What Is CBT for Insomnia, and Why Is It the First-Line Treatment?

2:48 am. I have been looking at the ceiling long enough that I have stopped forming complete thoughts and started just observing the shapes my brain makes in the dark. This is the fourth night in a row. I have tried melatonin, chamomile, a lavender pillow spray that cost more than dinner, and a meditation app that keeps telling me to observe my breath with curiosity. My breath is not interesting. My ceiling is not interesting. I am extremely awake.

If this sounds familiar, you are probably part of a large and exhausted club. Chronic insomnia affects roughly 10% of adults at any given time, and most of them have already tried the standard advice: consistent bedtime, dark room, no screens. The advice is not wrong. It is just radically incomplete. It is like telling someone with a broken bone to eat more calcium: technically relevant, practically useless as a treatment.

The treatment most people have not heard of

Cognitive Behavioral Therapy for Insomnia, abbreviated CBT-I, is a structured program that typically runs six to eight sessions. Both the American Academy of Sleep Medicine and the American College of Physicians recommend it as the first-line treatment for chronic insomnia, ahead of any medication [3] [4]. That recommendation is not tentative. It is based on decades of randomized trials showing durable results that medication has difficulty matching over the long term.

I did not hear about CBT-I from my doctor. My doctor offered trazodone and told me to try cutting caffeine after noon. Both were reasonable suggestions, and neither addressed the actual mechanism keeping me awake. I found CBT-I the way most people find it: searching the internet at 3am, exhausted enough to click past the first page of “sleep hygiene” listicles into something more substantive.

CBT-I has five components, and sleep hygiene is only one of them. The others are the ones that actually do the heavy lifting:

Sleep restriction: Reducing the time you spend in bed to match the time you actually sleep, then gradually expanding it. If you are sleeping five and a half hours but lying in bed for eight, your sleep window starts at five and a half hours. This is the hardest part and the most effective part.

Stimulus control: Reserving the bed for sleep only. If you are awake for more than about twenty minutes, you get up and go to another room until you feel sleepy. The mechanism: you are breaking the association between bed and wakefulness. The bed has become a cue for frustration and alertness. Stimulus control makes it a cue for sleep again.

Cognitive restructuring: Identifying and challenging the catastrophic thoughts that fuel insomnia. “If I don’t sleep tonight, I won’t function tomorrow” is a prediction that feels true at 3am and rarely turns out to be accurate in the way you feared. You are not replacing the thought with a happy one. You are replacing it with a more accurate one.

Relaxation training: Techniques like progressive muscle relaxation, used specifically to reduce physiological arousal. This is not “just relax.” It is a trained skill aimed at the body’s fight-or-flight system, practiced outside the bedroom so it does not become another performance task in bed.

Why it works when tips do not

Sleep hygiene listicles treat insomnia as a habits problem. CBT-I treats it as a conditioned arousal problem. The difference matters.

Chronic insomnia is typically maintained by a cycle: you sleep badly, so you start trying harder to sleep, so the bed becomes a place of effort and vigilance, so you sleep worse [1]. Every adaptation you make, going to bed earlier, lying in longer, napping, canceling plans because you might be tired, is a sensible-looking response that strengthens the cycle. You are training your brain that sleep requires management, and a brain that believes sleep requires management is a brain that stays alert in bed.

I made every one of those adaptations. I went to bed at 9pm to give myself the maximum sleep opportunity. I lay perfectly still when I woke at 2am, convinced that movement would wake me up further. I napped on Saturday afternoons. Each strategy was intuitive, and each one was maintaining the insomnia. I was putting enormous effort into sleep, and the effort was the problem, because sleep is not something you achieve through effort. It is something that happens when the conditions are right and you stop interfering.

