A 2018 research review brought together 41 carefully controlled studies involving nearly 3,000 people with conditions including OCD, anxiety and PTSD. Each study compared cognitive behavioural therapy, a structured and practical form of talking therapy, against a placebo condition designed to account for the effect of simply talking to someone.
CBT came out clearly ahead. For OCD specifically, the results were among the strongest of any condition examined. People who received CBT, particularly the exposure-based form, were nearly three times more likely to show meaningful improvement than those who did not.
What that means in ordinary terms is that intrusive thoughts, rituals and the exhausting mental loops OCD creates are not something a person simply has to accommodate. They are treatable, and the treatment has been tested against the most demanding comparison researchers can construct.
OCD is often accompanied by embarrassment, which is one reason people delay seeking help. If any of this is familiar, speaking with a psychologist is a practical and well-evidenced step.
What the study actually did
This paper is a meta-analysis, a study that statistically combines results from many separate trials to produce a more reliable overall picture than any single trial can give. Carpenter and colleagues pooled 41 randomised controlled trials involving 2,843 participants across six anxiety-related conditions: acute stress disorder, generalised anxiety disorder, obsessive compulsive disorder, panic disorder, PTSD and social anxiety disorder. In a randomised controlled trial, people are assigned at random to different treatments, which helps rule out the possibility that differences in outcome simply reflect who volunteered. These were also placebo-controlled trials, meaning participants received either CBT or a placebo condition, either a convincing but inactive psychological treatment or a placebo pill. That design separates the specific effects of CBT from general effects such as feeling cared for or expecting to improve.
Making sense of the numbers
Results are reported using Hedges' g, an effect size that expresses how large a difference is in standardised form so it can be compared across studies. As a rough guide, 0.2 is considered small, 0.5 moderate and 0.8 large. The study also reports an odds ratio, which compares the odds of a good outcome in one group against another. An odds ratio of 1 would indicate no difference.
What it found
Across all disorders combined, CBT outperformed placebo on the symptoms it targeted, with a moderate effect of g = 0.56, and produced smaller but genuine improvements in other anxiety symptoms, depression and quality of life. People given CBT were roughly three times more likely to show a meaningful response than those given placebo, with an odds ratio of 2.97.
The strength of the effect varied by condition. OCD, generalised anxiety disorder and acute stress disorder showed large effect sizes, while PTSD, social anxiety disorder and panic disorder showed only small to moderate effects. Effects were stronger when the analysis was limited to people who completed treatment rather than everyone enrolled, and individual therapy outperformed group therapy for social anxiety and PTSD. Treatments built mainly around exposure, meaning gradually confronting feared thoughts, situations or triggers rather than avoiding them, tended to produce larger effects than more purely cognitive approaches, although that particular difference was not statistically significant and could plausibly reflect chance.
Why this matters for OCD
OCD stood out, appearing among the conditions with the largest treatment effects in the analysis. For anyone dealing with intrusive thoughts and compulsions, that is an evidence-based reason for optimism, because this is not a condition where therapy offers only modest help. The trend towards exposure-based strategies performing especially well is also relevant, since exposure and response prevention is a core technique within CBT for OCD.
What this study does not tell us
A meta-analysis is only as good as the trials feeding into it, and this one focused specifically on placebo-controlled designs, so it does not capture every study of CBT ever conducted. It cannot explain why some disorders such as PTSD or panic disorder showed smaller effects, nor identify which individual factors predict who responds best. Averages describe groups rather than guaranteeing an outcome for any one person.
If OCD is affecting your life, this kind of evidence is good reason to consider speaking with a psychologist.