Many people with anxiety try yoga or mindfulness before considering formal therapy, and the appeal is understandable. It feels gentler, more accessible and less clinical. The question is how yoga compares with the best-supported treatment for anxiety when the two are tested directly.
A 2021 clinical trial by Simon and colleagues did exactly that, comparing yoga, cognitive behavioural therapy and stress education for people with generalised anxiety disorder, a condition in which worry becomes uncontrollable and consuming.
Generalised anxiety disorder is common and often goes untreated for years. People affected tend to worry excessively about health, work and relationships even when no obvious trigger is present. The constant mental noise is exhausting and can be genuinely disabling.
The trial found that CBT outperformed both yoga and stress education in reducing anxiety symptoms. Yoga showed some benefit and was not without value, but CBT produced consistently stronger results. The finding is useful because it clarifies the role of each. Yoga may help with managing day-to-day stress, and it is not a substitute for evidence-based therapy once anxiety has begun interfering with your life.
Effective help exists. If worry is running your life, a trained psychologist can offer more than coping techniques.
What the study actually tested
This trial, led by Simon and colleagues, was a randomised controlled trial: a design in which participants are assigned at random to different treatments so that the groups start out similar on average, allowing later differences to be attributed to the treatment itself. It had three arms. Participants with generalised anxiety disorder were randomised to Kundalini yoga, a style combining postures, breathwork and meditation, with 93 people; cognitive behavioural therapy, a structured talk therapy targeting unhelpful thought patterns and avoidance behaviours, with 90 people; or stress education, an active control condition involving general psychoeducation about stress, with 43 people. All three were delivered in a group format over twelve 120-minute sessions plus daily homework.
The trial was single-blind, meaning the outcome raters did not know which treatment a participant had received, which reduces bias in how improvement is judged. Researchers tested two separate questions: whether yoga and CBT were each superior to stress education, and whether yoga was noninferior to CBT, meaning close enough in effectiveness to count as an equally valid option.
What it found
Of 226 participants with confirmed GAD, 155, or 68.6%, completed the post-treatment assessment, with no meaningful difference in dropout between groups. Using independent raters' judgments of clear improvement, response rates were 54.2% for yoga, 70.8% for CBT and 33.0% for stress education. Both active treatments beat stress education. Yoga was about 2.5 times more likely to produce improvement, with an odds ratio of 2.46, and CBT about 5 times more likely, with an odds ratio of 5.00. The number needed to treat, meaning how many people would need to receive a treatment for one extra person to benefit compared with the control, was roughly 4.6 for yoga and 2.6 for CBT. Fewer people needed treating with CBT to produce an additional responder.
Yoga did not, however, meet the statistical bar for noninferiority to CBT. The gap between them, 16.6 percentage points, was too large for the test to support treating the two as equivalent.
Why it matters
This is among the more rigorous head-to-head comparisons of yoga against a first-line psychological treatment for GAD, rather than against no treatment at all. It establishes that yoga produces real, measurable improvement beyond an active control. It also establishes that for people whose worry has become disabling, CBT remains the stronger and better-supported option, and the authors concluded it should remain first-line.
Limits worth knowing
Participants were recruited from two academic specialty centres, so they may not represent everyone with GAD in the wider community. About a third did not complete the post-treatment assessment, which affects how confidently the results generalise. The trial also measured outcomes at 12 weeks, so it says nothing about how these effects hold over the longer term.
If worry has started running your life, well-tested treatments exist, and a psychologist is a reasonable place to start.