The harder you try, the less you sleep
Sleep and mental health. What really helps, what cognitive behavioural therapy for insomnia (CBT-I) is, and what to do with thoughts at 3 a.m. A webinar guide with infographics, polls, exercises and a printable night card.
Before you start. This guide is educational. It is not a diagnosis or a treatment plan. If something here matches your situation closely, take it as a reason to talk to someone, not a reason to self-treat. Where to find support.In short
This guide is based on the webinar “The harder you try, the less you sleep”, which I deliver for teams in organisations. It includes four self-reflection polls (only you see your answers), an interactive sleep-pressure chart, a short “is it time for CBT-I?” checklist, a defusion exercise, guided breathing and a printable night card. Choose a part in the menu, or start with the tiles above.
There is also a longer Polish guide on sleep and insomnia and a sleep self-check on this site (both in Polish).
Sleep and mental health
Three numbers
Many more have occasional symptoms. If you sleep badly, you are not unusual, and it is not a sign of weakness. Pooled across 21 studies that followed people forward in time: the sleep problem was measured first, the depression came after. Only the sleep was treated (an online CBT-I programme). Mental health symptoms improved too, and the sleep improvement statistically explained the change. The first two numbers say that sleep and mood travel together. The third shows the direction: only the sleep was changed, and the mind followed. Confidence: high for all three. The OASIS trial was conducted with university students, so the size of the effect may differ elsewhere, for example in a workplace population. Sources.
The relationship runs both ways
Most of us think of it in one direction: I feel stressed, so I sleep badly. That is true, but it is only half of the picture. The other half changes what you can do about it.
The sleep and mood loop
Weeks of poor sleep do not only follow a hard period. They also raise the chance that a hard period begins. Why is this good news? Because sleep is the part of the loop you can work on directly. We often cannot remove the stress at work or at home. But we can change how the night is organised. And when the night improves, mood very often follows.
Sleep problems often appear before a low period becomes obvious, and they often remain after the mood has lifted. Untreated insomnia is one of the better known risk factors for relapse. So in therapy we treat sleep as a target in its own right, not as a side effect that will sort itself out. This does not mean depression is “caused” by bad sleep. Sleep is one contributing factor among several, and it happens to be a workable one. Treating sleep helps; it does not replace treatment for depression.What we see in the clinic
One short night
The brain’s emotional alarm responds more strongly to unpleasant events, and the brake from the prefrontal cortex is weaker. Neutral faces and neutral messages are read as less friendly. The same meeting feels worse than it did last week. The meeting did not change. The reader did. Patience, humour and willingness to cooperate drop. This is why one bad night is rarely a private matter at work. Amygdala reactivity to unpleasant images A small laboratory study (n = 26), after about 35 hours without sleep, which is more than an ordinary bad night. The direction is well supported. Treat the number as an illustration, not a measurement of your Tuesday.Stronger reactions
A more negative filter
Less social bandwidth
None of this means you are fragile. It means the next day is being run on a system with less regulation available. It is worth knowing that before you make a decision or send a difficult message. After a bad night, the difficult email can wait until the afternoon.
If a colleague is unusually short with you, you are not necessarily seeing their character. You may be seeing their night. That is not an excuse for how people behave. But it is a useful second hypothesis before you take it personally.For anyone who works with people
What helps you sleep
Two systems, one condition
Sleep is not a decision. It is the result of two systems and one condition. Once you see that, a lot of common advice starts to make sense. And some of it stops making sense.
It builds up for every hour you are awake and discharges when you sleep. A long nap or a late lie-in spends it early, so there is less left for the night. Protect it: no long naps, one fixed wake-up time. It decides when sleep is biologically easy. It is set mainly by light and by a regular rhythm, not by how tired you feel. Set it: daylight in the morning, dim light late. The condition. If the body and mind are in alert mode, sleep does not arrive, even when pressure is high and the clock is right. This is where most insomnia actually lives. See how much sleep pressure is left for the night An illustrative sketch based on Borbély’s two-process model (time constants from the literature). The numbers are an illustration, not a measurement. You cannot produce sleep. You can only raise sleep pressure, keep the clock regular and lower activation. Then sleep happens by itself. Chronotype. Body clocks differ between people, partly genetically. For sleep, regularity matters more than the particular hour. Caffeine blocks the signal of sleep pressure (adenosine). The pressure is still there; you just stop feeling it. How long caffeine lasts varies enormously between people; more in the questions at the end.1. Sleep pressure
2. The body clock
3. Activation
Night owl or early bird? And caffeine?
