The evidence on meditation

Real modest benefits for anxiety, depression and pain — and no evidence it beats exercise, therapy or drugs, in a literature its own researchers have called overhyped.

What it is

The empirical question: what does meditation actually do, for whom, and compared with what?

This is the best-funded research programme covered by this site and one of the most methodologically troubled, and the trouble has been named most clearly by researchers inside the field rather than by its critics.

Where it came from

Serious clinical research began with the spread of MBSR from 1979 and accelerated enormously from the 2000s. Thousands of trials now exist, along with a large neuroimaging literature and substantial institutional investment.

How it is used

To justify clinical programmes, workplace and school interventions, and an app industry — and, more loosely, to lend scientific authority to contemplative claims generally.

What we can and cannot say

The most careful summary remains the 2014 systematic review by Madhav Goyal and colleagues, commissioned by the US Agency for Healthcare Research and Quality and published in JAMA Internal Medicine. It examined 47 trials with around 3,500 participants, restricted to studies with active control conditions — the design that actually tests whether a treatment does more than an equivalent amount of attention and effort.

Its findings:

Moderate evidence of improvement in anxiety, depression and pain
Low or insufficient evidence for stress, positive mood, attention, sleep, substance use and weight
No evidence that meditation programmes were better than active alternatives — exercise, medication, or behavioural therapies

That third point is the one most often dropped. The effect sizes for anxiety and depression are comparable to what antidepressants achieve in similar populations, which is a genuine result and is not a claim of superiority.

MBCT for depressive relapse is the standout, and it deserves its reputation. An individual patient data meta-analysis led by Willem Kuyken (2016) found that MBCT reduced relapse risk in recurrent depression comparably to maintenance antidepressant medication, with the benefit appearing greatest in those with more severe histories. That is a real clinical tool with a real evidence base, and it is the field's best result by some distance.

The field's own critique is the important document. In 2018 fifteen authors — including established meditation researchers — published "Mind the Hype" in Perspectives on Psychological Science, cataloguing the problems: inconsistent definitions of mindfulness, poor or absent active controls, small samples, weak blinding, inadequate reporting of adverse events, and public claims running far ahead of findings. When the researchers who built a field say this about it, an outside sceptic does not need to add much.

Publication bias is measurable here. Coronado-Montoya and colleagues (2016) examined mindfulness trials and found that the proportion reporting positive results was far higher than the statistical power of the studies could plausibly support — the signature of negative results not being published.

The neuroscience is weaker than its press. Early structural findings — Sara Lazar's 2005 report of increased cortical thickness in meditators, Britta Hölzel's 2011 grey-matter changes after eight weeks — were striking and had small samples. Larger and better-controlled replication attempts have produced mixed results, and several widely publicised structural claims have not held up. This is the ordinary trajectory of a young imaging literature and it is not a scandal; treating the early findings as settled was the error.

Adverse effects are real and were under-reported for years. A meta-analysis by Miguel Farias and colleagues (2020) found adverse events in roughly one in twelve participants — comparable to rates in psychotherapy, which is to say not alarming for a psychological intervention, but not the zero-risk profile meditation was marketed with. Trials frequently did not ask.

We can say the honest summary is this: meditation is a psychological intervention of modest, genuine benefit for some conditions, with a real clinical application in relapse prevention, no demonstrated advantage over other things that also work, real if uncommon adverse effects, and a research literature that overstated itself for two decades.

We cannot say what it does for the goals the traditions actually hold, because nobody is measuring those, as what meditation is sets out. A practice aimed at the end of craving is not refuted by a modest effect size on a depression inventory, and it is not vindicated by one either.

What it is

Where it came from

How it is used

What we can and cannot say

Further reading