Exercise as an intervention for depression gets a lot of enthusiastic coverage. The trial data supports it — with real, specific limits worth knowing before expecting too much from any one workout.
The evidence base
A 2025 systematic review and meta-analysis pooled 9 randomized controlled trials and 514 participants, looking specifically at high-intensity exercise and depression symptoms.1
The result
High-intensity exercise produced a statistically significant improvement in depression symptoms — an overall standardized mean difference of −0.23, and a more pronounced −0.44 on the Hamilton Rating Scale specifically.
Harder wasn't better
High intensity showed no significant advantage over moderate- or low-intensity exercise. The intensity of the workout, on its own, didn't predict a bigger benefit.
What did predict a bigger benefit
Duration and consistency did. Interventions running 12 weeks or longer outperformed shorter ones, aerobic exercise outperformed interval training, and benefits were strongest in adults 60 and older — the 30–60 age group saw a minimal effect in this analysis.1
Who this evidence best supports
The clearest signal is in adults 60 and older doing sustained aerobic work — that's the subgroup where the pooled benefit showed up most reliably. The 30–60 group is different: this analysis found a minimal effect there, which is a finding about this specific pool of high-intensity trials, not a verdict that exercise is useless for depression in your 30s or 40s. A handful of trials at one intensity band, split by age, is a thin basis for a strong claim either way.
The bigger point is the one we'd want any adjunct judged on: none of these trials tested exercise as a stand-in for clinical care. They tested it as an add-on, in people who were also being screened, monitored, or treated. That's the same adjunct-not-replacement logic we'd apply to any nutritional intervention under review — including the zinc-and-depression research we've covered elsewhere, even though zinc isn't a product in Proco's range yet — layered on top of proper treatment, not a substitute for it. If you're dealing with clinical depression, exercise is a lever to pull with your care team, not instead of one.
How this compares to the broader evidence base
A −0.23 to −0.44 effect sounds precise, so it's worth checking against the rest of the exercise-and-depression literature — and there's a lot of it. Cochrane's first major review pooled 35 trials and 1,356 participants and found a moderate overall effect, SMD −0.62. Restricted to only the trials with adequate allocation concealment, intention-to-treat analysis, and blinded outcome assessment, the pooled effect fell to −0.18 across 6 trials and 464 participants — no longer statistically significant.2 That's the single most important number here: a lot of the apparent benefit in the older literature came from lower-quality trials, and it shrank hard once bias controls tightened.
Cochrane's most recent update, published in 2026, added 35 new trials — 73 trials and roughly 5,000 participants in total. The primary comparison (57 trials, 2,189 participants) found a low-certainty SMD of −0.67 at end of treatment, similar effects to psychological therapy (moderate certainty, 10 trials), and comparable, low-certainty effects versus antidepressant medication. Restricted to the 7 highest-quality trials, the effect settled at −0.46. The authors noted that adding 35 new trials "had very little effect on the estimate of the benefit," and flagged that every included trial carries a high risk of performance bias — you can't blind someone to whether they're exercising.3
A −0.23 to −0.44 effect isn't an outlier — it sits at the smaller end of a literature that keeps finding the same story: real, positive, and modest, with a tendency to shrink further wherever the methods tighten.
Set against those numbers, the 2025 high-intensity-specific meta-analysis — −0.23 overall, −0.44 on the Hamilton scale — lands at or below the broader Cochrane range, not above it. Going harder didn't buy a bigger effect than exercise generally produces, and may have bought a smaller one. That lines up with the 2026 update's own finding that light-to-moderate intensity looked more helpful than vigorous exercise, and that 13 to 36 sessions was associated with the best results — consistent with this analysis's conclusion that duration and consistency mattered more than intensity.3
The 2025 paper has its own caveats worth naming. Egger's test found evidence of publication bias on the Hamilton Rating Scale outcome (p = 0.027), sensitivity analyses showed the pooled result could shift when individual studies were removed, and the authors were working with a small sample — 9 trials, 514 participants, 14 comparison groups. None of that erases the finding. It does mean "modest, and not the last word" is the right way to hold it — exactly how the study's own authors framed it.