What you do when sleep will not come
Sleep hygiene, honestly
Useful ground rules, not a treatment. On its own, sleep hygiene is the weakest of the recognised insomnia interventions.
The familiar rules, ranked by what they actually do The American Academy of Sleep Medicine (AASM, 2021) makes a conditional recommendation against sleep hygiene as a single-component therapy for chronic insomnia. Not because it is harmful. Because it is not enough, and people conclude that nothing works. Sleep hygiene is the floor of the building, not the building.
What the guidelines say
There is a second, subtler problem with the list. For someone already anxious about sleep, a long list of rules becomes six more ways to fail before midnight. Some people lie awake worrying that the bedroom is two degrees too warm. That worry costs them far more sleep than the temperature ever did.
The effort trap
Every other problem gets better when you try harder. Sleep gets worse. Effort is activation, and activation blocks sleep. This is why the title of this guide is not a slogan. The goal is not to try better. The goal is to stop making the night a task. Lying there deciding to fall asleep now is effort. And effort is activation. “Five hours left” turns the night into a countdown you are losing. Each new rule is one more exam you can fail tonight. How a few bad nights become months Stress triggers a few bad nights. That is normal, and it would usually pass. You start managing sleep. Earlier to bed to catch up. The evening out cancelled. The app. The clock. Sleep has become a project with a performance target. The bad nights continue after the stress is gone. The insomnia is now maintained by the solution, not by the original cause. Sleep medicine describes this with the “3P” model (predisposing, precipitating and perpetuating factors) and the concept of sleep effort. So the rest of this guide does not give you better techniques for trying. It takes the night out of the category of things you perform.Trying to sleep
Checking the clock
Perfect sleep rules
CBT-I
What CBT-I delivers
CBT-I is a short, structured programme, usually four to eight sessions, delivered by a psychologist, psychotherapist or doctor. Well-tested digital versions exist too. Here is what pooled data from 19 randomised trials show.
Change versus control (meta-analysis) Sleep efficiency is the share of your time in bed that you actually spend asleep. These are averages: some people do much better, some less well. CBT-I does not add hours straight away. It consolidates the hours you have. A solid night feels dramatically better than the same number of fragmented hours. Total sleep often grows later, once the system settles. Benefits are generally maintained after treatment ends. That is the main difference from sleeping medication, where the benefit tends to stop when the medication stops. American (AASM 2021) and European (2023) guidelines give multicomponent CBT-I their strongest recommendation: it should be offered first, before medication.Solid first, longer later
The gains last
The strongest rating
The three working parts
Rebuilds the link between bed and sleep, which insomnia has broken. safe to start on your own Shortens time in bed so that sleep becomes deep and continuous again. with guidance only Takes apart the catastrophic prediction that drives night-time arousal. safe to start on your ownStimulus control
Sleep consolidation
Cognitive work
Notice what is missing from this list: there is no relaxation trick to make you fall asleep faster. Every component removes something that is keeping you awake. It is subtraction, not addition. That is also why it keeps working after the programme ends: you have not acquired a technique you must keep performing. You have removed what was in the way.
Stimulus control
After weeks of bad nights you get into bed and your body already knows what is coming: lying awake, frustration, the clock. The bed stops being a cue for sleep and becomes a cue for being awake and annoyed. The body learns this the way it learns any association. Stimulus control unlearns it. It feels counter-intuitive, because the instruction is to spend less time in bed, not more.
