Insomnia

Insomnia is difficulty falling asleep, difficulty staying asleep, or waking far earlier than intended — combined with the daytime consequences of it. That second half matters. Someone who sleeps six hours and functions perfectly well does not have insomnia. Someone who sleeps six hours and spends the day exhausted, irritable and unable to concentrate does.

It is extremely common. Around a third of adults have some trouble with sleep in any given year, and roughly one in ten has the persistent form.

Short-lived and persistent are different problems

The distinction runs through everything that follows, and older writing on this subject frequently gets it the wrong way round.

Short-term insomnia lasts days or weeks and usually has an obvious cause: a bereavement, a deadline, a new baby, noise, a change of shift, jet lag, an illness. It generally resolves when the cause does, and needs little beyond patience.

Chronic insomnia is sleep difficulty on three or more nights a week, for three months or longer, with daytime consequences. A few weeks of bad sleep is not chronic insomnia and is not permanent, and describing it that way frightens people unnecessarily. Three months is the threshold at which the problem has usually stopped being about whatever started it, and started being self-sustaining.

That shift is the central thing to understand. Chronic insomnia is maintained not by the original trigger but by the responses to it: going to bed earlier to catch up, lying in at weekends, napping, staying in bed awake hoping sleep will arrive, and the anxiety about sleeping that grows out of all of it. Each of those is a sensible-sounding reaction, and each makes the problem more durable.

Spending longer in bed is the commonest response to sleeping badly, and one of the few that reliably makes it worse.

Sometimes the insomnia is a symptom

Before treating insomnia as the problem, it is worth establishing that it is. Poor sleep is a prominent feature of a long list of other conditions, and treating the sleep alone will not get far if one of these is behind it:

• Depression and anxiety. The relationship runs both ways — insomnia is a symptom of both, and persistent insomnia raises the risk of developing them. Early-morning waking in particular is associated with depression.
• Pain of any kind, which is probably the commonest physical cause.
• Sleep apnoea, which is substantially under-recognised in women and often presents as insomnia and exhaustion rather than as snoring. It is covered under sleep apnoea.
• Restless legs, where an urge to move the legs makes falling asleep difficult.
• The menopause, where night sweats and hormonal change disturb sleep considerably.
• An overactive thyroid, prostate trouble or anything else causing repeated trips to the lavatory, heartburn and acid reflux, asthma, heart failure, and neurological conditions including Parkinson's.
• Alcohol, which is discussed below, and medicines — including some antidepressants, steroids, beta blockers, decongestants and stimulants.
• Shift work, where the problem is a circadian one rather than insomnia as such.

Insomnia that has not responded to the measures below, or that comes with daytime sleepiness rather than daytime tiredness, is worth taking to a GP for exactly this reason. Tiredness and sleepiness are not the same: people with insomnia are usually exhausted but cannot nod off, while people falling asleep in the daytime more often have something else.

The treatment that works

This is where UK practice has moved furthest from what most writing on insomnia still says.

Cognitive behavioural therapy for insomnia is the first-line treatment for the chronic form. It is recommended ahead of medication, and in England it is available without a referral through NHS-commissioned digital programmes that run over several weeks — a course of structured sessions worked through at home rather than appointments with a therapist. Face-to-face courses exist but are scarcer.

It is not relaxation training with a different name, and it is not a list of bedtime tips. Two of its components do most of the work, and neither appears in older writing on this subject:

Sleep restriction deliberately reduces the time spent in bed to match the time actually spent asleep, then lengthens it gradually as sleep becomes more solid. Someone in bed for nine hours and asleep for five is instructed to be in bed for about five and a half. It is counter-intuitive, it is mildly unpleasant for the first week or two, and it is the single most effective component. The mechanism is simple: mild sleep deprivation is a powerful sleep-inducer, and consolidating sleep into a shorter window breaks the association between bed and lying awake.

Stimulus control re-establishes the bed as a place for sleeping. Bed is for sleep and sex and nothing else — no television, no phone, no working, no eating, no lying there awake. If sleep has not come within about twenty minutes, the instruction is to get up, go to another room, do something quiet and dull in low light, and return only when sleepy. The same applies to waking in the night.

Alongside those: cognitive work on the beliefs that keep people awake — that a bad night will ruin the following day, that eight hours is required, that sleep is slipping away permanently — and relaxation methods, which help but are the least important part.

Two honest caveats about it. The effects build over weeks rather than days, and people frequently abandon it in the uncomfortable early phase. And the sleep restriction phase can cause real daytime sleepiness while it is under way, which matters for anyone who drives or operates machinery — a reason to start it at a sensible point rather than before a week of long drives. Older writing describes the non-medical treatments as having no side effects at all, and that is not quite true of the effective one.

Sleep hygiene, in proportion

The familiar advice — regular hours, a cool dark quiet room, no screens late, no caffeine in the afternoon, winding down beforehand — is sensible and worth doing. It is also, on its own, the weakest of the measures described here. Trials of sleep hygiene by itself find little effect on established insomnia, and it is frequently used in research as the inactive comparison against which real treatments are measured.

That is not an argument against it. It is an argument against being told that the problem is bad habits, trying harder at the habits, and concluding from the failure that nothing works. The habits are a foundation; sleep restriction and stimulus control are the treatment.

