Types of Anxiety Therapy and What the Research Says

Therapists describe their work in many ways, and it is hard to tell from a website what the differences amount to. This article sets out what randomized trials have found for each main approach to anxiety. A randomized trial assigns people to treatments by chance, which makes it the fairest test of whether one works better than another.

Two cautions first. A result from a trial describes that group of participants. It does not predict what will happen for you. And not everyone with anxiety needs therapy at all, as do I need therapy for anxiety? explains.

Which therapy has the most research behind it?

CBT, by a wide margin. CBT is a family of structured therapies that work on patterns of thinking and behavior. A 2016 meta-analysis pooled 144 randomized trials that compared CBT with a waiting list, usual care or a pill placebo. Of those, 54 were trials for major depression and 90 were for three anxiety disorders: 24 for generalized anxiety disorder, 30 for panic disorder and 36 for social anxiety disorder. The pooled effects for the three anxiety disorders were large.

The same paper is frank about the limits. More than 80% of the anxiety trials used a waiting list as the comparison. The authors concluded that the effects are large against a waiting list and small to moderate against usual care or a pill placebo. Only 17.4% of the trials were rated high quality. For generalized anxiety disorder, the estimate dropped after adjusting for publication bias. The authors wrote that the effects "are still uncertain and should be considered with caution."

NICE, the United Kingdom's guideline body, tells clinicians to offer adults with social anxiety disorder individual CBT developed for that disorder. That is UK guidance, not a U.S. rule.

What is exposure, and why is it part of CBT?

Exposure is the part of CBT that deals with avoidance. The National Institute of Mental Health (NIMH) describes it as a type of CBT in which a person spends brief periods in a supportive environment learning to tolerate the distress caused by certain items, ideas or imagined scenes, continued until the fear linked to them goes down over time.

The CBT trials in the 2016 meta-analysis included exposure treatments for anxiety disorders. The article on panic attacks describes the form used for panic.

What do trials of psychodynamic therapy show?

Psychodynamic therapy is a talking therapy that looks at what drives a person's anxiety, including recurring themes and relationship patterns. It has been studied far less than CBT. A 2014 meta-analysis found 14 randomized trials with 1,073 patients across the anxiety disorders. The CBT paper above counted 90 trials for three anxiety disorders alone. The two reviews used different rules for which trials to include, so the counts are not exactly like for like, but the difference in scale is clear. The main trials are these.

• Generalized anxiety disorder, 2009. A German trial assigned 57 patients to CBT (29) or short-term psychodynamic therapy (28), for up to 30 weekly sessions. Both produced significant, large and stable improvements in symptoms of anxiety and depression, with no significant difference on the primary anxiety measure. On measures of trait anxiety, worry and depression, CBT was superior. The trial had no untreated group and, with 57 patients, was too small to show that the two are equivalent.
• Social anxiety disorder, 2013. A German multicenter trial assigned 495 patients to CBT (209), psychodynamic therapy (207) or a waiting list (79). Remission rates were 36%, 26% and 9%. Response rates were 60%, 52% and 15%. Both therapies did significantly better than the waiting list. CBT was significantly superior to psychodynamic therapy for remission but not for response, and did better on measures of social phobia and interpersonal problems, though not on depression. The effect sizes for the differences in remission and response were small. At follow-up two years later, response was approximately 70% and remission nearly 40% for both therapies, and no significant differences were found between them after 6 months.
• Panic disorder, 2007 and 2016. A New York trial of 49 adults compared panic-focused psychodynamic therapy with applied relaxation training, which the authors chose as a credible comparison therapy. It did not include CBT. The psychodynamic group had a significantly greater reduction in panic severity, and more of them responded (73% against 39%). The authors called this preliminary evidence from a small trial. A 2016 trial of 201 patients at two sites compared CBT, psychodynamic therapy and applied relaxation. Results differed by site. At Cornell, the three did not differ on the main measure, the speed of improvement in panic severity, and response was 71% for psychodynamic therapy, 65% for CBT and 30% for applied relaxation. At Penn, CBT and applied relaxation improved more than psychodynamic therapy on the main measure, and response rates of 60%, 63% and 48% did not differ significantly. Dropout was 41% for applied relaxation, 25% for CBT and 22% for psychodynamic therapy. The authors concluded that CBT showed the most consistent performance across sites and that psychodynamic therapy showed promise. Measured differences between the patients at the two sites did not explain the result.

