The best evidence says mindfulness meditation programmes produce small to moderate improvements in anxiety, depression and pain, with effect sizes of about 0.2 to 0.4 in the 2014 JAMA Internal Medicine review of 47 trials, and they do not beat other active treatments such as exercise, drugs or therapy. Claims that eight weeks of practice rebuilds the brain failed to replicate in a 218-person trial in 2022, a large UK school programme showed no benefit for pupils, and roughly 8% of participants across meditation studies report an adverse effect. Mindfulness is useful; it is not a cure-all.
- 0.2 to 0.4 is the typical effect size against active controls: real, but small to moderate.
- No edge over drugs, exercise or therapy in the 47-trial Goyal review.
- 218 adults in the largest controlled MBSR scan study showed no grey-matter change after 8 weeks.
- 8,376 pupils in the MYRIAD school trial got no benefit at one or two years.
- 8.3% of meditators in 83 studies reported an adverse effect, mostly anxiety or low mood.
Few practices have gone from monastery to boardroom as fast as mindfulness. Apps, corporate programmes, school curricula and Instagram all sell it, often with a picture of a glowing brain.
In 2018, a group of researchers who mostly study meditation themselves published Mind the Hype in Perspectives on Psychological Science, warning that poor methods and loose claims risked harming the public. Eight years on, the data have grown, and they mostly back the sceptics on the big claims while supporting a modest core. This essay sets out what holds and what does not.
In the spring of 1979, Jon Kabat-Zinn, then working in the anatomy department of the young University of Massachusetts Medical School, was on a two-week vipassana retreat at the Insight Meditation Society in Barre, Massachusetts. Around day ten, sitting in his room one afternoon, he had what he later called a "vision" lasting maybe ten seconds: a plan to teach meditation and yoga to hospital patients who would never walk into a meditation centre.
Back at the hospital he met the heads of the primary care, pain and orthopaedic clinics and asked what share of their patients they felt they could help. The answer was typically 10 to 20 percent. They agreed to refer the rest to an eight-week course, and in the autumn of 1979 the programme that became MBSR began. It was built as a modest complement for people falling through the cracks, not a cure for everything.
Kabat-Zinn J: Some reflections on the origins of MBSR, skillful means, and the trouble with maps, Contemporary Buddhism (2011)How well does mindfulness work for anxiety and depression?
It works moderately well, about as well as other decent interventions. The landmark analysis is Madhav Goyal and colleagues' 2014 review in JAMA Internal Medicine, which screened 18,753 citations and kept only 47 randomised trials (3,515 participants) that used active controls to account for placebo effects.
Mindfulness programmes had moderate evidence of improving anxiety, depression and pain, and low evidence for stress and quality of life.
Johns Hopkins researchers searched nine databases to November 2012 and included only randomised trials with active controls, which rules out the waitlist comparisons that inflate many meditation results. From 47 trials and 3,515 adults, mindfulness programmes improved anxiety (0.38 at 8 weeks, 0.22 at 3 to 6 months), depression (0.30 and 0.23) and pain (0.33). Evidence was low or insufficient for positive mood, attention, substance use, eating, sleep and weight, and there was no evidence meditation beat drugs, exercise or other behavioural therapies. The limit: trials were small and varied, and the strict inclusion rule threw out much of the literature, which cuts both ways.
Read the paper (free)Why the comparison group changes the answer
Against no intervention at all, effects look bigger. A 2021 PLOS Medicine meta-analysis by Julieta Galante and colleagues pooled 136 trials with 11,605 adults in non-clinical settings such as workplaces and universities.
Compared with doing nothing, mindfulness programmes reduced anxiety, depression and distress by roughly half a standard deviation. But the 95% prediction intervals crossed zero (for anxiety, from -1.19 to 0.06), meaning that in some settings a programme might show no benefit at all.
| Comparison | What it controls for | Typical effect |
|---|---|---|
| Versus no intervention (Galante 2021) | Nothing: time, attention and expectation all count as "effect" | About 0.45 to 0.56 |
| Versus active control (Goyal 2014) | Attention, expectation and time spent on a programme | About 0.22 to 0.38 |
See the data as a table
| Effect size (SMD) | |
|---|---|
| Anxiety vs no intervention (Galante 2021) | 0.56 |
| Depression vs no intervention (Galante 2021) | 0.53 |
| Distress vs no intervention (Galante 2021) | 0.45 |
| Anxiety vs active control, 8 wk (Goyal 2014) | 0.38 |
| Pain vs active control (Goyal 2014) | 0.33 |
| Wellbeing vs no intervention (Galante 2021) | 0.33 |
| Depression vs active control, 8 wk (Goyal 2014) | 0.3 |
Source: Galante et al., PLOS Medicine 2021; Goyal et al., JAMA Internal Medicine 2014 · Standardised mean differences shown as positive numbers meaning improvement. Comparisons with no intervention give larger numbers than comparisons with active controls, which is the point.
