Insomnia is often treated as a problem of sleep manners: less coffee, fewer screens, a cooler room, perhaps a stricter bedtime. Those things can help the edges. Chronic insomnia is usually more stubborn. When sleep becomes something the brain has learnt to resist, the stronger evidence points to cognitive behavioural therapy for insomnia, or CBT-I, not sleep hygiene alone.
Why insomnia becomes self-reinforcing
Most people know the first bad night. The more interesting biology begins after the tenth, twentieth, or hundredth. The bed starts to mean effort. Darkness stops being a cue for sleep and becomes a cue for monitoring: how long have I been awake, how much will tomorrow suffer, why is everyone else able to do this simple thing?
That loop matters because insomnia is not only a shortage of sleep. It is a state of conditioned arousal. The nervous system can be tired and alert at the same time. Heart rate may not be dramatic, and cortisol may not be visibly abnormal, but the sleeper is running a quiet prediction model: bed equals failure. CBT-I tries to retrain that model.
The science is clearer for chronic insomnia than for many popular sleep routines. The 2021 American Academy of Sleep Medicine guideline gave multi-component CBT-I a strong recommendation for adults with chronic insomnia disorder. That does not mean it works for everyone, or that it is pleasant in the first week. It means the balance of controlled trials favours it over advice that merely tidies up sleep habits.
Sleep hygiene is useful, but it is not treatment
Sleep hygiene is the familiar list: keep a regular wake time, reduce late caffeine, avoid heavy alcohol close to bed, make the room dark, quiet, and cool. These are reasonable foundations. They also tend to disappoint people with long-standing insomnia, because they do not directly change the learnt association between bed and wakefulness.
The distinction is not semantic. NHS guidance says insomnia usually improves with changes to sleeping habits, but when those changes are not enough after months of trouble, GPs may offer cognitive behavioural therapy for insomnia rather than long-term sleeping pills. In other words, habits are the floor. They are rarely the whole house.
This is why a person can do everything conventionally right and still be awake at 2am. They can avoid coffee after lunch, buy blackout curtains, and leave the phone outside the room. If the bed itself has become a place of threat and calculation, the problem is no longer just the environment. It is the brain’s expectation of what happens there.
What CBT-I actually changes
CBT-I is not general talk therapy with a sleep label attached. It is a structured treatment, usually delivered over several sessions, that uses behavioural and cognitive tools to rebuild sleep pressure, timing, and confidence.
Stimulus control is one piece. The bed is reserved for sleep and sex, not for long negotiations with the ceiling. If sleep does not come, the person leaves the bed and returns only when sleepy. The point is not punishment. It is reconditioning. The brain relearns that bed is where sleep happens, not where wakefulness is practised.
Sleep restriction therapy is another piece, and it is often the most misunderstood. Time in bed is temporarily limited to match the person’s actual sleep time, then gradually expanded as sleep becomes more consolidated. Someone spending eight hours in bed but sleeping five may initially be prescribed a shorter sleep window. That sounds counterintuitive, but the goal is to increase sleep drive and reduce fragmented, frustrated time in bed.
Cognitive work addresses the catastrophic arithmetic that keeps insomnia alive. A thought such as “I will be useless tomorrow unless I sleep now” is not treated as silly. It is tested. Many people function badly after poor sleep, but not always as badly as the 2am mind predicts. Loosening that certainty can lower arousal enough for sleep to regain some automaticity.
The evidence is strongest for chronic insomnia
CBT-I is best understood as a treatment for chronic insomnia disorder, not as a universal sleep upgrade. Chronic insomnia usually means difficulty falling asleep, staying asleep, or waking too early at least several nights a week, with daytime consequences, lasting three months or longer. That duration matters. A bad week after grief, travel, illness, or a work crisis is not the same clinical problem.
In older adults, the evidence remains encouraging. A 2022 systematic review and meta-analysis of CBT-I in older people concluded that it may be a safe and effective approach for improving insomnia symptoms. That is important because ageing changes sleep architecture. Deep sleep tends to decline, medical conditions become more common, and medication side effects become more consequential.