The five instructions Go to bed only when you are sleepy, not merely tired. Tired is heavy. Sleepy is when your eyes actually close. The bed is for sleep and sex only. No work, no scrolling, no series. Awake and frustrated for roughly 20 minutes? Get up. Another room, dim light, something quiet and boring. Do not check the clock to measure it; estimate. Return only when sleepy again. Repeat as many times as the night needs. If it is three times, it is three times. Get up at the same time every morning, whatever the night was like. Expect two or three harder nights before it settles. That is part of the method, not a sign it is failing. “I will be even more tired if I get up.” For two or three nights, yes. After that, almost always less. You are trading a short cost for the association being rebuilt. “I share a bed. I will wake my partner.” Move quietly, have a dressing gown and a lamp ready in the other room, and tell them in advance what you are doing and why. Most partners prefer this to someone turning over for two hours beside them. “How do I know it has been 20 minutes without looking?” You estimate. The point is not accuracy; the point is that you stop waiting indefinitely. And please do not set a timer. A timer is one more measurement of your performance. Live in one room? Sit up in a chair in that room, with a different light on. The change of place and posture is what matters. A book? A paper book in low light is fine. The phone is not.
Three objections I always hear
Sleep consolidation
Someone sleeping five hours but lying in bed for nine spends four hours training themselves to be awake there. So time in bed is reduced towards the actual sleep time, then extended again step by step as sleep becomes solid.
Sketch: time in bed and sleep Deliberately without hours: the sleep window is set and adjusted week by week by a clinician, against your sleep diary.
It works very well. It is the fastest-acting part of CBT-I, and the one people most often get wrong on their own: they set the window far too short, feel terrible after four days, stop, and conclude that CBT-I “does not work for them”. In fact they used the strongest component without the part that makes it safe. For a while it increases daytime sleepiness, which is exactly why it needs someone watching the numbers with you.
Treat this section as something to ask for by name, not as a plan to start tonight. Say to your doctor or therapist: “I would like to try sleep restriction” (or sleep consolidation). They will know exactly what you mean. For two weeks, write down each morning, on paper: when you went to bed, roughly how long it took to fall asleep, the times you woke, when you got up. Nothing clever. It is the first thing any clinician will ask for, and it will save you a session. A ready template is on the printable card. Do not use a sleep-tracking watch for this. It measures sleep poorly and turns it into a nightly score, which raises sleep effort. There is a name for that pattern: orthosomnia.Do not self-prescribe if
What you can do on your own now: a sleep diary
The catastrophic prediction
The catastrophic prediction loop ↺ the next night the thought returns stronger
The body does not treat that thought as an idea. It treats it as an announcement of threat. You are not arguing the thought away, and you are not replacing it with positive thinking. You are checking the prediction against the record and counting what it costs.
Asked honestly, most people report that the day after a bad night was unpleasant and entirely survivable. The disaster is predicted far more often than it arrives. Believing the prediction raises the stakes of the night, and raised stakes are activation. The thought is not just inaccurate. It is expensive. If someone told you this at 11 p.m., would you answer: “You are right, tomorrow is ruined”? We are generous with other people and severe with ourselves at two in the morning. When to do this? In the evening, or the next day, in daylight, when the mind is cooler. Not in bed at 3 a.m. At 3 a.m. you use the practices from part four. Agreed. Some days genuinely depend on being rested. Even then, believing the prediction at midnight only reduces the sleep you still might get.What actually happened last time?
What does believing it cost me right now?
What would I say to a colleague?
“But sometimes tomorrow really does matter”
When to ask for help
Talk to a doctor
Ask for CBT-I
Do not wait
Many people go years without asking for help with insomnia, even though a short treatment with strong evidence exists. Asking for help with sleep is not an admission that you are not coping. It is like seeing a physiotherapist about a knee: a specific problem with a specific treatment. If your employer offers an employee assistance programme, sleep is a perfectly good reason to use it. You do not need to be in crisis.
The mind at 3 a.m.
What comes at night
Emptying the mind is not possible, and trying to do it is more effort, which means more activation. The goal is different. But first, one question.
Why night-time thinking convinces
Rumination uses the same words as planning, so the mind treats it as useful work. But it circles instead of concluding. At 3 a.m. there is no task, no colleague, no daylight, nothing to contradict the story. A worry that would look ordinary at noon fills the whole room. The mind raises a problem and finds every route to it closed until morning. That gap between urgency and impossibility is itself activating.It feels like problem-solving
Nothing else is competing
You cannot act on it
Planning
Rumination
So the problem is not the content of the thoughts. At 3 a.m. the content is almost never solvable anyway. The problem is the relationship: you are treating the mind as a council that must be answered.
