Exercise for Depression: What the Evidence Actually Says

Exercise has consistent, moderate-sized antidepressant effects in randomised trials. That is a real finding and it is worth taking seriously. It is also a finding that has well-known quality limits: most trials cannot blind participants and staff to which arm someone is in, so expectancy effects could inflate the observed benefits. Major clinical guidelines, including NICE in the UK and the American Psychological Association in the US, treat exercise as a useful option or adjunct rather than a replacement for talking therapy or medication, particularly for more severe depression.

This guide walks through what the recent evidence actually shows, what kinds and amounts of activity have been studied, the plausible biological and psychological mechanisms, and where exercise fits within a stepped-care approach. It is calm and educational, not a treatment plan for any individual.

The essentials at a glance

• The largest recent network meta-analysis (Noetel and colleagues, BMJ 2024) found moderate reductions in depressive symptoms across exercise modalities in 218 unique studies and 14,170 participants.
• That evidence base is rated low to very-low confidence by CINeMA, mainly because few studies blind participants and staff to the exercise arm, so expectancy effects could inflate the observed benefits.
• Walking or jogging, yoga, strength training, mixed aerobic exercise, and tai chi or qigong all showed moderate effects, with effects roughly proportional to the intensity prescribed.
• NICE (UK, NG222, 2022) lists group-based exercise as a treatment option in less severe depression; for more severe depression, treatments with greater therapist contact are considered first.
• APA (US, 2019) recommends psychotherapy or a second-generation antidepressant as initial treatment for adult depression. Exercise fits as a supportive, complementary option rather than a stand-alone substitute for moderate-to-severe presentations.

What the evidence actually says

The 2013 Cochrane review by Cooney and colleagues pooled 39 trials with 2,326 participants and found a moderate clinical effect of exercise on depression compared with no treatment or control. When the analysis was restricted to high-quality trials only, however, the effect shrank to small and was no longer statistically significant. The authors called explicitly for larger, higher-quality trials. That caveat became the standard caution about the exercise-for-depression literature for the rest of the decade.

Schuch and colleagues revisited the question in 2016 with an explicit publication-bias correction. Across 25 randomised controlled trials of exercise versus control, 9 of which were in major depressive disorder, the pooled standardised mean difference after trim-and-fill adjustment was 1.11 (95% CI 0.79 to 1.43), with a fail-safe N of 1,057. Even after correcting for the publication bias the field was worried about, the adjusted effect remained clinically meaningful.

The most recent and largest synthesis is the Noetel and colleagues 2024 network meta-analysis in the BMJ, covering 218 unique studies, 495 arms, and 14,170 participants. Against active control conditions such as usual care or attention placebo, the modality-level effects (Hedges' g, where more negative values indicate larger symptom reduction) were: walking or jogging -0.63 (95% CI -0.80 to -0.46), yoga -0.55 (-0.73 to -0.36), strength training -0.49 (-0.69 to -0.29), mixed aerobic exercise -0.43 (-0.61 to -0.25), and tai chi or qigong -0.42 (-0.65 to -0.21). CINeMA confidence in these estimates was rated low for walking or jogging and very low for the other modalities, driven primarily by the inability of trialists to blind participants and staff to the exercise arm. The Noetel authors are explicit that expectancy effects cannot be ruled out and that their review was not designed to comprehensively sample the psychotherapy and pharmacotherapy literatures, so cross-comparisons against those treatments inside the same network should not be read as head-to-head efficacy claims. A correction notice (BMJ 2024;385:q1024) refined the baseline-SD standardisation but preserved the modality ordering.

Taken together, the picture has improved since 2013, but the inherent unblindability of exercise interventions means the confidence rating stays low. The honest summary is moderate effects, low to very-low confidence, and an open question about how much of the effect is specific to exercise versus general expectancy and behavioural-activation effects shared with other active interventions.

