The claim is everywhere: people who exercise more tend to report better mental health, so exercise must be the reason. Then researchers compared genetically identical twins — and the story became much less tidy.
Some twin studies suggest that part of the activity–depression link reflects shared familial or genetic influences. Yet randomized trials in people with depression show that exercise interventions can reduce symptoms. That is not a scientific contradiction. It is a lesson in asking the right question.
Quick answer
Can exercise help depressive symptoms?
Randomized evidence supports exercise as an effective intervention for reducing depressive symptoms in many people with depression. But twin research shows that the reason more active people often report fewer symptoms in everyday life cannot be reduced to exercise alone.
Key takeaways
- A Dutch twin-family study found no fewer anxious/depressive symptoms in the more active identical co-twin.
- A much larger international twin meta-analysis found within-pair associations between leisure activity and fewer depressive symptoms.
- An Australian pandemic twin study found weak within-pair evidence, suggesting familial factors mattered.
- A 2024 BMJ network meta-analysis of 218 studies found symptom reductions for several exercise modalities.
The twin result that challenged the simple story
In a Dutch sample of 5,952 twins plus siblings and parents, the more active member of an identical twin pair did not automatically report fewer anxious and depressive symptoms than the less active co-twin. Longitudinal changes in exercise also did not predict symptom reductions in that dataset.
The authors interpreted much of the ordinary population association as shared genetic influence. That is a powerful warning against reading every correlation as a treatment effect.
Then 31,596 twins added a twist
A meta-analysis of 11 twin studies — published online in November 2025 in a January 2026 journal issue — examined 31,596 like-sex twins aged 32 to 99. Physical leisure activity was partly heritable, but negative associations with depressive symptoms remained within twin pairs.
The authors described the findings as consistent with, but not proof of, a causal benefit of leisure activity. That wording is exactly the level of caution the result deserves.
Why randomized trials change the question
A randomized trial does something a population twin comparison does not: it actively assigns an intervention. A 2024 BMJ network meta-analysis included 218 studies and 14,170 participants meeting clinical cut-offs for major depression.
Walking or jogging, yoga, strength training and mixed aerobic exercise were among modalities associated with moderate symptom reductions compared with active controls. That supports treatment effects without proving that all population differences between active and inactive people are caused by exercise.
What looks like a contradiction is really triangulation
Twin studies ask whether an association survives tighter control for shared genes and family background. Randomized trials ask what happens when exercise is introduced as an intervention. Both questions matter.
The most useful conclusion is neither “exercise cures depression” nor “exercise does nothing”. It is that intervention evidence is meaningful while the natural relationship between activity and mental health is biologically and socially more complicated.
Explore how TwinPare approaches activity and recovery context
Source notes
The sources have been verified and editorially reviewed for this article. The limitations below show which level of conclusion the sources support.
- [de-moor-2008] Testing causality in the association between regular exercise and symptoms of anxiety and depression Marleen H. M. De Moor; Dorret I. Boomsma; Janine H. Stubbe; Gonneke Willemsen; Eco J. C. de Geus. Archives of General Psychiatry, 2008. Evidence type: Longitudinal genetically informative twin-family study Limitation: The population association weakened under genetically informative analyses. This does not show that exercise cannot improve symptoms when used as an intervention in people with depression. PubMed DOI
- [nygaard-2026] Leisure activity engagement and depressive symptoms: meta-analysis of 11 studies of middle-aged and older twins aged 32-99 Marianne Nygaard; Miriam A. Mosing; Deborah G. Finkel; Margaret Gatz; Matt McGue; IGEMS Consortium. The Journals of Gerontology: Series A, 2026. Evidence type: Meta-analysis of 11 twin studies with co-twin-control analyses Limitation: Published online 11 November 2025 in Volume 81, Issue 1 (January 2026). Within-pair associations are consistent with, but do not prove, causal effects of leisure activity on depressive symptoms. Oxford Academic DOI
- [caruana-2024] Physical activity and depression during the COVID-19 pandemic in Australia: A twin study Julia M. Caruana; Jesse T. Young; Jessica Tyler; Paulo H. Ferreira; Sue Malta; John L. Hopper; Lucas Calais-Ferreira. Mental Health and Physical Activity, 2024. Evidence type: Adult twin study using within-individual and within-pair regression during the COVID-19 pandemic Limitation: Self-reported activity, pandemic context and weak within-pair evidence limit causal interpretation and generalisation. DOI DOI
- [noetel-2024] Effect of exercise for depression: systematic review and network meta-analysis of randomised controlled trials Michael Noetel et al.. BMJ, 2024. Evidence type: Systematic review and network meta-analysis of randomized exercise trials Limitation: Participants met clinical cut-offs for major depression. Intervention effects should not be confused with explanations for observational population associations, and exercise should not be framed as a replacement for clinical care. PubMed PMC DOI
Editorial source review
This section shows how the article's key factual claims are linked to the source.
Phrasings that require caution
- Do not say exercise prevents depression based on these sources.
- Do not say twin studies prove exercise has no causal effect.
- Keep treatment evidence distinct from general-population associations.
- Do not imply exercise replaces psychotherapy, medication or other clinical care.
| ID | Claim | Source support | Caution |
|---|---|---|---|
| D2-C1 | In the Dutch genetically informative study, the more active identical co-twin did not show fewer anxious/depressive symptoms. | 2008 | Use the study's population and outcome in the phrasing. |
| D2-C2 | The 31,596-twin meta-analysis found significant negative within-pair associations between depressive symptoms and leisure-activity domains. | 2026 | Consistent with but not proof of causality. |
| D2-C3 | The BMJ review included 218 unique studies and 14,170 participants and found symptom reductions for several exercise modalities. | 2024 | Use the study's population and outcome in the phrasing. |