The Universal Prescription
Close your eyes. Focus on your breath. Let your thoughts pass without judgment. This sequence is now the default recommendation for stress, anxiety, insomnia, chronic pain, and a dozen other conditions. The NHS prescribes mindfulness-based cognitive therapy to prevent depression relapse. Calm has been downloaded over 150 million times. Headspace runs corporate wellness programs for companies including Google, Adobe, and General Electric. The American Psychological Association lists mindfulness as an evidence-based treatment. The implicit promise is clear: meditation is safe for everyone, a practice with unlimited upside and no downside.
In 1977, the American Psychiatric Association published a position statement urging well-controlled studies to evaluate "the possible specific usefulness, indications, contraindications, and dangers of meditative techniques." It took 43 years for anyone to do it systematically.
What the Review Found
Miguel Farias and colleagues at Coventry University searched five medical databases for every study that reported adverse events during or after meditation. They found 6,742 citations. After screening, 83 met the inclusion criteria, covering 6,703 participants across experimental trials, observational studies, and case reports from 1974 to 2019.
Of those 83 studies, 55 reported at least one adverse event. That is 65% of studies, not 65% of participants. The pooled prevalence of adverse events across all study types was 8.3%, meaning roughly one in twelve people who meditate experience something harmful. The most common categories: anxiety appeared in 33% of studies reporting adverse events, depression in 27%, and cognitive anomalies (disorientation, confusion, false memories, attention problems) in 25%. Less frequent but present: gastrointestinal problems and suicidal behaviors, each appearing in 11% of the reporting studies.
The numbers shifted dramatically by study design. Randomized controlled trials reported a 3.7% adverse event rate. Observational studies reported 33.2%. The gap reflects how adverse events are measured, not how often they occur. Clinical trials are required to track only "serious adverse events" as defined by regulatory agencies, which means hospitalization, disability, or death. Anything that does not reach that threshold goes unrecorded. Observational studies, which simply ask people what happened, catch everything from mild distress to full-blown panic.
Who Gets Hurt
The review's most unsettling finding is negative: adverse events occurred in people with no prior psychiatric history. Meditation did not merely worsen existing conditions. It appeared to generate new ones. Case studies documented first-onset psychosis in previously healthy individuals following intensive meditation retreats. Experimental studies found elevated cortisol, increased blood pressure, and trauma re-experiencing in participants who entered studies screening as mentally well.
Willoughby Britton at Brown University spent a decade investigating this phenomenon. Her 2021 study used a structured 44-item assessment tool to interview 96 participants after completing an 8-week mindfulness-based cognitive therapy program. The numbers were stark. 83% reported at least one meditation-related side effect. 58% experienced something they described as negative. 37% reported that meditation impaired their daily functioning. And 6% to 14% described lasting bad effects, defined as functional impairment persisting more than one month. Dissociation predicted the worst outcomes: feeling emotionally flat or disconnected after meditation, which some practitioners initially experienced as relief, was the strongest predictor of enduring harm.
A 2022 population-based study by Goldberg and colleagues, published in Psychotherapy Research, surveyed a nationally representative sample of American adults and found that approximately 10% of those who meditated reported adverse effects. This confirmed Farias's clinical estimate in a real-world population.
The Strongest Case Against
The most rigorous defense of meditation comes from the same researchers who document its harms. Britton herself argues that adverse events are not a reason to abandon meditation any more than side effects are a reason to abandon surgery. The 8.3% adverse event rate matches what psychotherapy produces. SSRIs cause sexual dysfunction in 25–73% of users, weight gain in 55%, and emotional blunting in roughly 40–60%. By that standard, meditation is safer than the treatments it competes with.
Methodological critics note that the 8.3% figure pools together wildly different study designs, intervention lengths, and definitions of "adverse." A participant who reports temporary restlessness during a 10-minute guided session is counted alongside someone hospitalized after a 10-day silent retreat. The observational studies that drive the 33.2% rate rely on self-report and retrospective recall, both vulnerable to memory bias and the tendency to attribute any negative experience to whatever new practice preceded it.
Richard Davidson, whose lab at the University of Wisconsin-Madison has published some of the most-cited evidence for meditation's neural benefits, co-authored the Goldberg population study. His framing: "These findings point to the need for more systematic monitoring, not less meditation." The comparison to psychotherapy is instructive. Nobody calls therapy universally dangerous because some patients deteriorate. They screen for contraindications, train providers, and track outcomes. Meditation has not yet built that infrastructure.
What We Didn't Prove
The Farias review cannot establish causation. Systematic reviews pool existing studies; they do not control variables. A person who begins meditating while already developing depression may attribute the depression to meditation. The review also cannot identify which meditation technique causes which adverse effect, because 71% of included studies used mindfulness or mindfulness-based interventions, leaving loving-kindness, transcendental meditation, and Vipassana underrepresented.
The 95% confidence interval for the overall prevalence spans from 5% to 12%, which means the true rate could be as low as one in twenty or as high as one in eight. The gap between experimental (3.7%) and observational (33.2%) rates suggests that measurement methodology explains a large share of the variance. Studies that look harder find more adverse events, but studies that look harder may also define "adverse" more loosely.
Intensive retreats (more than five hours of daily practice) produce qualitatively different effects than app-guided 10-minute sessions, and the review does not separate dose-response relationships cleanly. Someone using Headspace for ten minutes before bed almost certainly faces different risks than someone sitting a 10-day Vipassana retreat in silence.
The Bottom Line
Meditation helps most people who try it. But the claim that it is universally safe is not supported by the evidence. About one in twelve meditators experience adverse effects, most commonly the same anxiety and depression that meditation is supposed to treat. The prevalence is comparable to psychotherapy's adverse event rate, and the finding has been replicated across clinical, experimental, and population-based samples. Meditation deserves the same informed consent, screening, and monitoring infrastructure that every other mental health intervention requires.
What You Can Do
Start with short, guided sessions rather than intensive silent retreats. The dose-response data consistently shows that adverse events cluster at higher intensities. If you have never meditated before, a 10-minute guided practice carries measurably less risk than a multi-day silent retreat with eight or more hours of daily practice.
Monitor yourself for signs that meditation is making things worse, not better. Increased anxiety, emotional numbness, dissociation, intrusive thoughts, or difficulty concentrating after sessions are signals to stop or change approach. Britton's research found that dissociation, which can feel like welcome relief at first, is the strongest predictor of lasting negative effects.
Tell your instructor or therapist if you notice adverse effects, because the research shows that most meditation teachers are not trained to identify or manage them. If you are taking a mindfulness-based program recommended by a clinician, ask whether they screen for trauma history and psychotic disorders, since both are established risk factors for meditation-related harm.
Do not assume that all meditation techniques carry the same risk profile. Open monitoring practices, which involve observing whatever arises in awareness without directing attention, appear to produce more challenging experiences than focused attention practices like breath counting. If one approach causes distress, switching techniques may resolve it without abandoning meditation entirely.