Does meditation really change the brain in eight weeks?
The best test so far says no, at least not in brain structure. The Reddit thread linked above, built around a Washington Post interview with Harvard's Sara Lazar, reflects a story that spread widely after small studies around 2011 reported grey-matter changes after an 8-week mindfulness-based stress reduction (MBSR) course.
In 2022, Tammi Kral and colleagues at the University of Wisconsin combined two randomised trials with 218 meditation-naive adults:
- 75 did MBSR;
- 73 did a matched active programme;
- 70 waited.
Scans before and after found no evidence that MBSR changed grey matter volume, density or cortical thickness compared with either control, even in the regions earlier studies had highlighted.
What the null result does and does not mean
That does not mean meditation does nothing to the brain. Functional changes, such as quieter activity in the default mode network in long-term meditators, show up in several small studies.
It means that the most quoted claim, visible structural growth after two months, did not survive the largest controlled test. Our post on whether meditation changes the brain goes through the imaging literature in detail, and the replication crisis in psychology explains why small early studies so often shrink.

Does mindfulness work in schools?
Not as a universal programme, according to the largest trial ever run. The UK's MYRIAD trial randomised 84 secondary schools and 8,376 students aged 11 to 13 to either a school-based mindfulness curriculum or their usual social-emotional teaching.
At one year, there was no benefit for pupils on risk of depression, wellbeing or behaviour. Among students already at risk of mental health problems, scores were slightly worse with mindfulness, though the differences were small and not clinically relevant.
The two-year follow-up, published in BMJ Mental Health in 2026, broadly confirmed it: little advantage for pupils, with some lasting gains in teacher-reported wellbeing and school climate in one cohort.
The uptake problem
Part of the problem was uptake. A 2023 MYRIAD paper found that 4,232 students who received the programme practised outside class about once on average over the whole course. You cannot benefit from a practice you do not do.
For students, sleep and good study methods have firmer evidence; see our ranking of study techniques by evidence.
| Claim you will hear | What the evidence says | Key study |
|---|---|---|
| Mindfulness reduces anxiety and depression | Yes, modestly (effect sizes about 0.2 to 0.4 against active controls) | Goyal 2014, 47 RCTs |
| It works better than other treatments | No evidence it beats drugs, exercise or therapy | Goyal 2014 |
| Eight weeks grows grey matter | Not replicated in the largest controlled trial | Kral 2022, n = 218 |
| Teaching it in every school helps children | No benefit for pupils at 1 or 2 years | MYRIAD, 84 schools, 8,376 pupils |
| Meditation is harmless | Adverse effects in about 8% overall; higher in observational studies and long retreats | Farias 2020; Van Dam 2025 |
| Brief app sessions are risky | No more distressing than browsing the internet as usual | Simonsson 2026, n = 5,049 |

Can meditation be harmful?
Yes, for a minority. A 2020 systematic review in Acta Psychiatrica Scandinavica by Miguel Farias and colleagues found adverse effects reported in 55 of 83 studies covering 6,703 meditators.
The pooled prevalence was 8.3%, but it was 3.7% in experimental studies and 33.2% in observational ones, which look harder for problems. The most common were:
- anxiety;
- depression;
- unusual thinking or perception.
Some occurred in people with no previous mental health history.
The authors searched five databases to October 2019 and screened 6,742 citations, keeping 83 studies of meditation (not yoga postures) with 6,703 adult participants. Some 65% of studies reported at least one adverse event. Total prevalence was 8.3% (95% CI 5% to 12%), from 3.7% in experimental studies to 33.2% in observational ones. Anxiety (in 33% of studies with adverse events), depression (27%) and cognitive anomalies (25%) were most common; suicidal behaviour and gut problems were least frequent (11% each). The limit: most trials do not ask about harms systematically, so the true rate is uncertain in both directions.
Read the paperWho is most at risk, and who is not
Newer data refine the picture. In 2025, Nicholas Van Dam and colleagues surveyed nearly 900 American meditators chosen to reflect the national meditating population:
- about 60% reported at least one unusual effect from a 30-item checklist;
- about 30% found an effect challenging or distressing;
- 9% reported impaired functioning.
Distress in the 30 days before meditating and intensive silent retreats were risk factors.
At the other end, a 2026 randomised trial of 5,049 adults in the US and UK found that a single 5- or 10-minute digital practice was no more likely to cause a distressing experience than using the internet as usual. Ten-minute sessions were somewhat more likely to than 5-minute ones.
What is the "McMindfulness" critique?
