General information, not medical advice. If a meditation practice is working for you, nothing here is a reason to stop. Nobody should stop prescribed medication or leave therapy on the basis of anything in this article.
In 2018, fifteen researchers published a paper in Perspectives on Psychological Science titled “Mind the Hype.” The author list included Willoughby Britton, Sara Lazar, David Vago and Clifford Saron — people who have spent their careers studying meditation, several of whom practise it.
The paper is a detailed indictment of the evidence base they helped build. And the most interesting thing about the eight years since is how little has changed in how meditation is discussed.
The definitional problem
The paper’s first point is that there is no universally accepted technical definition of mindfulness and no broad agreement on its components. In practice the word covers a ten-minute app session, an eight-week clinical programme with a trained instructor and group meetings, a three-month silent retreat, and compassion training — activities with little in common beyond the label.
This matters because findings do not transfer between them. Evidence for structured group-based programmes in clinical populations is routinely cited in the marketing for consumer apps. Those are different interventions tested on different people.
The authors also cite a stage audit of the field: roughly 30% of mindfulness-based intervention research had moved beyond early development, about 9% had reached efficacy testing against an active control in a research clinic, and about 1% had been conducted outside research settings at all.
The comparison-group problem
Most of the good news about meditation comes from trials that compared it to nothing — a waitlist. No class, no teacher, no group, no ritual, no weekly commitment.
Simon Goldberg and colleagues reviewed 142 randomised trials published between 2000 and 2016 and found that only 40.9% of comparisons used an active therapy control, and only 32.4% assessed and reported whether the intervention had actually been delivered as designed. They also looked for improvement over time in any quality indicator and found none.
The AHRQ-commissioned review by Madhav Goyal and colleagues, published in JAMA Internal Medicine in 2014, took the unusual step of restricting its meta-analysis to trials with active controls — 47 trials, 3,515 participants. Against non-specific active controls, mindfulness programmes produced standardised effects of roughly 0.2 to 0.4 for anxiety, depression and pain: real, modest, and roughly comparable to what antidepressants achieve in primary care. Against specific active treatments — drugs, exercise, other behavioural therapies — they found no evidence of superiority.
A literature too positive to be true
In 2016, Stephanie Coronado-Montoya and colleagues examined 124 mindfulness therapy trials in PLOS ONE. They found that 108 (87%) reported at least one positive outcome in the abstract and 109 (88%) concluded the intervention was effective.
Modelling the literature at a plausible effect size, the expected number of positive trials given the field’s statistical power was about 66. The observed number was 108 — roughly 1.6 times more positive results than the studies could support.
Only 21 of the 124 trials had any prior registration record, and none adequately pre-specified a single primary outcome. “Study after study shows” is being read as corroboration when it is a recognised signature of selective reporting.
The question the field stopped asking
The “Mind the Hype” authors noted that fewer than a quarter of meditation trials actively assess adverse events, relying instead on spontaneous reporting — against effectively universal active monitoring in drug trials.
Willoughby Britton built an instrument and asked. In a 2021 study in Clinical Psychological Science, an independent assessor administered a 44-item structured interview to 96 participants after three variants of an eight-week mindfulness-based cognitive therapy course, querying each item for occurrence, causal link, duration and impact on functioning.
83% reported at least one meditation-related side effect. Effects with negative valence occurred in 58%, and effects with a negative impact on functioning in 37%. Lasting negative effects occurred in 6–14%.
Britton’s own framing of this result is the one to keep: these are rates similar to other psychological treatments. The finding is not that meditation is dangerous. It is that meditation is an active intervention, and active interventions have side effects — which is exactly what the field had stopped checking.
Miguel Farias and colleagues reviewed 83 studies covering 6,703 meditators in Acta Psychiatrica Scandinavica in 2020 and found a pooled adverse event prevalence of 8.3% — but 3.7% in experimental studies against 33.2% in observational ones. The nine-fold gap between designs is itself the evidence about detection. The commonest categories were anxiety (33% of the studies reporting events), depression (27%) and cognitive anomalies (25%). Severe outcomes were the tail, not the centre.