Modality and dose

Five modalities were each linked to moderate antidepressant effects in Noetel and colleagues 2024: walking or jogging (g around -0.63), yoga (around -0.55), strength training (around -0.49), mixed aerobic exercise (around -0.43), and tai chi or qigong (around -0.42). The CINeMA confidence caveat above applies to every one of these point estimates: low for walking or jogging and very low for the others, driven by inadequate blinding rather than by inconsistent direction of effect. Within the same network analysis, effects scaled with the intensity prescribed (vigorous around g = -0.74, light around g = -0.58), suggesting that higher intensity is associated with a somewhat larger symptom reduction on average, while a wide range of intensities still produced benefit.

For a baseline of how much general physical activity adults are encouraged to do, the World Health Organization's 2020 Guidelines on Physical Activity and Sedentary Behaviour are useful. Adults aged 18 to 64 are advised to do at least 150 to 300 minutes per week of moderate-intensity aerobic activity, or 75 to 150 minutes of vigorous-intensity aerobic activity, or an equivalent combination, plus muscle-strengthening activities on at least 2 days per week involving all major muscle groups. The WHO also notes that physical activity confers mental health benefits, including reduced symptoms of anxiety and depression. Hitting at least the lower end of that range with a modality that is realistic to sustain is more practically useful than chasing a heroic dose that does not last beyond a few weeks.

Plausible mechanisms

The antidepressant effects of exercise appear to operate through several converging pathways rather than a single proven causal chain. Brain-derived neurotrophic factor (BDNF) acutely rises with aerobic exercise and is associated with hippocampal volume changes over weeks of training. Regular activity appears to dampen the cortisol hyper-response associated with chronic stress and depression, an HPA-axis effect that has been described across multiple physiological reviews. Regular activity is also associated with lower systemic low-grade inflammation, which has in turn been associated with depressive symptoms in observational and mechanistic work.

Alongside the biology, there is a strong behavioural pathway. Getting up, leaving the house, completing a planned activity, and often interacting with other people reverses the avoidance pattern characteristic of depression and plausibly contributes to rebuilding reward sensitivity and self-efficacy. This behavioural-activation component is shared with several evidence-based talking therapies, which is one reason exercise integrates well as an adjunct rather than a competitor. Specific numerical claims about BDNF dose-response or cortisol magnitude are intentionally left out here because the underlying literature is heterogeneous.

Exercise as an option or adjunct, not a substitute

The major clinical guidelines treat exercise as a useful component of depression care, not as a replacement for psychotherapy or pharmacotherapy. NICE guideline NG222 (UK, 2022) places group-based exercise inside the menu of treatment options for less severe depression: "a group physical activity intervention provided by a trained practitioner, more than one session per week for 10 weeks, including moderate-intensity aerobic exercise." For more severe depression, group exercise still appears in the menu, but NICE adds the explicit caveat that "the potential advantages of providing other treatment choices with more therapist contact should be carefully considered first."

The American Psychological Association's 2019 Clinical Practice Guideline for the Treatment of Depression Across Three Age Cohorts recommends, for initial treatment of adult depression, offering either psychotherapy (such as behavioural therapy, cognitive therapy, CBT, mindfulness-based cognitive therapy, interpersonal psychotherapy, or psychodynamic therapy) or a second-generation antidepressant (such as an SSRI, SNRI, or NDRI), with shared decision-making between clinician and patient. The APA does not single out exercise as a first-line monotherapy for adult major depressive disorder.

The plain summary across both guidelines is the same: if symptoms are moderate or severe, professional care is the priority, and exercise sits alongside that care rather than in place of it.

When to seek professional help

Persistent low mood for two weeks or more, loss of interest in usual activities, sleep or appetite disruption, difficulty functioning at work or in relationships, or any thoughts of self-harm or suicide are reasons to talk to a clinician. A primary care doctor is a reasonable first contact in most health systems and can refer onward to mental health services where needed. Speaking to a clinician earlier rather than later widens the range of effective options.

