That last part surprises almost everyone, including a lot of therapists.
Most people picture CBT for panic as learning to calm down. Breathing techniques, thought records, a cheat sheet of coping statements for the bad moments. Those things exist, and some of them help. But they are not the part that does the work, and mistaking the comfort for the cure is one of the most common reasons people finish a course of therapy still afraid of their own heartbeat.
What keeps the panic cycle going?
A panic attack sustains itself through interpretation. Your body produces a sensation, your mind reads it as a sign of catastrophe, the fear response amplifies the sensation, and the louder sensation confirms the interpretation.
A 2018 meta-analysis in PLOS One by Ohst and Tuschen-Caffier found that catastrophically misreading bodily sensations is what most distinguishes panic disorder from other anxiety disorders, with medium to large effects compared with both healthy people and people with other anxiety conditions. The sensation is not the problem. The sentence you attach to it is.
Then avoidance arrives and welds the whole thing shut. You stop drinking coffee, stop the gym, stop the highway, and every avoided thing quietly teaches your brain that it was right to be afraid. The fear never gets a chance to be disproven. Panic disorder develops almost entirely through this loop of learning.
What are the parts of CBT for panic attacks?
Standard CBT for panic has four moving parts, and they are meant to work together rather than as a menu.
- Psychoeducation: a working explanation of what a panic attack actually is, so the sensations stop being mysterious evidence of dying.
- Cognitive restructuring: examining and testing the catastrophic interpretations. Not positive thinking. Accurate thinking.
- Interoceptive exposure: bringing on the feared body sensations on purpose, in a controlled way.
- Situational exposure: returning, gradually, to the places and activities that panic has taken away.
The protocol used in the major trials, including Barlow and colleagues’ 2000 JAMA study of 312 people with panic disorder, combined interoceptive exposure, cognitive restructuring, and breathing retraining across 11 sessions in 12 weeks.
What is interoceptive exposure?
Interoceptive exposure means deliberately producing the physical sensations you are afraid of so your nervous system can learn, from direct experience, that they are not dangerous. “Interoceptive” just means sensations coming from inside your own body.
In practice it looks strange and it is disarmingly simple. In session, that might mean:
- Breathing quickly through a straw to bring on breathlessness
- Spinning in a chair to produce dizziness
- Running in place or up stairs to raise your heart rate
- Holding your breath to create chest tightness
- Staring at a light or a pattern to induce a sense of unreality
Each one recreates a piece of a panic attack on purpose. You stay with it. It peaks. It passes. Nothing happens. And then you do it again, until a pounding heart is just a pounding heart.
This is the part clients look at me sideways for the first time I describe it. It sounds backwards, because every instinct says the sensations are the enemy. The logic is the same as any exposure work: you cannot argue your body out of a fear it learned through experience. You have to give it a different experience.
Which parts of CBT actually do the work?
The evidence points to exposure, and points away from relaxation.
The most direct test of this is a 2018 component network meta-analysis in Psychological Medicine by Pompoli and colleagues, which took apart 72 trials involving 4,064 people to see which pieces of CBT carried the effect. Interoceptive exposure had the strongest positive association with remission (incremental odds ratio 1.49). Muscle relaxation went the other way, and clearly: an incremental odds ratio of 0.59 for remission, meaning treatments that included it did worse. Breathing retraining sat near neutral for remission at 0.84.
The gap between the best and worst combinations of CBT components was an odds ratio of 7.69. Same label, “CBT,” radically different results depending on what is actually inside it. This is why asking a therapist what their panic treatment contains is a fair and useful question.
The clinical reading of this is not that relaxation is bad. It is that relaxation can become a safety behavior. If you use breathing to make the sensations go away, you have taught yourself one more time that the sensations needed to go away. That is the opposite of what exposure is for.
I still teach breathing skills to some clients. I am just careful about when, and clear about what job they are doing.
How many sessions of CBT does panic take?
Most CBT for panic attacks runs roughly 12 weekly sessions. The Barlow trial used 11 sessions across 12 weeks of acute treatment, and most published protocols land in the 12 to 15 range.