A separate 2017 study by Jared Lindahl and colleagues interviewed practitioners about difficult meditation experiences in detail, and found substantial impairment and in some cases hospitalisation — but that study deliberately recruited people who had had difficult experiences, using outlier sampling and teacher referral, so its percentages describe severity within a selected group and are not prevalence figures for meditators generally. Quoted without that sentence attached, they are badly misleading, which is why they usually are.
What genuinely holds up
None of this is a case that meditation does not work. Two findings are among the strongest in behavioural medicine, and both come from structured, teacher-led programmes rather than apps.
Mindfulness-based cognitive therapy prevents depressive relapse. Willem Kuyken and colleagues pooled individual patient data on 1,258 patients with recurrent depression across nine trials, published in JAMA Psychiatry in 2016. Relapse over 60 weeks was 38% on MBCT against 49% in comparison arms — hazard ratio 0.69 against usual care, 0.79 against active treatments, and 0.77 against maintenance antidepressants. Patients with more residual symptoms benefited most.
Mindfulness-based stress reduction matched a first-line drug for anxiety. Elizabeth Hoge’s team randomised 276 adults with anxiety disorders to eight-week MBSR or escitalopram, reported in JAMA Psychiatry in 2023. MBSR met the prespecified non-inferiority margin. Study-related adverse events were reported by 78.6% of the drug arm against 15.4% of the MBSR arm, with ten withdrawals for adverse events in the drug group and none in MBSR.
Note what that second trial actually shows: both options have side effects. The honest comparison is between two active treatments, not between a drug and something inert.
Three questions worth keeping separate
Most of the confusion in this topic comes from collapsing three distinct claims.
Does it beat nothing? Reliably, yes.
Does it beat a plausible alternative? Sometimes — clearly for depressive relapse, credibly for anxiety — but the head-to-head evidence is thin and often underpowered. When a meta-analysis reports no difference against an active control, that frequently means too few studies to tell, not a demonstrated equivalence.
Is it harmless? No, and the field has not been measuring properly. Difficulties are usually transient and mild, and the actionable advice is straightforward: tell your teacher or clinician rather than pushing through, treat intensity as a dose, and if you have a trauma or psychosis history, practise with clinical support rather than alone with an app.
The story underneath
The reason this is worth reading is not that a wellness practice has been overhyped. It is that the people who built the evidence base published a detailed, specific, public critique of it — and the popular coverage did not change at all.
Eight years on, the result that genuinely holds up, MBCT cutting depressive relapse from 49% to 38% and outperforming maintenance antidepressants, remains among the least-reported findings in the field. The claims that dominate coverage are the ones resting on waitlist comparisons. And a discipline that quietly stopped asking whether its intervention has side effects had to be reminded by one of its own researchers to start again.
Sources
- Van Dam NT et al. Mind the Hype: A Critical Evaluation and Prescriptive Agenda for Research on Mindfulness and Meditation. Perspectives on Psychological Science 2018;13(1):36-61.
- Goyal M et al. Meditation Programs for Psychological Stress and Well-being. JAMA Internal Medicine 2014;174(3):357-368.
- Britton WB, Lindahl JR, Cooper DJ, Canby NK, Palitsky R. Defining and Measuring Meditation-Related Adverse Effects in Mindfulness-Based Programs. Clinical Psychological Science 2021;9(6):1185-1204.
- Farias M, Maraldi E, Wallenkampf KC, Lucchetti G. Adverse events in meditation practices and meditation-based therapies: a systematic review. Acta Psychiatrica Scandinavica 2020;142(5):374-393.
- Coronado-Montoya S et al. Reporting of Positive Results in Randomized Controlled Trials of Mindfulness-Based Mental Health Interventions. PLOS ONE 2016;11(4):e0153220.
- Kuyken W et al. Efficacy of Mindfulness-Based Cognitive Therapy in Prevention of Depressive Relapse. JAMA Psychiatry 2016;73(6):565-574.
- Hoge EA et al. Mindfulness-Based Stress Reduction vs Escitalopram for the Treatment of Adults With Anxiety Disorders. JAMA Psychiatry 2023;80(1):13-21.