Practical starting points

A few general points are reasonably well supported. Picking a modality that fits the person's life and current capacity matters more than picking the modality with the marginally larger point estimate in a meta-analysis. Building up toward the WHO 150 to 300 minutes per week range gradually, rather than going from zero to a heroic schedule in a single week, tends to be more sustainable. Group-based or buddy-based exercise carries a behavioural-activation bonus on top of the physical dose, because it adds structure and social contact. Consistency across weeks beats occasional intense single sessions. For anyone already in clinical care for depression, mentioning a new exercise plan to the treating clinician is sensible, so the activity complements rather than collides with ongoing treatment.

Frequently asked questions

Can exercise replace antidepressants for depression?

For more severe depression, no. Both NICE NG222 (UK, 2022) and the APA 2019 Clinical Practice Guideline position psychotherapy and second-generation antidepressants as the initial treatment options, with exercise as a supportive complement. For less severe depression, NICE lists group-based exercise as one viable first-line option among several.

How much exercise is needed for a mood benefit?

The WHO 2020 guidelines set the adult baseline at 150 to 300 minutes per week of moderate-intensity aerobic activity, or 75 to 150 minutes of vigorous-intensity, plus muscle-strengthening on at least 2 days per week. The Noetel 2024 network meta-analysis found effects roughly proportional to the intensity prescribed, with a wide range of intensities still producing benefit on average.

Which type of exercise is best for depression?

The Noetel 2024 network meta-analysis found moderate antidepressant effects across walking or jogging, yoga, strength training, mixed aerobic exercise, and tai chi or qigong. CINeMA confidence in these estimates was low to very low, driven mainly by inadequate blinding. The most acceptable and sustainable modality for the individual usually beats the modality with the slightly larger point estimate.

Why is the evidence on exercise for depression rated low quality?

Few exercise trials successfully blind participants and staff, because it is hard to hide whether someone is in the exercise arm. Most studies are therefore rated at high risk of bias from expectancy effects. The effects survive a publication-bias correction (Schuch and colleagues, 2016) and replicate across modalities (Noetel and colleagues, 2024), but the CINeMA confidence ratings stay low for that reason.

Is it safe to start exercising while taking antidepressants?

For most adults this is consistent with general physical-activity guidance, but anyone in active clinical care for depression should mention a new exercise plan to the treating clinician so it complements rather than collides with ongoing treatment.

References

1. 1Effect of exercise for depression: systematic review and network meta-analysis of randomised controlled trials (Noetel M, Sanders T, Gallardo-Gomez D, et al., BMJ 2024;384:e075847, 14 February 2024) · BMJ (via PMC, PMC10870815). Accessed 2026-05-28.
2. 2Correction: Effect of exercise for depression (BMJ 2024;385:q1024, 28 May 2024) · BMJ (via PMC, PMC11131084). Accessed 2026-05-28.
3. 3Exercise for depression (Cooney GM, Dwan K, Greig CA, et al.), Cochrane Database of Systematic Reviews 2013, Issue 9, Art. No.: CD004366.pub6 · Cochrane Library. Accessed 2026-05-28.
4. 4Exercise as a treatment for depression: A meta-analysis adjusting for publication bias (Schuch FB, Vancampfort D, Richards J, Rosenbaum S, Ward PB, Stubbs B), Journal of Psychiatric Research 2016;77:42-51 · Elsevier / Journal of Psychiatric Research. Accessed 2026-05-28.
5. 5Depression in adults: treatment and management (NICE Guideline NG222, June 2022), Recommendations chapter · National Institute for Health and Care Excellence (UK). Accessed 2026-05-28.
6. 6Clinical Practice Guideline for the Treatment of Depression Across Three Age Cohorts (2019), Adults overview · American Psychological Association. Accessed 2026-05-28.
7. 7Guidelines on Physical Activity and Sedentary Behaviour (2020) · World Health Organization. Accessed 2026-05-28.
8. 8Physical activity (fact sheet) · World Health Organization. Accessed 2026-05-28.
9. 9988 Suicide & Crisis Lifeline · Vibrant Emotional Health / SAMHSA. Accessed 2026-05-28.
10. 10Samaritans: 116 123 (UK and Ireland), 24/7, free from landlines and mobiles · Samaritans. Accessed 2026-05-28.