CBT-I — the gold-standard insomnia treatment

Why CBT-I exists

Sleep medication works — short-term. Run any benzodiazepine or Z-drug against placebo for two weeks and you'll see a statistically significant benefit. Run them out to twelve months and the benefit shrinks or disappears, while tolerance, rebound insomnia, and (for older adults) fall risk grow.

CBT-I was developed to target the perpetuating factors of chronic insomnia directly — the behaviours and beliefs that keep insomnia going long after the original trigger has passed. It addresses the disorder rather than masking the symptom.

The five components

1. Sleep restriction therapy

The single most powerful component, developed by Spielman in 1987. Counterintuitive: you temporarily reduce your time in bed to consolidate sleep.

Mechanism: if you spend 9 hours in bed but only sleep 6, your sleep efficiency is 67%. Restricting you to 6.5 hours in bed rapidly drives efficiency above 85% — sleep becomes consolidated rather than fragmented. Once efficiency is consistently high, time in bed is extended in 15-minute increments week by week.

It feels brutal in week one. By week two, most patients are sleeping better than they have in years.

2. Stimulus control

Developed by Bootzin in 1972. The core insight: in chronic insomnia, the bed has become a conditioned cue for wakefulness, frustration, and arousal — the opposite of what it should be.

The rules:

• Bed is for sleep and sex only. Nothing else.
• Get into bed only when sleepy (not just tired — sleepy).
• If you can't sleep within ~20 minutes, leave the bedroom. Do something quiet and dim. Return only when sleepy.
• Keep a consistent wake time every day, regardless of how the night went.
• No daytime napping.

3. Cognitive restructuring

Targets the dysfunctional thoughts that maintain insomnia. The classic pattern: “I have to sleep right now or tomorrow will be a disaster.” “I'll never sleep normally again.” “If I don't get eight hours I'll get sick.”

These thoughts produce arousal, which produces insomnia, which confirms the thoughts. CBT-I systematically identifies and challenges them — replacing catastrophising with realistic appraisal (“I've had bad nights before and managed the next day”).

4. Sleep hygiene education

Largely the same as what's on our hygiene page: consistent timing, dark/cool room, caffeine cut-offs, alcohol restriction. Importantly, CBT-I treats hygiene as necessary but not sufficient. Hygiene alone almost never resolves chronic insomnia — it's baseline maintenance.

5. Relaxation training

Progressive muscle relaxation, paced breathing, or guided imagery — used to reduce cognitive and physical arousal at bedtime. Adjunctive rather than core. Not all CBT-I protocols include it; some patients find it helpful, others find it adds another performance pressure.

What the trials actually show

Morin et al (2009) ran the definitive head-to-head: CBT-I alone versus zolpidem alone versus combined, in adults with chronic insomnia.

• At 6 weeks: all active treatments worked, roughly equivalently.
• At 12 months: CBT-I gains were maintained. Zolpidem gains were largely lost.
• Combined treatment was the best at 6 weeks but no better than CBT-I alone at 12 months.

Replicated repeatedly. The 2021 AASM guideline reviewed every trial since and concluded: CBT-I should be offered first, with medication considered only when CBT-I is unavailable or has failed.

How to access CBT-I

Three routes, in approximate order of efficacy:

• A trained CBT-I therapist — best evidence, highest cost, often hardest to access. Ask your GP for a referral or search the British Sleep Society / AASM directories.
• Digital CBT-I via validated apps — Sleepio (NHS-approved in the UK) and Somryst (FDA-approved in the US) have head-to-head trial data showing comparable effects to therapist-led CBT-I for moderate cases.
• Self-guided CBT-I via books or workbooks — modestly effective. Worth trying before nothing.

When CBT-I isn't appropriate

Sleep restriction can be unsafe in:

• Bipolar disorder (sleep restriction can trigger mania).
• Severe untreated depression with suicidal ideation.
• Untreated sleep apnoea (treat apnoea first).
• Active substance withdrawal.
• Conditions where daytime sedation creates safety risks (e.g. commercial drivers).

Get an assessment before starting if any of these apply.