Why exercise might help depression
The trials above measure an outcome without settling on a mechanism, and it's worth being honest about what's hypothesis versus what's established. A recent bibliometric review of the field maps a few candidate pathways that researchers keep circling back to, none of them proven as the mechanism.4
BDNF and neuroplasticity. Exercise appears to upregulate brain-derived neurotrophic factor, a protein tied to hippocampal neuroplasticity — the brain's capacity to form and reorganize connections. Lower BDNF has been associated with depression, and animal studies have been used to validate exercise's effect on BDNF levels specifically.4
Monoamine signaling. Earlier work — roughly 2006 to 2014, per the same review — focused on exercise's effects on monoamine transmitters, including serotonin (5-HT), the same broad neurotransmitter class targeted by SSRIs. It's one of the older, more established threads in the mechanism literature, running in parallel to the trials measuring clinical outcomes rather than growing out of them.4
Inflammation. Depression has repeatedly been linked to elevated inflammatory markers, and the same review flags exercise's effect on inflammatory cytokines as an emerging area — changes in cytokines after exercise are being explored as possible clinical markers, though this thread is newer and less developed than the BDNF and monoamine work.4
Behavioral activation and self-efficacy. Not every candidate pathway is biological. A study of community-dwelling older adults found that exercise-induced mood and exercise self-efficacy each independently predicted lower depressive symptoms — mood offering distraction from negative states, self-efficacy building confidence in managing daily tasks.5 That's a plausible piece of why the benefit above showed up most clearly in the 60-and-older group: a felt sense of mastery may matter as much as anything happening at the level of neurotransmitters.
None of this is settled biology — these are leading hypotheses under active study, not mechanisms proven to drive the clinical effect measured in the trials above. Mechanism research and outcome trials are separate literatures that don't always move in lockstep.
One more wrinkle: anxiety doesn't respond the same way
We've covered the exercise-and-anxiety literature separately, and the comparison is worth making explicit — it argues against treating "exercise" as one undifferentiated intervention. In the depression trials above, aerobic exercise came out ahead, with interval training weakest. In the anxiety literature we reviewed for Which Type of Exercise Actually Helps Anxiety, resistance training and mind-body practice were the standouts, not straight aerobic work. Same broad category, different top performer, depending on the outcome you're targeting — matching the modality to the condition looks like it matters more than defaulting to "exercise" as one interchangeable prescription.
What the researchers concluded
The study authors themselves called the improvements "modest" and explicitly called for larger trials — consistent with what the broader Cochrane literature keeps concluding, with confidence in the exact size varying by how tightly a given trial controlled for bias. That's the standard we're holding this to as well.
What we take from this
Exercise has a real, measurable effect on depression symptoms in the trial data — sustained and aerobic beats intense and short, and the effect shows up more reliably after 12 weeks than after a handful of sessions, most clearly in adults 60 and older. The mechanism is still an open question, with BDNF, monoamine signaling, inflammation, and behavioral self-efficacy all plausible contributors rather than a settled answer. Weighed against decades of larger meta-analyses, this effect size sits on the smaller, more conservative end — not an outlier claim, but not the strongest number in the field either. It's one part of a picture, evidence-backed but modest, layered onto standard care rather than standing in for it.
This article is for informational purposes only and is not medical advice. It is not intended to diagnose, treat, cure, or prevent depression or any other condition. If you're experiencing symptoms of depression, speak with a healthcare provider — exercise is not a substitute for clinical care.
References
- Frontiers in Public Health (2025). Systematic review and meta-analysis of high-intensity exercise and depression symptoms. frontiersin.org/journals/public-health/articles/10.3389/fpubh.2025.1616925/full
- Cooney, G.M., et al. (2013). Exercise for depression. Cochrane Database of Systematic Reviews. pubmed.ncbi.nlm.nih.gov/24026850
- Clegg, A.J., et al. (2026). Exercise for depression. Cochrane Database of Systematic Reviews. cochranelibrary.com/cdsr/doi/10.1002/14651858.CD004366.pub7/full
- Huang, D., Yang, H., Xu, J., & Hou, L. (2025). Neurobiological mechanisms of the effect of exercise on depressive disorder: analysis using CiteSpace. Frontiers in Psychiatry, 16, 1600286. frontiersin.org/journals/psychiatry/articles/10.3389/fpsyt.2025.1600286/full
- Miller, K.J., Mesagno, C., McLaren, S., Grace, F., Yates, M., & Gomez, R. (2019). Exercise, mood, self-efficacy, and social support as predictors of depressive symptoms in older adults: Direct and interaction effects. Frontiers in Psychology, 10, 2145. frontiersin.org/journals/psychology/articles/10.3389/fpsyg.2019.02145/full
Related reading
- Which Type of Exercise Actually Helps Anxiety
- Burnout and Cognitive Failure: What the Research Shows
Want to know what's actually in what you take? Join the list — no spam, just it straight.