Meditation: the evidence, sorted by how well it holds up

Meditation research has expanded enormously and the quality is uneven. Sorting the findings by how well they hold up is more useful than summarising the volume.

Reasonably well supported

Anxiety, depression and pain, modest effects. A widely cited systematic review commissioned by a US health agency examined mindfulness meditation programmes and found moderate evidence of small improvements in anxiety, depression and pain, and low or insufficient evidence for most other outcomes examined.

That is a measured conclusion from a rigorous review, and it remains a reasonable summary. Meditation programmes help somewhat with these outcomes, comparably to other active interventions, and they are not transformative.

Structured programmes have better evidence than informal practice. Most of the good research examines defined eight-week programmes with trained instructors, standardised curricula and daily home practice. Findings from these do not automatically transfer to using an app for five minutes.

Relapse prevention in recurrent depression. Mindfulness-based cognitive therapy has a reasonable evidence base for preventing relapse in people with a history of recurrent depression, and it appears in clinical guidelines in several countries for that specific indication.

Mixed or weak

Physiological outcomes. Blood pressure, immune markers, inflammatory measures. Studies exist, results are inconsistent, effect sizes where found are small, and reviews have generally been cautious.

Cognitive performance. Attention, memory and executive function. Some positive findings, considerable inconsistency, and concerns about small samples and inadequate control conditions.

Brain structure changes. A widely publicised area, and one where subsequent examination has been sobering. Several early findings have not replicated in larger studies, and reviews have raised concerns about sample sizes and analytical flexibility in neuroimaging research generally.

The methodological problems

Blinding is impossible. Participants know whether they are meditating.

Control conditions are frequently inadequate. Comparison against a waiting list measures the effect of doing something versus nothing, plus expectation, plus attention from instructors. Active control conditions matched for time, attention and expectation are much rarer and produce smaller effects when used.

Outcomes are usually self-reported, which is exactly where expectation effects operate.

Selection. People who volunteer for meditation studies are frequently people who expect it to help.

Publication bias, which reviews in this area have specifically identified.

None of this means meditation does nothing. It means the effect sizes in the literature are probably inflated and the true effects are likely smaller than headline figures.

The adverse effects nobody mentions

This is the under-reported part.

Research examining adverse effects has found that a minority of meditators experience unpleasant or distressing experiences, including anxiety, depersonalisation, re-experiencing of traumatic material, disturbed sleep and, in a small number of cases, more serious psychological difficulty.

A frequently cited review found that a meaningful proportion of participants in studies reported at least one unpleasant experience, and that most meditation research has not systematically recorded adverse events at all — which is a notable gap by the standards applied to any other intervention.

Risk appears higher with intensive practice — long retreats — and in people with a history of trauma or psychiatric illness.

The practical implication: anyone with a history of trauma, psychosis, bipolar disorder or severe anxiety should approach intensive practice with professional guidance rather than on their own, and any practice that consistently produces distress should be stopped and discussed with a clinician.

Meditation in the Ayurvedic context

The classical framework treats meditative and contemplative practice as part of a broader regimen, alongside diet, routine and conduct, rather than as a standalone intervention.

That is a different framing from the clinical trial model, and it is not testable in the same way. It is worth noting that the tradition itself does not present meditation as a treatment for a diagnosis.

What is reasonable

A short daily practice is low-cost, low-risk for most people, and has modest evidence for improvements in anxiety and mood.

That is a decent proposition and it should not be oversold. Meditation is not a treatment for a diagnosed mental health condition on its own, and it should not replace treatment that has better evidence.

If you are being treated for depression, anxiety or another condition, meditation may be a useful addition and should be discussed with the clinician managing your care rather than substituted for it.

Do not stop prescribed treatment for a mental health condition in favour of meditation. If a practice consistently produces distress, dissociation or disturbing experiences, stop and speak to a clinician.