CBT-I (cognitive behavioral therapy for insomnia, the structured-protocol kind that actually has a syllabus, not the talk-about-your-sleep-feelings kind)…
Sections
CBT-I (cognitive behavioral therapy for insomnia, the structured-protocol kind that actually has a syllabus, not the talk-about-your-sleep-feelings kind) is the actual first-line treatment for chronic insomnia and almost nobody offers it, so people end up on Ambien for a decade instead. That’s the situation. The treatment works better than the sleep meds, the effect lasts after you stop, and the whole thing takes about six to eight weeks.
The reason it’s not offered comes down to a couple of unglamorous things. Most primary care doctors aren’t trained in it, writing a prescription takes thirty seconds while doing CBT-I takes a clinician hours of skilled time, and insurance is fine paying for the pills. So that’s what most people get handed. A small prescription, indefinitely, and the assumption that the sleep meds are the actual treatment, which they aren’t.
What CBT-I actually involves
There are five or six pieces. The two that do most of the heavy lifting are sleep restriction and stimulus control. The others are sleep hygiene, cognitive work on the beliefs about sleep, and some relaxation training. Most clinicians teach all of them but the engine is the first two, and if your clinician spends most of the time on sleep hygiene without ever bringing up restriction, you’re not getting the real protocol, you’re getting the sanitized version that’s easier to teach and doesn’t actually work as well.
Sleep restriction is exactly what it sounds like and it’s the part people hate. You figure out roughly how many hours you actually sleep on a given night (usually less than you think, and we use a sleep diary, not your memory) and you restrict your time in bed to that number. If you sleep five hours and spend nine in bed, we put you in bed for five and a half hours. You go to bed later and get up at the same time every day, weekends included. No exceptions. The discipline is the medicine.
It’s miserable for the first week or two. You’re tired, sometimes more tired than before you started. That’s intentional, building sleep pressure is the whole mechanism, your body has to actually want to sleep before it’ll do it on demand. After a couple weeks, you start falling asleep fast and staying asleep, and then we add fifteen minutes back at a time until you find what your real sleep need actually is, which is almost always shorter than what you thought.
Stimulus control
The bed is for sleep and sex. That’s it. No phone, no TV, no reading, no laying there at midnight worrying about not sleeping. If you’re in bed for twenty minutes and you’re not asleep, you get up, you go sit in another room with the lights low, and you do something boring. When you feel sleepy you come back. If you’re not asleep in twenty minutes, you get up again.
The point is to retrain the brain to associate the bed with sleep, not with the experience of laying there worrying about not sleeping. If you’ve been a chronic insomniac for years, your brain has done the opposite association without anybody asking it to. Walking into the bedroom triggers wakefulness, the smell of the sheets is now the smell of a bad night. Stimulus control reverses that, slowly, by making bed boring unless you’re actually sleeping in it.
The cognitive part
Most chronic insomniacs are running a set of beliefs about sleep that are making the problem worse. The big ones… I need eight hours or I can’t function tomorrow, if I don’t sleep tonight I’m going to be wrecked, sleep medication is the only thing that works for me, I’ve always been a bad sleeper and this is just who I am.
Some of those are wrong in ways the data actually addresses. Most adults function fine on six and a half to seven hours. One bad night doesn’t wreck the next day for most people. Most chronic insomniacs sleep more than they think they do, the sleep diaries routinely catch about an hour of sleep the person was sure they didn’t get. The catastrophizing about not sleeping is doing as much damage as the not sleeping, sometimes more, and the cognitive work is identifying those beliefs and arguing with them with actual evidence. It sounds soft, it’s actually surprisingly direct work.
The guy I think about
Say you’ve got a guy who owned a small business, had been on Ambien for nine years, came in because his PCP wouldn’t refill it anymore without him trying something else first. He didn’t want to be there. He thought the Ambien was working fine and the doctor was being a pain in the ass for making him jump through hoops.
His sleep diary told a different story. He was in bed nine hours a night and getting maybe five and a half hours of actual sleep. He’d take the Ambien, sleep four solid hours, wake up at 3am, lay there until 5am, doze off again until his alarm went off. He thought of himself as a guy who needed nine hours to function. He’d been operating on five and a half for nine years and his business hadn’t fallen apart, which was the data point he didn’t want to look at, because looking at it meant the story he’d been telling himself about his own sleep need was wrong.