Some specifics, with the ones that are overstated marked as such:

• Consistent waking time is the most useful single habit, and more useful than a consistent bedtime. Weekend lie-ins undo a week of progress.
• Caffeine has a long half-life and is in tea, cola, energy drinks and chocolate as well as coffee. Sensitivity varies a great deal, and stopping it after midday for a fortnight is a reasonable experiment.
• Alcohol brings sleep on faster and then wrecks the second half of the night, suppressing deeper sleep and causing repeated waking. Using it as a sleep aid is one of the surer routes into chronic insomnia.
• Nicotine is a stimulant, and withdrawal during the night fragments sleep.
• Exercise helps sleep and should not be given up. The old instruction to finish five or six hours before bed is stronger than the evidence supports: for most people evening exercise does no harm, though vigorous exercise in the last hour can delay sleep onset.
• Daytime naps are worth dropping while treating insomnia, because they reduce the sleep pressure the treatment depends on. This is specific to insomnia rather than general advice against napping.
• The bedroom should be cool, dark and quiet, which is uncontroversial. Heavy curtains, earplugs and a steady background sound all help where noise or light is the problem.
• The clock should be turned away. Watching the time while trying to sleep raises arousal and achieves nothing, and the same applies to sleep-tracking devices, which make a surprising number of people more anxious about sleep rather than less.
• A heavy meal late is uncomfortable and can cause reflux, which is worse lying down. Beyond that, the advice to eat turkey, milk or peanuts to release serotonin is folklore — the amount of tryptophan involved has no measurable effect on sleep, and getting up to eat in the night works against stimulus control.
• Writing down tomorrow's worries before bed is a genuinely useful technique for people whose minds race, and one of the few bits of traditional advice with decent evidence.
• A new mattress may well be more comfortable. It is not a treatment for insomnia, and a great deal is sold on the suggestion that it is.

Medicines, and what the UK actually does

Sleeping tablets have a place, and it is narrower than their reputation.

For short-term insomnia, where sleep is badly disrupted and the situation is acute, a short course of a hypnotic is reasonable. UK prescribing guidance is specific about how: the lowest effective dose, for three to seven nights only, intermittently rather than every night where possible, and not repeated as a matter of course. Zopiclone and zolpidem are the usual choices, at the lower doses in older people.

For chronic insomnia, hypnotics are not the answer, and this is the point on which older writing is most misleading. Tolerance develops within weeks, so the dose stops working. Dependence follows, both physical and psychological. Stopping causes rebound insomnia worse than the original problem, which is experienced as proof that the tablets were needed. And the daytime effects are real: impaired concentration and coordination, a measurably raised risk of road accidents, and in older people a raised risk of falls, fractures and confusion. Describing short courses as carrying no risk of addiction understates all of this.

A newer class of medicine acting on the orexin system is now licensed in the UK for insomnia lasting three months or more where symptoms are substantially affecting daytime life, with a review after three months and a maximum of around a year. It is for people in whom CBT for insomnia has not worked or is not available, rather than a replacement for it.

Melatonin deserves a clear statement because the position differs by country. In the UK melatonin is a prescription-only medicine. It is not sold in health food shops here, whatever American writing on the subject says, and products bought online from elsewhere are unregulated and of uncertain content. Where it is prescribed, it is a modified-release preparation for adults over about 55, for a limited period — useful for jet lag and for some circadian problems, and modest in ordinary insomnia.

Over-the-counter sleep aids are sedating antihistamines, usually diphenhydramine or promethazine. They are not recommended in UK guidance for insomnia. The evidence that they improve sleep is poor, tolerance develops in days, and they leave a hangover of grogginess the following morning. In older people they carry an anticholinergic burden associated with confusion, falls and urinary retention, and are best avoided altogether. Some formulations also contain a painkiller, which is worth checking before taking anything else containing the same.

Herbal remedies

Several are sold for sleep. Most are ineffective rather than dangerous, and one is genuinely worth a warning.

Valerian is the most commonly recommended. Trials are numerous, small and inconsistent, and the overall picture is of little or no benefit beyond placebo. Rare liver reactions have been reported. Chamomile and lemon balm are pleasant and have not been shown to do much; chamomile can cause reactions in people allergic to related plants. Lavender, taken as a standardised oral preparation, has somewhat better evidence than the others, which is a low bar.

St John's wort should not be taken for sleep, and it appears on lists of sleep remedies for no good reason. It is not a sedative — it is studied as an antidepressant — and it is a potent inducer of liver enzymes, which means it reduces the effectiveness of a long list of prescribed medicines including hormonal contraceptives, anticoagulants, some heart and HIV medicines, immunosuppressants and several antidepressants. Combined with an antidepressant it can cause serotonin syndrome. Anyone taking prescribed medication should not take it at all.

Because herbal products are regulated as food supplements rather than medicines in most cases, the amount of active ingredient varies between brands and batches, and interactions are poorly documented. That is the general reason for caution rather than a verdict on any single one.

The thing nobody mentions

Severe sleep deprivation impairs driving in a way comparable to alcohol, and it does so without the person being aware of how impaired they are. Anyone who has gone several nights without proper sleep should not be driving long distances, and anyone who has felt themselves drifting at the wheel should stop. Where daytime sleepiness has persisted for three months or more, the DVLA has to be told — that applies to sleep conditions generally, not only to sleep apnoea.

When to see a GP

• Sleep difficulty on most nights for a month or more, with the days affected.
• Sleepiness during the day — nodding off while inactive — rather than tiredness, which points away from insomnia and towards something else.
• Low mood, loss of interest, or anxiety alongside the sleep problem.
• Snoring with pauses in breathing, or waking unrefreshed however long the night.
• Long-term use of sleeping tablets, which should be reduced gradually and with advice rather than stopped abruptly.
• Insomnia that has not improved after a proper attempt at the behavioural measures above.
• Sleep problems in pregnancy, or at any point where a medicine is being considered.

And immediately, for anyone whose thoughts in the sleepless hours have turned to harming themselves.