CATEGORY: EXERCISE & MOVEMENT
Exercise moderately reduced symptoms of depression—and worked about as well as talk therapy—in a Cochrane review that pooled 73 randomized trials involving nearly 5,000 adults diagnosed with depression.1 Compared with no treatment or an inactive control, exercise programs produced a moderate improvement in mood by the end of treatment, and in the head-to-head trials there was probably little to no difference between exercising and receiving psychological therapy.
Seventy-Three Coin Flips, Pooled Into One Answer
A Cochrane review is a systematic synthesis of randomized trials—studies in which people are assigned by chance to one treatment or another, the design that makes the groups alike in everything except the treatment itself and so allows a cause-and-effect conclusion. This one, updating a review last revised in 2013, gathered every eligible trial through November 2023 in which adults with depression were randomized to an exercise program versus no treatment, a waiting list, an inactive control, or an active alternative such as therapy or medication. Depression symptoms were measured on standard rating scales at the end of each program, and the reviewers graded how much confidence each pooled comparison deserves.
A Moderate Lift, and a Draw With the Talking Cure
Across 57 trials with 2,189 participants, people assigned to exercise finished with moderately lower depression scores than those assigned to no treatment or a control—though the reviewers rate the certainty of that estimate as low, largely because of weaknesses in how many of the trials were run. The most striking comparisons were the direct ones. In ten trials that pitted exercise against psychological therapy—approaches such as cognitive behavioural therapy, a structured talk therapy that teaches people to recognize and redirect unhelpful thought patterns—the two treatments came out essentially even, a finding the reviewers grade with moderate certainty. Five trials comparing exercise with antidepressant medication also found little to no difference, though that evidence is thinner and rated low certainty.
Where the Evidence Wears Thin
The review is as candid about weaknesses as it is about results. Nobody can be blinded to whether they are exercising, and most trials relied on participants rating their own mood—both of which can inflate apparent benefits. When the reviewers restricted the analysis to the seven strictest trials, the benefit over no treatment shrank, though it did not disappear. Few trials followed people after the programs ended, so the durability of the effect is uncertain. Unwanted effects, meanwhile, were uncommon on both sides of the ledger: occasional muscle and joint complaints among exercisers, while trial participants taking the antidepressant sertraline reported side effects such as fatigue and sexual dysfunction.
Thirty-Five New Trials Barely Moved the Estimate
Perhaps the most telling detail is one the authors volunteer themselves: adding 35 trials and more than 2,500 participants since the 2013 edition changed the estimate of exercise’s benefit very little. The answer has now held steady across more than a decade of accumulating research—unusual stability in a field where enthusiasm often outruns evidence. People in the exercise trials were also no more likely to drop out than those receiving other treatments, suggesting the prescription is one people can actually follow. For a condition as common as depression, the practical meaning is that exercise now sits on the evidence-backed menu alongside therapy and medication for symptom relief at the end of treatment—with the honest footnote that the head-to-head comparisons rest on a relatively small number of modest trials.
Key Takeaways
- Across 73 randomized trials and nearly 5,000 adults with depression, exercise moderately reduced symptoms compared with no treatment, though the certainty of that estimate is rated low.
- In direct comparisons, exercise performed about as well as psychological therapy (moderate certainty) and about as well as antidepressants (low certainty, based on a few small trials).
- The benefit was smaller—but still present—when only the most rigorously run trials were counted, and little is known about how long it lasts after a program ends.
- This update added 35 trials to the 2013 review and left the core estimate essentially unchanged, a sign of an unusually stable finding.
References
- Clegg AJ, Hill JE, Mullin DS, et al. Exercise for depression. Cochrane Database of Systematic Reviews. 2026;1(1):CD004366. link