After CBT-I: What to Do When Insomnia Comes Back

5:52 am. I slept well for three months. Then last Tuesday happened: a 2am wake-up, the old familiar pattern, the ceiling, the clock, the thoughts. Wednesday was worse. Thursday I was counting sheep and hating myself for counting sheep.

It is back, I thought. Everything I did was temporary. I am an insomniac again.

I was wrong about the third sentence, but right about the first two. The insomnia was back. It had not been permanently fixed. That is not what CBT-I promises, and understanding what it does promise is the difference between a setback and a relapse.

Why bad nights return

Spielman’s 3P model explains this cleanly [2]. Three factors interact: predisposing factors (the biological and temperamental vulnerabilities that made you susceptible to insomnia in the first place), precipitating factors (the event that triggered the first episode), and perpetuating factors (the behaviors and cognitions that kept it going once it started).

Your predisposing factors do not disappear after treatment. They are part of your physiology. Some people have a nervous system that runs hotter, a stress response that activates more easily, a tendency toward rumination that extends into the night hours. CBT-I does not change your temperament. It addresses the perpetuating factors: the behavioral and cognitive patterns that turned a few bad nights into chronic insomnia.

When a new precipitating event arrives, and one always does eventually (a stressful week, an illness, travel across time zones, a major life change), the predisposing vulnerability reactivates. You have a bad night. Maybe two. This is normal. It is not a sign that the treatment failed. It is a sign that you are a human with a nervous system, and your nervous system was always going to respond to stress the way it responds to stress.

The critical moment is what happens next. Before CBT-I, you responded to bad nights with the perpetuating behaviors: extending time in bed, going to bed early, lying awake hoping for sleep, napping to compensate, catastrophizing about the consequences. Those responses turned a few bad nights into months of insomnia. The responses were the problem, not the trigger that started them.

After CBT-I, you have a choice. You can respond the old way, or you can apply what you learned. The choice is not easy at 2am. But it is available to you in a way it was not before treatment, because you have seen the mechanism from the inside and you know which responses feed it.

The early intervention window

The transition from a few bad nights to chronic insomnia does not happen instantly. There is a window, typically one to two weeks, during which the perpetuating cycle is re-establishing but has not yet consolidated. The bed-wakefulness association is weakening again but has not fully reformed. The catastrophic thoughts are returning but have not yet become automatic. This is when the CBT-I tools are most effective at preventing relapse.

The principles are the same ones you learned during treatment:

Keep your wake time fixed. Do not sleep in to compensate. The alarm goes off, you get up. The circadian anchor holds the system in place. Sleeping until 10am on Saturday after a bad Friday night feels restorative, but it shifts your circadian rhythm and dilutes the sleep pressure you need for Saturday night. The anchor is the most important thing to hold.

Do not extend time in bed. The temptation after a bad night is to go to bed an hour early. This dilutes sleep pressure and increases the time available for lying awake. Keep your sleep window roughly where it was when things were going well. If you were going to bed at 11pm and sleeping until 6:30am, do not start going to bed at 9:30pm. Those extra ninety minutes will not be sleep. They will be the conditions for the cycle to restart.

Get out of bed if you are awake. If you wake at 2am and cannot fall back asleep within about twenty minutes, get up. The bed-wakefulness association reactivates faster than you think. A few nights of lying awake in bed can undo weeks of conditioning. I know the living room is cold. I know the couch is uncomfortable. The discomfort is temporary and the reconditioning is the point.

Do not nap. The afternoon nap discharges the sleep pressure you need for tonight. Resist it, even though the tired afternoon feels like it demands one. The pressure is working for you. Let it build.

Name the catastrophic thought. “It’s back, I’ll never sleep again, everything is ruined” is a prediction, not a fact. You have tested this prediction before. It was wrong before. It is probably wrong now. The thought feels different this time, heavier, more certain. It always feels different this time. That is how catastrophic thoughts work: they present each instance as unique and unprecedented, even when the evidence shows a pattern of overestimation.

