva and colleagues, participants held the same calorie deficit for two weeks with either 8.5 or 5.5 hours of sleep opportunity.
On the short nights, the proportion of weight lost as fat dropped by more than half, while loss of lean mass rose. Participants also reported more hunger.
Training quality suffers too, in ways that are harder to measure but easy to recognise: heavier warm-ups, lower motivation, worse technique on the last sets, and a slower return to baseline between sessions.
None of this makes sleep more important than training or nutrition. It makes sleep the variable that determines how much of the other two you get to keep. Our articles on lowering body fat and on progressive overload cover the other two sides of that.
What about screens, alcohol and naps?
These three come up constantly and the evidence behind them is more mixed than their prominence suggests, so it is worth being precise about what is known.
Alcohol is the most clear-cut. It shortens the time to fall asleep and then degrades the second half of the night, which is why a nightcap produces sleep that feels unrefreshing.
Screens have two separate effects that often get conflated. The light itself does something measurable: Chang and colleagues had participants read in the evening from either a light-emitting device or a printed book, and after the device they took longer to fall asleep, had suppressed melatonin, a later circadian clock and reduced alertness the next morning.
But what you do on the device counts as well. Work email, arguments and infinite scroll are arousing, and arousal keeps you awake regardless of what colour the light is. A blue light filter partly addresses the first problem and does nothing at all about the second.
Naps are neutral to helpful for most people if they are short and early. A twenty minute nap before mid-afternoon rarely affects night-time sleep. Long or late naps reduce the sleep pressure you need in the evening, which matters most if you already struggle to fall asleep.
What should you change first?
Most people start with the weakest intervention and stop before reaching the strong ones. This order reflects effect size rather than ease.
- Fix your wake time before your bedtime. A consistent wake time anchors your body clock; bedtime will follow it. Bedtime alone does not work in reverse.
- Get daylight within an hour of waking. Ten to twenty minutes outside does more for your clock than any indoor lighting change in the evening.
- Audit caffeine honestly. Count everything, including pre-workout and afternoon tea, and shift the large doses earlier before cutting them out.
- Train regularly, at a time you will keep. Consistency of the habit beats optimisation of the timing.
- Then adjust the environment. Dark, cool, quiet. Worth doing, and the point at which most articles begin.
- If nothing has changed after several weeks, ask your GP about CBT-I. This is the step with the strongest evidence and the one people reach last.
Change one thing at a time and give it two weeks. Changing five things at once tells you nothing about which one worked.
When is it more than habits?
Some sleep problems are not habit problems, and no amount of routine adjustment will resolve them. Speak to a GP if any of the following apply.
Loud snoring with pauses in breathing, gasping or choking during the night, or waking unrefreshed despite adequate time in bed, can indicate sleep apnoea, which needs assessment rather than better habits. Persistent difficulty sleeping for more than three months is chronic insomnia and has a specific evidence-based treatment.
Sleep problems also travel with low mood, anxiety and chronic stress in both directions. If your sleep changed alongside how you have been feeling generally, that is worth raising with a doctor rather than treating as a separate technical problem.
Persistent daytime sleepiness that affects driving or work warrants prompt medical advice regardless of cause.
Key points
- Seven or more hours is the consensus recommendation for healthy adults, as a floor rather than a target.
- Sleep hygiene advice did not meet the criteria for an effective standalone intervention in the treatment literature.
- Light sets your central clock and food sets part of your body clocks. The two can drift out of step independently.
- One coffee four hours before bed measured as harmless; four coffees at once disrupted sleep up to twelve hours out.
- People cannot reliably perceive how caffeine affects their own sleep, so self-assessment is not a good guide here.
Frequently asked questions
What is the 3-3-3 rule for sleep?
The 3-3-3 rule is a popularised routine, usually given as no alcohol three hours before bed, no food three hours before bed, and no screens or work three hours before bed, though several versions circulate. It packages a few reasonable ideas into something memorable rather than reflecting a specific study, and the three hour figure is not derived from trial data. Treat it as a structure some people find easy to follow, and note that it omits caffeine, which has better evidence behind it than any of the three.