We restricted him to six hours in bed. First week was rough, he was angry about it and called twice asking for permission to give up. Second week he was sleeping through. By week four he was off the Ambien. We added time back gradually and landed at seven hours in bed, six and three quarters of sleep, no medication, no melatonin, nothing. Years later it’s still holding. He sleeps less than he used to plan for, sleeps better than he ever did, and didn’t believe any of it would work for the first month of the protocol.
The catastrophizing about not sleeping is doing as much damage as the not sleeping. The cognitive work is identifying those beliefs and arguing with them with evidence.
What’s nice to hear
Most coverage of CBT-I leads with how miserable the first two weeks are, so here’s the other half. When the protocol clicks, it tends to click pretty cleanly. The transition from “I’m not sleeping” to “I’m sleeping fine, I just sleep less than I thought I needed” is genuinely satisfying when it happens, because it’s not a maintenance situation, it’s a real fix that you got to by doing something hard. Patients usually report sleeping better than they did before whatever event kicked off the insomnia in the first place. Better, deeper, more efficient. And then they don’t think about sleep anymore, which is the actual definition of good sleep… not thinking about it.
How to access it in OR/WA
The barrier is finding somebody who actually does CBT-I. Most therapists don’t. There are a handful of trained CBT-I clinicians in Oregon and Washington and you can find them through the Society of Behavioral Sleep Medicine directory online. There are also app-based versions with decent data, particularly Sleepio and the VA’s CBT-i Coach (free, even if you’re not a veteran). They’re not as good as a real clinician working with you, but they are better than nothing and they are dramatically better than another year of Ambien.
If you’re going to do it through an app, do the whole protocol. Don’t pick the parts you like and skip the parts you don’t. The sleep restriction piece is the part you don’t want, and skipping it kills the whole thing, which is the part nobody warns you about because the apps want you to keep using them and the bad first week is when most people quit.
Sleep restriction
Restrict time in bed to your actual sleep amount, build sleep pressure, add time back gradually. The miserable first two weeks are the whole mechanism. Skip this and you’re doing yoga, not CBT-I.
Stimulus control
Bed is for sleep and sex. Twenty minutes awake, get up. Boring room, boring activity, come back when sleepy. Reverses the bedroom-equals-anxiety association the chronic insomniac brain has built.
Cognitive work on sleep beliefs
The eight-hour-rule, the catastrophizing about tomorrow, the “I’ve always been a bad sleeper” story. Most of these are wrong in ways the data actually addresses.
About the sleep meds themselves
Before we land the plane, a word on the Z-drugs (Ambien, Lunesta, Sonata, the class of sleep meds that mostly replaced the older benzo-style sleep meds). They work in the short term. They are not really treatment, they are time-buying. Long-term use carries some unglamorous baggage… rebound insomnia when you try to come off, some increased risk of next-morning grogginess, parasomnia stuff that doesn’t happen to most people but does sometimes happen, and the general issue that the longer you take them the more your brain comes to think of sleep as something the pill does instead of something it does. Trazodone, used off-label at low doses, is the one a lot of clinicians reach for instead because the abuse and tolerance profile is gentler, and it’s a reasonable bridge while you do the actual work. None of these meds are evil. They’re just not the treatment, and a lot of people on Ambien for a decade were never told the treatment exists.
Bottom line
If you’ve been on a sleep med for over a year, CBT-I belongs on the table. The protocol is uncomfortable for a few weeks and effective for years afterward. The fact that it’s not the default treatment in the field is a failure of the system, not a real question about whether it works. Six weeks of doing something hard, against another decade of taking a pill that doesn’t really fix the underlying thing, isn’t a close call.
Sources
- Qaseem A, Kansagara D, Forciea MA, et al. Management of Chronic Insomnia Disorder in Adults: A Clinical Practice Guideline From the American College of Physicians. Ann Intern Med. 2016;165(2):125-133. PMID 27136449.
- Trauer JM, Qian MY, Doyle JS, et al. Cognitive Behavioral Therapy for Chronic Insomnia: A Systematic Review and Meta-analysis. Ann Intern Med. 2015;163(3):191-204. PMID 26054060.
- Edinger JD, Arnedt JT, Bertisch SM, et al. Behavioral and psychological treatments for chronic insomnia disorder in adults: an American Academy of Sleep Medicine clinical practice guideline. J Clin Sleep Med. 2021;17(2):255-262. PMID 33164742.