Improvement usually starts well before the end. What often takes longer is the avoidance, particularly for people who have spent years narrowing their world, or who are dealing with agoraphobia alongside the panic. Long-standing avoidance takes more repetitions to unwind than the attacks themselves.
Does CBT for panic work over telehealth?
Remote CBT for panic performs comparably to in-person CBT. A 2021 meta-analysis in the Journal of Anxiety Disorders by Efron and Wootton pooled 21 studies and 1,604 participants, finding a large within-group improvement (Hedges’ g = 1.18) and essentially no difference from active in-person treatments (g = 0.02). Videoconference-delivered CBT specifically produced a large effect (g = 1.40).
For panic this matters more than it does for most conditions, because the drive to the office can itself be the feared situation. Sessions are available throughout California and Colorado.
What does a CBT session for panic actually look like?
Early sessions are mostly mapping. What happens, in what order, in what situations, and what you have quietly stopped doing. Most people are startled by how long the avoidance list turns out to be once it is written down.
Middle sessions are where the exposure work lives, in session and between sessions. This is the working part of therapy and it is collaborative. You are never ambushed, and nothing happens that you have not agreed to.
Later sessions are about consolidating: what you now know, what to do if panic shows up again, and how to keep from rebuilding the avoidance out of habit.
This is the structure behind panic attack and panic disorder therapy here, and it sits inside a broader Integrated Panic Recovery approach. If you are still weighing your options, the most effective treatment for panic disorder compares CBT against medication and what happens after treatment ends.
The short version
- CBT for panic attacks targets interpretation and avoidance, not calmness.
- Interoceptive exposure is the active ingredient, the component most associated with remission in the largest component analysis to date.
- Relaxation training is not neutral. It was associated with worse remission rates, likely because it functions as a safety behavior.
- Expect about 12 weekly sessions, with improvement usually starting earlier and avoidance taking longest.
- Remote delivery works. The research shows no meaningful difference from in-person treatment.
- You will not be ambushed. Exposure is planned, graded, and agreed to in advance.
The thing worth holding onto is that panic is learned, which means the learning can be redone. Not by getting better at calming down, but by finding out, over and over in small controlled doses, that the alarm was never telling the truth.
New, severe, or unfamiliar physical symptoms, especially chest pain, warrant a medical evaluation first. Ruling out a medical cause is part of good care.
Frequently asked questions
Is CBT for panic attacks the same as regular talk therapy?
No. CBT for panic is structured and skills-based, with a specific target each session and work between sessions. Talking about your history has its place, but insight alone does not typically resolve panic, because panic is maintained by learning rather than by not understanding it.
Will therapy make me have a panic attack on purpose?
Interoceptive exposure deliberately brings on panic-like sensations, not full attacks, and always in a graded, agreed-upon way that you control. You start with the mildest exercises and move up only when you are ready. Most people find the anticipation harder than the exercise.
What if CBT has not worked for me before?
It is worth asking what the previous therapy actually contained. A course of CBT that was mostly coping skills, thought records, and relaxation is a different treatment from one built around interoceptive and situational exposure, even though both get called CBT. It is also worth looking at whether unprocessed trauma is feeding the panic.
References
- Ohst B, Tuschen-Caffier B. Catastrophic misinterpretation of bodily sensations and external events in panic disorder, other anxiety disorders, and healthy subjects. PLOS One, 2018. Read meta-analysis
- Pompoli A, Furukawa TA, Efthimiou O, Imai H, Tajika A, Salanti G. Dismantling cognitive-behaviour therapy for panic disorder: a systematic review and component network meta-analysis. Psychological Medicine, 2018;48:1945–1953. Read study
- Barlow DH, Gorman JM, Shear MK, Woods SW. Cognitive-behavioral therapy, imipramine, or their combination for panic disorder: a randomized controlled trial. JAMA, 2000;283:2529–2536. Read study
- Efron G, Wootton BM. Remote cognitive behavioral therapy for panic disorder: a meta-analysis. Journal of Anxiety Disorders, 2021;79:102385. Read meta-analysis