CBT-I does not reverse the normal ageing of sleep. It does not promise the deep sleep profile of a twenty-five-year-old. The more modest and useful aim is better sleep efficiency: less time awake in bed, more predictable sleep timing, and less fear around the night.
Why pills are not the same answer
There are situations where medication has a role. Short-term pharmacological treatment may be reasonable during acute distress, whilst CBT-I is unavailable, or when symptoms remain severe despite good behavioural treatment. The concern is not that all sleep medicines are bad. It is that they solve a different problem.
A sedative can increase the probability of sleep on a given night. It does not necessarily retrain the brain’s relationship with bed, time, and threat. Some medicines also bring morning sleepiness, falls risk, tolerance, interactions, or rebound insomnia when stopped. Those trade-offs become more important with age, alcohol use, frailty, and polypharmacy.
The most recent nuance comes from combination treatment. A 2026 American Academy of Sleep Medicine guideline on CBT-I plus medication suggests combination treatment over medication alone, but suggests against adding medication to CBT-I when CBT-I alone is the comparator. The practical reading is conservative: CBT-I sits closer to first-line care; medication may be an adjunct for selected cases, not the foundation for most chronic insomnia.
When CBT-I needs medical screening first
Not every poor sleeper should jump straight into sleep restriction. Some symptoms point to another sleep disorder or medical driver. Loud snoring, witnessed pauses in breathing, morning headaches, resistant hypertension, and marked daytime sleepiness can suggest obstructive sleep apnoea. Restless legs, chronic pain, depression, thyroid disease, menopause symptoms, reflux, and medication effects can all fragment sleep.
This matters because CBT-I can be helpful alongside other care, but it should not be used to explain away red flags. A person with untreated sleep apnoea may feel exhausted because breathing repeatedly collapses during the night, not because they have failed to think correctly about sleep. A person with severe depression may need mental-health treatment as well as insomnia care.
The clinical route is usually straightforward: a primary-care clinician or sleep specialist takes a history, checks for breathing and movement symptoms, reviews medication and alcohol use, and decides whether sleep testing or other assessment is needed. For uncomplicated chronic insomnia, formal overnight testing is often unnecessary. For suspected sleep apnoea, it becomes much more relevant.
What this means in practice
- Do not confuse clean sleep habits with insomnia treatment. Keep the room dark, cool, and quiet, but recognise that chronic insomnia usually needs more than environmental tidying.
- Use a fixed wake time before chasing a perfect bedtime. Morning timing is one of the strongest circadian anchors, and it is usually more controllable than sleep onset.
- Leave the bed when wakefulness becomes established. If the bed has become a place for effort and worry, stimulus control is meant to rebuild the bed-sleep association.
- Consider CBT-I before long-term sleep medication. Medicines can have a role, but the best-supported first-line treatment for chronic insomnia is behavioural and cognitive.
- Screen for other sleep problems. Snoring, breathing pauses, severe daytime sleepiness, restless legs, and pain deserve medical attention rather than another round of sleep hygiene advice.
- Expect the first stage to feel harder, not easier. Sleep restriction can temporarily increase tiredness. It should be done with proper guidance, especially in older adults or people with medical conditions.
What we don’t know
CBT-I has a better evidence base than most insomnia advice, but access remains a problem. Trained clinicians are unevenly available, waiting lists can be long, and private treatment is not affordable for everyone. Digital CBT-I programmes may help close that gap, but they vary in quality and are not the right fit for every patient.
We also know less about some important subgroups than the headlines imply. People with complex psychiatric illness, neurodegenerative disease, severe chronic pain, shift-work schedules, and multiple sleep disorders are often harder to study and harder to treat. In those cases, CBT-I may still be useful, but it is usually one part of care rather than the whole answer.
The largest misunderstanding is gentler: CBT-I is not a promise of perfect sleep. It is a way to make sleep less effortful and more consolidated. For chronic insomnia, that is often the more realistic target. The aim is not to command sleep. It is to stop training the brain to fear the night.
Photo: JP Valery on Unsplash.