Before bed, write briefly: what matters tomorrow, the first action, and anything unresolved. The effect is not organisational. The mind gets evidence that it has been heard, and a mind that has been heard argues less at three. From research: in a sleep-laboratory study with polysomnography, people who spent 5 minutes at bedtime writing a specific to-do list for the coming days fell asleep faster than people who wrote about tasks already completed; the more specific the list, the faster they fell asleep (Scullin et al., 2018; healthy young adults, n = 57).In the evening: three lines
And during the night? A keyword on a notepad by the bed is fine. Opening a notebook and working is not. The rule: capture, do not process. Your best ideas come at night? It happens. They rarely survive daylight, so the same rule applies.
Defusion
Instead of “I have ruined the project”, say inwardly: “I am having the thought that I have ruined the project.” It sounds like a word game. It is not. The content stays the same; your distance from it changes, and so does the alarm it sets off. Same content, smaller alarm. “Noted. You get fifteen minutes tomorrow at nine.” And keep the appointment, at least sometimes. If you never keep it, the mind stops believing you. Suppression rebounds. An appointment is usually accepted. The breath. The weight of the body on the mattress. This is not a sleep trick. If you use it as a trick, it becomes effort. It is simply somewhere other than the story to rest attention. The mind will wander again. Returning is the practice, not staying. What you type stays in this browser tab and disappears when you close it. A thought is an event in the mind, not an instruction and not a verdict. You can let it be there and do nothing about it. At 3 a.m. that is usually the wisest available option. Where this comes from: defusion comes from acceptance and commitment therapy (ACT); the attentional element comes from mindfulness practice. The evidence base for insomnia specifically is smaller than for classical CBT-I, though the results are promising (review: Salari et al., 2020; randomised trial of MBTI: Ong et al., 2014). Very common, and usually a sign it was being used as a sleep technique. The expectation is what activates, not the practice. And if you cannot empty your mind: good. Nobody can. That is not the task here.Name the frame
Postpone, do not suppress
Return to the body
“I tried meditation and it made me more awake”
Everything in this guide has the same shape: we stop fighting the night and remove what is in the way. Stimulus control removes the wrong association. Dropping the catastrophic prediction removes the alarm. Defusion removes the obligation to answer the mind. None of it is a technique for producing sleep. That is the point.
Four steps and breathing
“Thinking is happening.” Not good, not bad. Just noticed. Say it plainly: “I am not trying to fall asleep. I am resting until morning.” Breathe in normally, a little longer out. Ten breaths. No counting system to get right. Still awake and frustrated? Get up, dim light, something quiet. Return when sleepy.
Notice and name
Drop the goal
Lengthen the out-breath
Decide at about 20 minutes
Step two is the one that does the work. Lying in bed resting, awake, is a perfectly acceptable outcome for a night. And here is the paradox of this guide one more time: that resting state is also the state from which sleep most often arrives. Not because you used it as a trick. Because you stopped trying.