The evidence for durability

The good news from the research is that CBT-I’s effects are remarkably durable. Trauer et al.’s meta-analysis found that improvements in sleep onset latency, wake after sleep onset, and sleep efficiency were maintained across follow-up periods [4]. The gains did not vanish when the program ended. They persisted, because the behavioral changes were self-reinforcing: better sleep reduced anxiety about sleep, which reduced arousal, which maintained better sleep.

Morin et al.’s trial showed gains sustained at 6 and 24 months [1]. Two years after treatment, the CBT-I group was still sleeping better than they had before the program. Not perfectly. Not every night. But substantively and measurably better, in a way that medication groups did not maintain once the medication stopped.

Edinger et al. found reduced relapse rates in the CBT-I group compared to other treatments at long-term follow-up [3]. The people who learned the behavioral skills relapsed less often, and when they did have setbacks, the setbacks were shorter and less severe. This is the difference between a treatment that addresses symptoms and a treatment that addresses the mechanism: symptom treatments stop working when you stop taking them; mechanism-based treatments leave you with tools that remain available.

The durability reflects a fundamental difference between CBT-I and medication. Medication effects end when the medication stops. CBT-I teaches skills that remain available after treatment ends. Those skills need to be reactivated during setbacks, but they do not need to be re-learned from scratch. The wiring is still there. You just need to use it.

Booster sessions

For setbacks that do not respond to self-management within two to three weeks, a booster session with a CBT-I therapist can recalibrate the treatment. Buysse et al.’s research on brief behavioral interventions suggests that one to two focused sessions can restore the treatment effect without requiring a full course of therapy [5].

A booster session typically involves reviewing the sleep diary, recalculating sleep efficiency, re-establishing the sleep restriction window, and troubleshooting the specific perpetuating behaviors that have re-emerged. It is a tune-up, not a rebuild. The therapist can see patterns you cannot: a gradual drift in bedtime that you did not notice, an increase in time-in-bed that crept in over weeks, a resumption of clock-watching that you did not register as a monitoring behavior.

If you do not have access to a therapist, a digital CBT-I program can serve a similar function. Returning to the app for a few weeks during a setback provides the structure and tracking that self-management alone sometimes lacks. The structure matters. During a setback, the motivation to follow the protocol dips at exactly the moment when following it matters most, and an external system that tracks compliance can compensate for the internal system that is flagging.

Self-monitoring

One of the most effective relapse prevention strategies is periodic check-ins with a sleep diary. You do not need to track every night indefinitely. But keeping a diary for one week every month or two provides early warning if sleep efficiency is declining or if time in bed is creeping upward.

The drift is usually gradual. Bedtime moves fifteen minutes earlier. A weekend nap becomes a habit. The alarm gets hit once or twice. Individually, each change is small. Collectively, they reconstruct the conditions for insomnia. A periodic diary check catches the drift before it becomes a pattern. I do a check-in week every six weeks or so. Most of the time, the numbers are fine and I file the diary and move on. Twice, the numbers showed a decline I had not felt yet, and I tightened my window before the problem became visible.

What makes this hard

The hardest part of a setback is the emotional response. Three months of good sleep creates an expectation that the problem is solved. When it comes back, even briefly, the disappointment is disproportionate to the actual disruption. Two bad nights after three good months is a minor fluctuation in any objective framework. It does not feel minor. It feels like proof that you are broken in a way that cannot be fixed.

The catastrophic thought, the one that says “this is permanent, the treatment was a mirage,” is the most dangerous response because it generates the anxiety that fuels the very cycle you are trying to prevent. It is the perpetuating factor in its purest form: a thought about insomnia that creates insomnia. And it is persuasive at 2am in a way that logic cannot easily counter, because the emotional weight of the thought overwhelms the evidence against it.

You have a counter to this thought now. You did not have it before CBT-I. The counter is your own data: the diary entries from the three good months, the sleep efficiency numbers that showed improvement, the evidence that your sleep system works when the perpetuating factors are absent. The data does not make the thought disappear. It makes the thought arguable, and arguable is enough to keep you from rebuilding the cycle around it.

The bad nights will come back. They come back for everyone. What you do with them is what determines whether they stay.

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.