The critique is that modern mindfulness was stripped of the ethical and philosophical frame it came from and repackaged as a productivity tool. In Buddhist practice, sati (mindfulness) is one limb of the Noble Eightfold Path, practised alongside right intention, right speech and right livelihood.
Corporate and app versions usually keep the attention technique and drop the rest. The word "McMindfulness", made familiar by Ronald Purser's 2019 book McMindfulness, captures the worry that a stressed worker is taught to breathe through a stressful job rather than question it.
A questionnaire score is not the same as a skill, and a skill is not the same as an eight-week course.
One word, four different things
Van Dam's Mind the Hype paper makes a related scientific point: "mindfulness" now names several things at once, and studies often measure one while claiming another:
- a trait (how mindful you tend to be);
- a state (how present you are right now);
- a set of practices;
- a range of programmes, from MBSR to apps.
The paper lists the movement's reach, from psychotherapy substitute to corporate wellbeing to training "more resilient soldiers", as the reason precision matters.
A team of contemplative scientists reviewed the state of mindfulness research, from its difficulty in defining mindfulness to methodological weaknesses in how it is measured and trained, possible adverse effects, and the over-reading of brain imaging. They set out a research agenda covering better measurement, properly controlled training studies, routine monitoring of adverse effects and caution with neuroimaging claims. The limit: a critical review and agenda rather than new data, written by researchers who still see value in the field.
Read the paper (free)So what can mindfulness actually do?
It can reduce anxiety, low mood and pain modestly, for people who practise. It works best as one tool among several, chosen because it suits you, not because it is magic.
- Expect modest effects. An effect size of 0.3 means a noticeable but not dramatic change for the average person. Judge it over 8 weeks, not 8 days.
- Compare against alternatives. Exercise, therapy and medication performed at least as well in Goyal's review. Pick what you will actually do.
- Start short. Five-minute sessions carried the lowest risk in the 2026 trial. Build up, and skip long silent retreats until you know how you respond.
- Choose by evidence, not branding. Our ranking of meditation techniques by evidence compares the main types, and our guide to starting meditation keeps it simple.
- Keep the frame. Mindfulness was taught as part of an ethical life. Attention training without any reflection on how you live is the version critics worry about.
None of this makes meditation less interesting. It makes it more honest, which is the spirit of everything we write in Manthan.
What's new
- Brief digital meditation no riskier than browsing as usual in 5,049-person RCT JMIR Mental Health, Sep 2026
- MYRIAD two-year follow-up: school mindfulness still no better than usual teaching for pupils BMJ Mental Health, 2026
- About 30% of US meditators report challenging effects, 9% impairment ScienceDaily, Nov 2025
Questions people ask
Is mindfulness scientifically proven?
It is supported for modest reductions in anxiety, depression and pain, based on dozens of randomised trials. It is not proven to beat other treatments, to change brain structure in eight weeks, or to help children as a universal school programme.
Did a Harvard study show meditation changes the brain in 8 weeks?
Small studies around 2011 reported grey-matter changes after an 8-week MBSR course. A 2022 trial with 218 people, including an active control group, found no structural change, so that claim is not established.
Can meditation make anxiety worse?
For some people, yes. Anxiety was the most common adverse effect in Farias's 2020 review. Recent distress and long intensive retreats raise the risk. Short sessions appear low-risk for most people.
Is mindfulness the same as Vipassana?
Modern mindfulness programmes such as MBSR borrowed from Buddhist insight practice, including Vipassana, but removed most of the religious and ethical teaching. A traditional 10-day Vipassana course is far more intensive than an app session; see our post on whether a Vipassana course is worth it.
Sources
- Goyal et al.: Meditation programs for psychological stress and well-being, JAMA Internal Medicine (2014)
- Galante et al.: Mindfulness-based programmes in nonclinical settings, PLOS Medicine (2021)
- Kral et al.: Absence of structural brain changes from MBSR, Science Advances (2022)
- Montero-Marin et al.: School-based mindfulness training in the MYRIAD trial, Evidence-Based Mental Health (2022)
- MYRIAD: Do adolescents like school-based mindfulness training? JAACAP (2023)
- MYRIAD two-year follow-up, BMJ Mental Health (2026)
- Farias et al.: Adverse events in meditation practices, Acta Psychiatrica Scandinavica (2020)
- Van Dam et al.: Mind the Hype, Perspectives on Psychological Science (2018)
- ScienceDaily: Van Dam et al., meditation-related adverse effects, Clinical Psychological Science (2025)
- Simonsson et al.: Adverse experiences in brief meditation practices, JMIR Mental Health (2026)
- Harvard Medical School: Harvard neuroscientist on meditation and the brain
- Kabat-Zinn J: Some reflections on the origins of MBSR, skillful means, and the trouble with maps, Contemporary Buddhism 12(1) (2011)

