Exercise as "medicine" for anxiety and depression: what the science actually says in 2026

For years, "get some exercise, it'll do you good" has been one of those pieces of advice given to almost everyone — vaguely true, but too generic to be genuinely useful in a clinical context. How much exercise? What type? Who does it work for, and who doesn't it help? In 2026, a series of high-level synthesis studies finally began answering these questions with a degree of rigor we didn't have before, shifting the status of physical exercise from "nice supporting advice" to an intervention with an evidence base comparable to already-established treatments.

What we thought we knew

The idea that physical exercise has a positive effect on mood isn't a recent discovery: dozens of studies over the past thirty years have documented associations between regular physical activity and reduced depressive and anxious symptoms. The problem was fragmentation: hundreds of small studies, often with different methodologies, heterogeneous populations (children, adults, cancer patients, people with chronic conditions), and not always consistent results. Existing meta-analyses tended to focus on specific subgroups or single exercise modalities, making it difficult to draw generalizable conclusions about what actually works, and for whom.

How the new studies were built

The 2026 leap in quality comes mainly from a paper published in the British Journal of Sports Medicine: a meta-meta-analysis — that is, a synthesis of syntheses, aggregating results from multiple already-published meta-analyses rather than starting over from individual primary studies. Munro and colleagues followed the PRIOR framework (Preferred Reporting Items for Overviews of Reviews) and included 63 studies, comprising 81 distinct meta-analyses, for a total of over 1,000 component studies and nearly 80,000 participants overall (Munro et al., 2026). Only randomized controlled trials were considered eligible: the most rigorous methodological standard for establishing a causal link.

In parallel, an Australian research group led by Ben Singh conducted a similar study focused specifically on children and adolescents, published in the Journal of the American Academy of Child & Adolescent Psychiatry, including 180 randomized trials on depression (over 34,000 participants) and 55 on anxiety (nearly 25,000 participants) (Singh et al., 2026).

Rounding out the picture, the Cochrane Collaboration published its 2026 update of its long-standing systematic review on exercise and depression, including 73 randomized trials with nearly 5,000 adults (Clegg et al., 2026) — an important update, since Cochrane reviews are generally considered the gold standard for methodological rigor and transparency.

What they found — and why it's surprising

The central finding is clear: physical exercise reduces both depressive and anxious symptoms, with effects ranging from moderate to robust. In Munro et al. (2026), the overall effect reaches an SMD (standardized mean difference, the standard measure for comparing effect size across different studies) of −0.61 for depression and −0.47 for anxiety — values that fall in the "moderate-to-large" range by common statistical conventions, and are comparable in magnitude to those reported for many standard psychotherapeutic interventions.

Perhaps even more surprising is the Cochrane finding: in direct comparison, physical exercise showed similar effectiveness to psychotherapy in treating depression, and a comparable effect even relative to antidepressant medication — though for that latter comparison the certainty of evidence was lower (Clegg et al., 2026). So we're not talking about a simple "added bonus," but about an intervention that, within the limits of available evidence, holds up against treatments we consider first-line.

A second notable point concerns which type of exercise works best, and for what. Aerobic activity (running, swimming, cycling, dancing) proved most effective overall on both fronts, but resistance training (weights) emerged as particularly effective specifically for anxious symptoms (Munro et al., 2026) — a finding that contradicts the widespread idea that only cardiovascular activity "releases" stress.

When the effect becomes real

An interesting finding also concerns program duration and structure. In the study on children and adolescents, shorter intervention programs (under 12 weeks) showed larger effects than those extended for longer (Singh et al., 2026) — a result that might seem counterintuitive, but is consistent with the idea that adherence and the novelty of the intervention play an important role in perceived effectiveness. Cochrane, for its part, notes that the most solid benefits are observed in those who complete between 13 and 36 sessions (Clegg et al., 2026): it's not enough to "exercise every now and then" — a minimum structure of consistency is needed for the effect to consolidate. In other words, physical exercise functions as a cumulative intervention, not a switch that flips on at the first gym session.

What these studies can't tell us

It should be said clearly that even the most rigorous syntheses inherit the limitations of the studies that compose them. Heterogeneity among the primary trials remains high across all three sources cited, meaning results vary considerably from study to study, and the aggregate estimates should be read as central tendencies rather than fixed values applicable to every individual case. Cochrane itself notes that many of the included trials remain small and at significant risk of bias, and that higher-quality studies are needed — a few large, well-conducted trials rather than many small ones — to establish with greater certainty which protocols work best, for whom, and whether benefits are maintained over time (Clegg et al., 2026). Furthermore, these works focus on depressive and anxious symptoms in general: they don't tell us whether or how the effect changes in the presence of more severe clinical presentations or complex psychiatric comorbidities, nor do they replace an ongoing treatment — they're an important piece of the therapeutic picture, not the only answer.

Clinical and social implications

For clinical practice, these findings offer a solid basis for including structured physical activity in care plans for anxiety and depression explicitly — not as a side suggestion, but as an intervention with its own evidence base — potentially calibrated: aerobic exercise if the primary goal is the depressive component, a mix including resistance work if the anxious component predominates.

On the social and public health front, the challenge becomes real accessibility: knowing that exercise works isn't enough if time, safe spaces, financial resources, or support to get started are lacking — especially for the most vulnerable populations, those in whom the burden of anxiety and depression is often highest. Integrating physical activity programs into mental health services, schools, and community settings today appears to be a direction with solid scientific grounding, rather than simply a good intention.

The article that prompted this analysis:

590.full.pdf

Download PDF • 1.14MB

References

Clegg, A. J., Hill, J. E., Mullin, D. S., Harris, C., Smith, C. J., Lightbody, C. E., Dwan, K., Cooney, G. M., Mead, G. E., & Watkins, C. L. (2026). Exercise for depression. Cochrane Database of Systematic Reviews, 1(1), CD004366. https://doi.org/10.1002/14651858.CD004366.pub7

Munro, N. R., Teague, S., Somoray, K., Simpson, A., Budden, T., Jackson, B., Rebar, A., & Dimmock, J. (2026). Effect of exercise on depression and anxiety symptoms: Systematic umbrella review with meta-meta-analysis. British Journal of Sports Medicine, 60(8), 590–599. https://doi.org/10.1136/bjsports-2025-110301

Singh, B., Bennett, H., Miatke, A., Dumuid, D., Curtis, R., Ferguson, T., Brinsley, J., Szeto, K., Eglitis, E., Zhou, M., Simpson, C. E. M., Petersen, J. M., Firth, J., & Maher, C. A. (2026). Systematic umbrella review and meta-meta-analysis: Effectiveness of physical activity in improving depression and anxiety in children and adolescents. Journal of the American Academy of Child & Adolescent Psychiatry, 65(2), 171–186.