CBT-I does not tell you to stop worrying about sleep. It changes the conditions so there is less to worry about. Sleep restriction builds up homeostatic sleep pressure until the body’s drive to sleep overwhelms the conditioned arousal. Stimulus control retrains the bed-sleep association. Cognitive restructuring addresses the catastrophic predictions. The program works from the outside in: change the behavior, and the brain follows.

What the research shows

A 2015 meta-analysis of 20 randomized controlled trials involving 1,162 participants found that CBT-I significantly improved sleep onset latency (how long it takes to fall asleep), wake after sleep onset, sleep efficiency and subjective sleep quality [5]. The effects were not small, and they were durable at follow-up. People who completed CBT-I were still sleeping better months after the program ended.

The comparison with medication is telling. Sleeping pills (particularly the Z-drugs like zolpidem) work in the short term, but their effects tend to diminish over weeks to months, and they carry risks of dependency and rebound insomnia on withdrawal. CBT-I’s effects, by contrast, tend to persist and sometimes improve after the program ends, because the behavioral changes are self-reinforcing [1]. A brain that has learned to associate the bed with sleep does not need nightly medication to maintain that association.

A landmark study by Morin and colleagues found that while combined CBT-I and medication produced the best short-term results, CBT-I alone produced the best outcomes at long-term follow-up [1]. The medication group relapsed more often once the pills stopped. The behavioral treatment group maintained their gains because the changes were structural, not chemical.

Digital CBT-I programs have also shown efficacy. A randomized trial of the Sleepio platform found significant improvements versus placebo, making CBT-I accessible to people without a nearby sleep specialist [6]. The effect sizes were smaller than face-to-face therapy, but the access problem is severe enough that a moderate-effect treatment available tonight beats a large-effect treatment available in three months.

What makes this hard

I will not pretend CBT-I is pleasant. Sleep restriction, the most potent component, means being deliberately sleep-deprived for the first week or two. You are going to bed later than you want to and getting up at the same time every morning regardless of how the night went. You will feel worse before you feel better. The first week is the point at which most people who are going to quit, quit.

The dropout rate in trials reflects this. Not everyone completes the program. But among those who do, the success rate of 70-80% is remarkably good for a behavioral intervention, and it comes without the side effects, dependency risk or rebound insomnia that accompany long-term use of sleeping pills.

The other hard part is access. There are not enough trained CBT-I therapists. Most primary care doctors do not deliver it. Many have not heard of it. The result is a treatment gap: the guidelines say CBT-I first, and the healthcare system says here is a prescription. Digital programs help fill the gap, and a good self-help manual (say, “Quiet Your Mind and Get to Sleep” by Colleen Carney and Rachel Manber) can guide you through the principles. But if you can work with a trained provider, the evidence favors it [2].

What remains uncertain

CBT-I is well established as effective for most people with chronic insomnia. What we know less about is who it does not work for, and why. The 20-30% who do not respond fully are not well characterized in the literature. Research into predictors of treatment response is ongoing, and the answer will likely involve subtypes of insomnia that respond to different combinations of components.

We also do not have a definitive answer on the optimal number of sessions, or on whether specific components (sleep restriction alone, for example) are sufficient for some people. The field is moving toward personalized protocols, but it is not there yet. What we do know, with a confidence the field rarely affords, is that CBT-I works for the majority of people who commit to it. The evidence is not ambiguous. The treatment is not speculative. It is the standard of care, and the only thing preventing more people from benefiting is the gap between what the guidelines recommend and what most doctors offer. What we do know, with a confidence the field rarely affords, is that CBT-I works for the majority of people who commit to it. The evidence is not ambiguous. The treatment is not speculative. It is the standard of care, and the only thing preventing more people from benefiting is the gap between what the guidelines recommend and what most doctors offer.

If your sleep is affecting daily life, talking to a GP or sleep specialist is a good first step. The American Academy of Sleep Medicine has a provider finder at sleepeducation.org. If insomnia is accompanied by depression, anxiety or other concerns, those are worth raising with your doctor as part of the same conversation.