Pooled together, the 2014 meta-analysis found psychodynamic therapy significantly more effective than control conditions and not significantly different from other active treatments, at the end of treatment or at follow-up. Its authors also reported medium to high variation between studies. NICE, in its UK guidance on social anxiety disorder, suggests short-term psychodynamic therapy for adults who decline both CBT and medication, and tells clinicians to be aware of its more limited clinical effectiveness and lower cost effectiveness compared with CBT, self-help and medication.

Put plainly: the trials suggest psychodynamic therapy helps people with anxiety disorders. It has not been tested nearly as much as CBT. In the largest head-to-head trial CBT produced more remissions at the end of treatment, and no significant difference was found at later follow-up. In the two-site panic trial, CBT was the more consistent of the two. Psychodynamic therapy is not as well established as CBT, and it is wrong to say it does not work.

What about ACT and mindfulness-based therapies?

Acceptance and commitment therapy (ACT), mindfulness-based cognitive therapy (MBCT) and mindfulness-based stress reduction (MBSR) have a smaller research base. A 2021 meta-analysis covered 23 randomized trials with 1,815 adults with anxiety disorders, most at unclear risk of bias. Added to usual treatment, all three had short-term effects on anxiety compared with usual treatment alone. Compared with CBT, ACT and MBCT showed comparable effects, and MBSR showed significantly lower effects. Analyses at 6 and 12 months found no significant differences from usual treatment or from CBT.

The authors wrote that it remains unclear whether these treatments have effects beyond those of placebo mechanisms. NICE advises against routinely offering mindfulness-based interventions for social anxiety disorder.

What does Dr. Rabin offer, and what should you ask?

Dr. Rabin's anxiety therapy is psychodynamic, attachment-based, trauma-informed and mindfulness-based. The work looks at what drives the anxiety, such as recurring themes, relationship patterns and the ways a person learned to protect themselves, as well as how to respond to anxious thoughts and physical sensations in the moment. She also offers CBT.

As this article has shown, the approach her anxiety page describes is not the one with the largest research base for anxiety disorders. CBT is. That is a statement about how much each has been studied, and about some trial results, not a prediction for any one person. The practice compares the two in psychodynamic vs. CBT therapy.

Ask any therapist you are considering how they would approach what you are dealing with and whether a structured treatment would suit you better. You can ask Dr. Rabin on a free consultation of about 15 minutes, requested through the contact page. How to choose an anxiety therapist has more questions to bring.

Keep reading

• How to Choose an Anxiety Therapist
• What to Expect in Anxiety Therapy
• How Long Does Anxiety Treatment Take?
• All anxiety guides

Sources

• Cuijpers P, Cristea IA, Karyotaki E, Reijnders M, Huibers MJ, How effective are cognitive behavior therapies for major depression and anxiety disorders? A meta-analytic update of the evidence, World Psychiatry (2016)
• Leichsenring F et al., Short-term psychodynamic psychotherapy and cognitive-behavioral therapy in generalized anxiety disorder: a randomized, controlled trial, American Journal of Psychiatry (2009)
• Leichsenring F et al., Psychodynamic therapy and cognitive-behavioral therapy in social anxiety disorder: a multicenter randomized controlled trial, American Journal of Psychiatry (2013)
• Leichsenring F et al., Long-term outcome of psychodynamic therapy and cognitive-behavioral therapy in social anxiety disorder, American Journal of Psychiatry (2014)
• Milrod B et al., A randomized controlled clinical trial of psychoanalytic psychotherapy for panic disorder, American Journal of Psychiatry (2007)
• Milrod B et al., Psychotherapies for Panic Disorder: A Tale of Two Sites, Journal of Clinical Psychiatry (2016)
• Keefe JR, McCarthy KS, Dinger U, Zilcha-Mano S, Barber JP, A meta-analytic review of psychodynamic therapies for anxiety disorders, Clinical Psychology Review (2014)
• Haller H, Breilmann P, Schröter M, Dobos G, Cramer H, A systematic review and meta-analysis of acceptance- and mindfulness-based interventions for DSM-5 anxiety disorders, Scientific Reports (2021)
• NICE, Social anxiety disorder: recognition, assessment and treatment, CG159, Recommendations (2013)
• National Institute of Mental Health, Psychotherapies (last reviewed February 2024)

Last reviewed October 2026. This guide is general information, not medical advice or a diagnosis. If you are in crisis, call or text 988, or call 911 in an emergency.