Sit how you are. You do not need to close your eyes. Breathe in normally; let the out-breath be a little longer. If the mind goes somewhere, that is what minds do: come back. Practise during the day, so you do not reach for the phone at night. The point is for the rhythm to become familiar. If you are still awake afterwards, that is not failure. Nothing here is a switch. It is a way of spending the night that does not make the next one worse. Over a week or two, that is what changes the pattern.Practice: ten breaths with a longer out-breath
Questions and work
Common questions
How much and when The usual recommended range for adults is seven to nine hours, but the range between people is wide. How you function during the day matters more than the number. “Eight hours for everyone” is a simplification, and a rigid target easily becomes one more performance goal. Hirshkowitz et al., 2015 (National Sleep Foundation). Honestly, one extra hour is fine for most people. Three hours is a small jet lag every week: the clock shifts, and falling asleep on Sunday evening gets harder. The research literature calls it social jetlag. Wittmann et al., 2006. Short (around 20–30 minutes) and early in the afternoon. If you have trouble sleeping at night, better to skip them: every nap spends some of the sleep pressure you need in the evening (see the chart). Common, and usually not sinister. It becomes a problem when you start expecting it, because the expectation itself activates. If it comes with loud snoring or gasping for breath, talk to a doctor. Substances and medication Weak for ordinary insomnia: in a meta-analysis it shortened time to fall asleep by only a few minutes on average. AASM guidelines do not recommend it for chronic insomnia. It is more useful for body-clock problems such as jet lag. Ask a doctor before using it. Ferracioli-Oda et al., 2013; Sateia et al., 2017 (AASM). Caffeine blocks the signal of sleep pressure. The pressure is still there; you just stop feeling it. Its half-life varies enormously between people. In a study with objective sleep measurement, a large dose (400 mg) taken even six hours before bed reduced sleep by more than an hour. A safe rule: coffee before early afternoon, and if you sleep badly, see what changes when you move it earlier. Drake et al., 2013. Alcohol speeds up falling asleep, but in the second half of the night sleep becomes lighter and more broken. When alcohol becomes the way through the night, it is a signal to talk to a professional (see when to ask for help). Ebrahim et al., 2013. They have a place: short term and with a doctor. Never stop them on your own. CBT-I is recommended as the first-line treatment, and it also helps if you and your doctor want to reduce medication gradually. Technology They measure sleep poorly, especially time awake. A nightly “sleep score” easily turns sleep into an exam and raises sleep effort. This pattern has been described as orthosomnia. If you want to measure something, a paper diary filled in each morning is better. Baron et al., 2017. Digital CBT-I programmes have real evidence. Generic sleep-sound apps do not. The difference is whether the app changes your behaviour or just plays something. Zachariae et al., 2016 (meta-analysis). The light effect exists, but it is small compared with content and timing. In a laboratory study, reading on a light-emitting e-reader for several evenings delayed falling asleep by around ten minutes. A work email at 11 p.m. does more harm than the colour of the screen. Blue-light glasses: if they help you, fine, but they are not the main lever. Chang et al., 2015. Special situations The rules change: the fixed wake-up time rule does not transfer directly to a shift rota. This is worth planning individually, in a consultation. The night is not in your control. Shift the goal from fixing sleep to protecting the wake-up time where you can, sharing the night-time load, and not adding rules you cannot keep. A real and very common cause of night waking. Medical conversation first, behavioural work alongside it. CBT-I improved sleep in perimenopausal and postmenopausal women with hot flushes too. McCurry et al., 2016. Frequent, distressing nightmares have their own effective treatments, such as imagery rehearsal therapy. They are worth asking about in a consultation, especially if they started after a difficult event. Morgenthaler et al., 2018 (AASM). Exercise and food Overall, yes: regular physical activity is linked with better sleep. Intense exercise very late suits some people and not others. It is individual. Kredlow et al., 2015 (meta-analysis). A very large late meal disturbs sleep. Going to bed hungry does too. Nothing dramatic either way.How many hours of sleep do I need?
Do weekend lie-ins help?
Naps: yes or no?
I wake at the same hour every night. Is something wrong?
Melatonin?
Caffeine: until when?
A glass of wine to help me sleep?
Sleeping pills
Sleep trackers and apps
Can an app help?
Screens and blue light
I work shifts
I have small children
Menopause
Nightmares
Does exercise help sleep?
Evening meals
Sleep and teams
Sleep is not only a private matter. Research on leadership shows that a leader’s poor night carries over into how they treat their team the next day, and that late-night smartphone use for work lowers next-day engagement.
When someone is unusually short with you, consider their night before you judge their character. A message sent at 11 p.m. invites work into sleep time. Scheduling it for the morning costs nothing. Where possible, hold important conversations and decisions after a good night, not after a night flight or an on-call shift. Say openly that sleep is a good enough reason to use the support programme. It lowers the threshold. Barnes et al., 2015; Lanaj, Johnson and Barnes, 2014.The second hypothesis
Late messages
Decisions after a bad night
Support is for everyone
To finish
One change for two weeks
After a guide like this, it is tempting to change everything: new bedtime, no coffee, no phone, an app, breathing. That usually lasts about four days and then collapses, and the conclusion is: “none of this works for me”. But reforming your entire night at once is just a larger version of trying harder. One change, two weeks, then decide whether to add anything. Write it on paper somewhere you will see it tonight. Not in your phone at bedtime. Cannot choose? Take the wake-up time. Everything else builds on it.Which one thing will you try this week?
