How to Actually Do CBT-I: The Evidence-Based Fix for Chronic Insomnia

What CBT-I actually is

CBT-I is a structured protocol developed over decades of sleep research that addresses the behavioral and cognitive patterns maintaining chronic insomnia. The landmark name in the field is Charles Morin at Laval University, whose 1993 study in the Journal of Consulting and Clinical Psychology established CBT-I as superior to sleep medication for long-term outcomes. The American College of Physicians formally endorsed it as the first-line treatment for chronic insomnia in 2016, ahead of any pharmacological option.

CBT-I has three core components: sleep restriction therapy, stimulus control, and cognitive restructuring. Most people who fail with informal sleep hygiene tips have not tried any of these. They are mechanistically different from advice like "avoid screens before bed" and produce different physiological outcomes.

The standard CBT-I course runs 6 to 8 weeks, either with a therapist, via digital platforms like Sleepio (which has clinical trial data) or the free CBTI Coach app from the VA, or using structured workbooks. The most important variable is adherence to sleep restriction, which is the component most people abandon prematurely.

Sleep restriction: the hardest and most powerful component

Sleep restriction therapy sounds counterintuitive: you temporarily limit time in bed to the amount you are actually sleeping. If your sleep diary shows you average 5.5 hours of sleep while spending 8 hours in bed, your initial time in bed window is set to 5.5 to 6 hours. The goal is to build intense sleep pressure through adenosine accumulation, forcing consolidation of fragmented sleep into a shorter, deeper window.

The first week is difficult. You will be sleepy earlier than your current bedtime, which is the point. Sleep pressure (adenosine drive) and circadian rhythm alignment converge at a specific window when sleep is easiest. Sleep restriction forces you into that window rather than lying awake in bed for hours building anxiety about not sleeping.

Sleep efficiency is the key metric: total sleep time divided by time in bed, multiplied by 100. A person spending 9 hours in bed but sleeping 5.5 has 61% efficiency. The goal of sleep restriction is to raise this above 85% by building genuine sleep pressure and eliminating the wakefulness being accumulated in bed.

Stimulus control: rebuilding the bed-sleep association

Stimulus control is based on classical conditioning research. When a person spends hours lying awake in bed frustrated, anxious, or watching TV, the bed becomes a conditioned stimulus for wakefulness and arousal rather than sleep. Stimulus control systematically reverses this. The rules are straightforward but require strict adherence.

The get-out-of-bed rule is the most difficult for most people. It feels counterintuitive to leave bed when you are trying to sleep more. But staying in bed while awake strengthens the wakefulness association. Over 2 to 3 weeks of consistent stimulus control, most people find they fall asleep faster because the bed has been reconsolidated as a sleep cue rather than an arousal cue.

Cognitive restructuring: stopping the thoughts that keep you awake

The cognitive component addresses the catastrophizing thoughts that activate the prefrontal cortex when sleep does not come: "I will not be able to function tomorrow," "This is ruining my health," "I have not slept properly in months." These thoughts are not neutral observations. They generate cortisol, raise heart rate, and directly suppress the ventrolateral preoptic area (VLPO), the brain region responsible for initiating sleep.

Cognitive restructuring in CBT-I uses several tools. Decatastrophizing identifies the exaggerated prediction ("I will be completely non-functional") and replaces it with an evidence-based alternative ("I have functioned on poor sleep before; today will be harder but manageable"). Sleep effort paradox reframing teaches that trying hard to sleep activates arousal, the opposite of what sleep requires: the more you try, the less likely sleep becomes.

Worry time is a behavioral technique: set aside 20 minutes earlier in the evening to write down worries and possible solutions. When worries arise at night, acknowledge them and defer them to the designated worry window. This is not suppression; it is structured scheduling of cognitive work that does not belong in the presleep period.

What your wearable data shows during CBT-I

Week one of sleep restriction often shows lower sleep scores than your baseline, which is disorienting. You are sleeping fewer hours in a more consolidated window. Your Oura or WHOOP readiness score may drop transiently. This is expected and should not cause you to abandon the protocol. What to watch for is sleep efficiency rising above 80%, then 85%, which typically happens by week 2 to 3 if the rules are followed consistently.

HRV typically improves measurably by weeks 3 to 4, as consolidated sleep allows deeper parasympathetic recovery during the first half of the night. Sleep latency (time to sleep onset) is usually the first metric to improve, often by the end of week 2. Deep sleep percentage may initially appear lower as absolute time is constrained, then recovers as the window expands.

Digital CBT-I tools that have clinical evidence

Therapist-led CBT-I is the gold standard but access is limited by cost and provider availability. Several digital tools have randomized controlled trial evidence supporting their efficacy.

For people who want a book-based approach, "Say Good Night to Insomnia" by Gregg Jacobs (Harvard Medical School) is the most accessible self-administered CBT-I workbook. It follows the same protocol structure as therapist-led treatment and has been validated in clinical populations. The workbook approach requires discipline but costs less than a single therapy session.

Frequently asked questions