Meditation reduces stress

Meditation programs lower self-reported stress and distress by a small to moderate amount compared with doing nothing, across hundreds of randomized trials. The clearest single estimate comes from a 2023 analysis of individual records from 13 trials in 2,371 adults: in-person, teacher-led group mindfulness courses lowered distress for one to six months afterward, a result its authors rated with high confidence, and the benefit did not depend on how stressed people were at the start. Compared with other activities that take the same time and attention, such as relaxation training, exercise or health education, meditation courses have done little or no better in pooled analyses: a 2014 review found small advantages for anxiety and depression, with low-strength evidence for stress, and a 2021 review found an advantage for depression only. Two large trials published since, one of a meditation app and one of single short exercises, each found lower stress than after a comparison activity.

Nearly all of this evidence comes from questionnaires filled in by people who knew they were meditating, and results vary by setting: in a 2021 review of 136 trials in adults outside clinical care, the effect expected in a new setting ranged from a large drop in distress to a small rise. Unwanted effects occur. A 2020 review of 83 studies found them in about 8 percent of meditators overall, most often anxiety or low mood, and trials that do not ask about them directly are likely to miss some. Most trials tested mindfulness courses. Mantra meditation, including Transcendental Meditation, lowered stress more than waiting lists or usual care in a 2022 review and showed no clear difference from active comparisons, and app-based practice has been tested in fewer trials, most of them short.

In a 2014 review funded by the US Agency for Healthcare Research and Quality, 47 randomized trials (3,515 adults with a medical or psychiatric condition, or under stress) compared meditation programs with activities matched for time and attention. Mindfulness programs improved anxiety (effect size 0.38 at 8 weeks and 0.22 at 3 to 6 months) and depression (0.30 and 0.23) with moderate-strength evidence, and stress or distress with low-strength evidence. No meditation program did better than the drugs, exercise, relaxation or other behavioural therapies it was compared with, and the evidence on those comparisons was graded low or insufficient, so it does not show the programs were as good either. Only 10 of the 47 trials were at low risk of bias. One trial since has compared a course directly with a medicine. Among 276 US adults with a diagnosed anxiety disorder, an eight-week in-person mindfulness-based stress reduction course was no worse than the antidepressant escitalopram on a clinician's rating of anxiety at eight weeks, by a margin set in advance; when the same trial continued by video call, the two did not differ significantly but the course did not meet that test. The trial measured anxiety in people with a diagnosis, compared starting a course with starting the medicine, and did not test stopping a medicine someone already takes.

The 2021 review pooled 136 trials in 11,605 adults in 29 countries, 77 percent of them women. Compared with no intervention, mindfulness courses reduced distress (standardised mean difference -0.45 across 27 trials), anxiety and depression one to six months later. Against activities designed to match the attention and expectations of a course, only depression differed (6 trials), and against established alternatives such as exercise or other therapies there was no evidence of an advantage. Confidence in these results was rated moderate to very low, mainly because many trials were at high risk of bias, and trials in the United States reported smaller effects. The 2023 analysis of individual records found an effect of -0.32 on distress, with no variation between trials. The 2022 mantra review pooled 14 trials in 1,078 people for stress (effect size -0.45); the effect was clear in healthy adults and not in people with a health condition, was no longer clear in trials that followed people for more than four months, and none of the trials the review covered was at low risk of bias.

For apps, a 2021 review of 34 randomized trials found lower perceived stress after app-delivered mindfulness than in control groups (effect size 0.46 across 15 trials), smaller changes in anxiety and depression, and no clear change in general distress; comparisons against active control conditions rested largely on one trial. The largest app trial since assigned 2,182 health service staff in England to a meditation app or to a health service website of advice on work stress. After four and a half months, stress had fallen by about 2.5 points on a 42-point scale with the app and about 1 point with the website. People used the app more than the website and rated it as more credible; the difference remained when the credibility ratings were taken into account, and the authors state that greater use, rather than meditation itself, may explain it. In a 2024 study at 37 sites, 2,239 adults with an average age of 22 did one short self-guided mindfulness exercise or listened to a story: stress reported shortly afterward was lower after each of four exercises, by amounts the authors describe as small. That study tested a single exercise and short-term stress only.

Beyond questionnaires, a 2017 review of 45 randomized trials that compared meditation with an active control activity found lower cortisol, blood pressure, heart rate and some inflammation markers after meditation when all types were pooled. A 2023 review of 32 randomized trials of mindfulness courses found only a small difference in markers of stress and inflammation, too small to rule out chance, and reported low study quality and signs that favourable results were more likely to be published. Both reviews are about body markers of stress rather than how people feel; the 2017 review's trials included adolescents and people with various health conditions, and few high-quality studies have looked at meditation and blood pressure itself. In the 2020 adverse-event review, the share of meditators reporting an unwanted effect was 3.7 percent in experimental studies and 33.2 percent in surveys of longer-term practitioners, many of whom had attended retreats. Six studies reported suicidal thoughts or behaviour, and case reports described psychosis, mania and depersonalization; 17 of the 31 people in those case reports had no earlier mental health history. In the 2014 review, only 9 of 47 trials reported on harms at all, and in the 2022 mantra review only 4 of 51 did. In the United States, the share of adults who meditate rose from 7.5 percent in 2002 to 17.3 percent in 2022. UK guidance on mental wellbeing at work (NICE, 2022) recommends that employers offer mindfulness, yoga or meditation to all staff and states that such individual approaches should not replace action on the causes of stress at work; its committee found no evidence on long-term effects.