My one change for two weeks
Five things to take away
If you change one thing after reading this, make it the wake-up time. Everything else needs it to build on.
Night card
One page with what matters most, and a second with a two-week sleep diary. Keep it by the bed, or in the room you go to at night.
The harder you try, the less you sleep · Jarosław Chybicki · chybicki.pl/en/materials/sleep Doctor: snoring with pauses in breathing, restless legs, falling asleep in the day without meaning to, coming off medication. CBT-I: 3+ nights a week, for over 3 months, affecting your days. Do not wait: low mood or anxiety through the day, loss of interest, alcohol or medication to get through the night, worrying thoughts. Support: your employer’s assistance programme (EAP), or a psychologist or psychotherapist. Urgent, in Poland: 116 123 · 800 70 2222 (24/7) · 112. Elsewhere: your local emergency number. Educational material, not a diagnosis. Jarosław Chybicki · chybicki.pl Fill it in each morning, on paper, roughly. Do not look at the clock during the night. Bring the diary to a consultation. Do not analyse it every day: it is not a score to improve.Night card
Stimulus control: 5 rules
At 3 a.m.: 4 steps
When a thought arrives
In daylight, not at night
My one change for 2 weeks
In the evening: 3 lines
When to ask for help
Sleep diary: 2 weeks
Day Date Went to bed at Fell asleep after about (min) Woke up: how often, how long Got up at Nap / alcohol / caffeine after 2 p.m. How I felt in the day (1–5) 1 2 3 4 5 6 7 8 9 10 11 12 13 14
Sources
Everything in this guide came from somewhere. Where the evidence is strong, it says so. Where it is thinner, it says that too. The sleep field is full of confident claims that do not hold up: eight hours for everyone, blue light as the main villain, apps promising to fix your brain waves. If you are going to change something about your nights, you are entitled to know how good the evidence is.
Claim Source Confidence Around 10% of adults have chronic insomnia disorder; CBT-I is first-line Edinger et al., AASM clinical practice guideline, J Clin Sleep Med, 2021 high Insomnia: roughly 2.6× the odds of later depression Baglioni et al., meta-analysis of 21 longitudinal studies, J Affect Disord, 2011 (OR 2.60, 95% CI 1.98–3.42) high Treating sleep alone improves mental health symptoms Freeman et al., OASIS randomised controlled trial, n = 3,755, Lancet Psychiatry, 2017 high student sample CBT-I: −19 min to sleep, −26 min awake, about +10 points efficiency Trauer et al., meta-analysis of 19 randomised trials, Ann Intern Med, 2015 high Sleep hygiene is not effective as a stand-alone treatment Edinger et al., AASM guideline, 2021 (conditional recommendation against) high Over 60% greater amygdala reactivity after a sleepless night Yoo, Gujar, Hu, Jolesz & Walker, Curr Biol, 2007 (n = 26) moderate small sample Defusion and acceptance-based work with night-time thoughts ACT literature; a smaller evidence base for insomnia than classical CBT-I (Salari et al., 2020) moderate Full reference list (including the extended sections)
Where to find support
If your employer has an employee assistance programme (EAP), you can talk to a psychologist there confidentially and free of charge. Sleep is a good enough reason. You can also see a psychologist, psychotherapist or doctor privately.
When sleep has not returned to normal for weeks, or comes with tension, anxiety or low mood. When the difficulties have lasted more than three months. Ask for it by name and bring your sleep diary. For snoring with pauses in breathing, restless legs, falling asleep during the day without meaning to, or coming off medication. If you are having thoughts of harming yourself, do not stay alone with them. In Poland: Outside Poland, call your local emergency number.A consultation with a psychologist
CBT-I
A doctor
When you need help now
Sleep is not something you do.
It is something that happens when you stop doing.