Why am I tired but unable to fall asleep?
Feeling exhausted while lying awake usually means your sleep pressure and your body clock are out of step, or that your arousal level is too high to let sleep start. Late caffeine, a late or irregular wake time, evening screen use that is mentally stimulating, and stress all contribute. The most common single cause is an inconsistent wake time, because sleep pressure builds from when you got up rather than from how tired you feel.
Do sleep trackers actually help?
Sleep trackers are reasonably good at estimating how long you slept and considerably less good at classifying sleep stages, so the deep sleep figure on your app is an estimate rather than a measurement. They can be useful for spotting trends across weeks. They can also make things worse for people who become anxious about the scores, which is common enough to have a name in sleep medicine. If checking your sleep score affects your mood in the morning, stop using it.
Does melatonin work?
Melatonin is a signal that tells your body it is night, so it works best for problems of timing such as jet lag or a shifted body clock, and considerably less well as a general sleeping aid. Doses sold over the counter are often far higher than the amounts studied for circadian effects. Because it interacts with some medications and is regulated differently across countries, check with a pharmacist or your doctor before using it rather than treating it as a harmless supplement.
Is it bad to sleep in at the weekend?
Sleeping in an extra hour at the weekend is unlikely to cause problems, but a three hour lie-in effectively shifts your body clock, which is why Sunday nights are so often difficult. That pattern has a name, social jetlag, and it is most common in late chronotypes working early schedules. If you consistently need a long weekend lie-in, that is a sign you are running a sleep debt during the week rather than a habit worth optimising. Keeping your wake time within about an hour of your weekday time is the pragmatic compromise.
Want your training built around the sleep and recovery you actually get? Book a free intake session.
Sources
Scientific claims in this article were verified via PubMed.
- Watson NF, Badr MS, Belenky G, et al. Recommended amount of sleep for a healthy adult: a joint consensus statement of the American Academy of Sleep Medicine and Sleep Research Society. Sleep. 2015;38(6):843-844. DOI
- McCurry SM, Logsdon RG, Teri L, Vitiello MV. Evidence-based psychological treatments for insomnia in older adults. Psychol Aging. 2007;22(1):18-27. DOI
- Trauer JM, Qian MY, Doyle JS, Rajaratnam SMW, Cunnington D. Cognitive behavioral therapy for chronic insomnia: a systematic review and meta-analysis. Ann Intern Med. 2015;163(3):191-204. DOI
- Gardiner CL, Weakley J, Burke LM, et al. Dose and timing effects of caffeine on subsequent sleep: a randomized clinical crossover trial. Sleep. 2025;48(4):zsae230. DOI
- Burke TM, Markwald RR, McHill AW, et al. Effects of caffeine on the human circadian clock in vivo and in vitro. Sci Transl Med. 2015;7(305):305ra146. DOI
- Xie Y, Liu S, Chen XJ, Yu HH, Yang Y, Wang W. Effects of exercise on sleep quality and insomnia in adults: a systematic review and meta-analysis of randomized controlled trials. Front Psychiatry. 2021;12:664499. DOI
- Nedeltcheva AV, Kilkus JM, Imperial J, Schoeller DA, Penev PD. Insufficient sleep undermines dietary efforts to reduce adiposity. Ann Intern Med. 2010;153(7):435-441. DOI
- Chang AM, Aeschbach D, Duffy JF, Czeisler CA. Evening use of light-emitting eReaders negatively affects sleep, circadian timing, and next-morning alertness. Proc Natl Acad Sci U S A. 2015;112(4):1232-1237. DOI
- Wittmann M, Dinich J, Merrow M, Roenneberg T. Social jetlag: misalignment of biological and social time. Chronobiol Int. 2006;23(1-2):497-509. DOI
- Wehrens SMT, Christou S, Isherwood C, et al. Meal timing regulates the human circadian system. Curr Biol. 2017;27(12):1768-1775. DOI
