"Can a Run Really Match Therapy or Antidepressants?"

For anyone in the doldrums, few prescriptions feel more tiresome than the brisk instruction to exercise more, yet two large analyses published in early 2026 give that familiar advice unexpected scientific weight. Both papers suggest that regular physical activity reduces the symptoms of depression and anxiety by roughly as much as a course of therapy or a prescription of antidepressants, a striking parity that has spread quickly through the health press. The experts who produced the work, however, have reservations of their own about how far that comparison can really be pushed.

The first study, published in January by researchers based across Britain and Ireland, took the form of a Cochrane review, a well-regarded meta-analysis of health-care research that pooled the results of 69 randomised controlled trials measuring exercise's effect on depression. The second, appearing the following month in the British Journal of Sports Medicine, was a so-called meta-meta-analysis that drew on more than 1,000 trials involving close to 80,000 participants in total. Each team concluded, in the cautious phrasing typical of the genre, that exercise reduces depressive and anxious symptoms by roughly as much as the conventional treatments doctors usually prescribe.

There are, inevitably, important caveats. Meta-analyses can only be as reliable as the trials they pool, and exercise studies are particularly prone to being skewed by the way they have to be designed. Participants cannot be blinded, since they obviously know whether or not they are doing kettlebell swings, which makes their self-reported mood especially vulnerable to whatever favourable expectations they happen to bring into the study. For this and related reasons, the Cochrane authors judged every trial they included to carry a high risk of bias.

The meta-meta-analysis introduces a second, subtler problem. It did not test exercise against other interventions directly; instead, its exercise results were stacked against the outcomes of separate trials of therapy and antidepressants. Drug trials, unlike exercise trials, are typically well blinded and produce strong placebo effects, which makes it harder for them to register a clean advantage over the comparison condition. "I don't think it's a fair comparison," says Jonathan Roiser, a professor of neuroscience at University College London.

For all that, the practical signal is reasonably consistent. Aerobic workouts such as running, walking and cycling appear to be beneficial across the board, while group-based or supervised sessions tend to outperform solitary effort when depression is the target, with benefits that accrue gradually over several months of regular practice. For anxiety, by contrast, lower-intensity activity seems to deliver the best results.

Why exercise works on mood is harder to pin down. The popular notion that it generates a chemical high by triggering a release of endorphins, a class of opioid produced by the body itself, has surprisingly little scientific support. A 2021 study found that blocking runners' opioid receptors reduced neither the euphoria they reported after a session nor the subsequent drop in anxiety. Researchers now suspect that endocannabinoids, chemicals produced by the body and brain that bind to the same receptors as the active molecules in cannabis, are likelier to be responsible for these short-term boosts.

Several other pathways also seem to be engaged. Exercise appears to dampen inflammation, improve brain plasticity, and increase the transmission of dopamine, the neurotransmitter that the brain uses to weigh effort against reward; lifting that transmission may help reverse the loss of motivation that defines clinical depression. Psychological benefits accrue on top: regular activity offers a tangible sense of achievement, agency and, in time, mastery, all of which are known to lift mood. Plenty of reasons, in short, to work up a sweat.