Does meditation work, and what for?

What did the definitive review find?

Moderate evidence for three outcomes and not much else. Goyal and colleagues reviewed 41 randomised trials comparing meditation programmes against active controls, and found mindfulness meditation improved anxiety at an effect size of 0.40 at eight weeks, depression at 0.32, and pain at 0.33.

Those numbers are small. An effect size around 0.3 is a real difference you would struggle to notice in any individual person, and it accumulates across a population.

The comparison design is what makes this review strong. Trials measured meditation against an active control, meaning another activity requiring similar time and attention, which strips out the effect of simply doing something structured with a group.

What did not hold up?

More than most people expect. Evidence was low for improvement in stress and distress, and low for mental-health-related quality of life. For positive mood, attention and weight, the review found either low evidence of no effect or insufficient evidence to judge.

Attention deserves particular notice, since improved focus is among the most common claims made for meditation. In this review it sat in the insufficient category.

Mantra meditation, which includes transcendental meditation and similar practices, showed no effect on any of the psychological outcomes examined.

Is meditation better than the alternatives?

No, and the authors say so plainly. The comparative analyses found no evidence that meditation programmes were superior to any of the specific therapies they were tested against, including exercise and behavioural therapies.

That is worth holding onto. Meditation performs comparably to other active approaches on these outcomes, which makes it one reasonable option among several rather than the uniquely effective one it is often presented as.

Where that leaves you is practical. If you find meditation tolerable and other approaches unbearable, the evidence supports it. If you would rather run, the evidence does not say you are choosing worse.

Why does the popular picture differ so much?

Because much of the enthusiastic literature compares meditation against doing nothing. Against a waitlist, almost any structured practice looks effective, and part of what gets measured is attention, expectation and routine.

Effect sizes also faded with time. Anxiety fell from 0.40 at eight weeks to 0.22 at three to six months, and depression from 0.32 to 0.23. The benefit is real and it does not obviously compound.

What is worth doing?

If you want the version with evidence behind it, use a mindfulness-based programme for anxiety, low mood or pain, and expect a modest improvement over a couple of months. Those are the three outcomes the review supports.

Be sceptical of the wider claims. Better focus, higher positive mood and weight change are not supported at the same level, and a practice that helps your anxiety does not automatically deliver them.

If meditation is unpleasant for you, that is useful information rather than a failure. Comparable outcomes are available through approaches that performed just as well in the trials.