Can moving help people with depression?

Meta-analysis of randomised trials. 14,170 people. trial interventions lasted at least 1 week (typical example ~10 weeks); searches up to 3 June 2023; outcomes measured right after the programme or up to ~6 months later.

Many trials, added up. In each one, chance decided who did what.

Who: 218 RCTs of people meeting clinical cut-offs for major depression (clinician diagnosed OR self-reported scores above a clinical threshold, e.g. >13 on the Beck Depression Inventory-II); includes people with physical comorbidities and postpartum depression; adolescents (14+), adults and older adults

• Walking or joggingHelpedfelt less low
• Yoga or weightsHelpedfelt less low

People knew they were moving.

More to know, from our read of the paper:

• Certainty is weak: only one of 218 studies met Cochrane criteria for low risk of bias; CINeMA confidence was LOW for walking/jogging and VERY LOW for every other treatment (including dance, yoga, strength training, SSRIs, CBT)
• Few trials blinded participants or staff, so effects could include expectancy effects (people expecting exercise to help)
• Effect sizes are against 'active controls' grouped together (usual care, placebo tablet, stretching, educational control, social support); only some of these are true placebos
• Dance (g -0.96, the biggest number) rests on 5 small studies (n=107), mostly young women (88% female, mean age 31); the authors explicitly say this 'prohibits us from recommending dance more strongly'
• Age/sex moderators (strength better for women/younger, yoga better for men/older) are study-level, not individual-level, and may be confounded

218 tests with 14,170 people with depression, added up (a network meta-analysis of randomised controlled trials). By chance, people got an exercise plan or other care, like usual care, stretching or a pretend pill. Walking or jogging, yoga and strength training all helped people feel less low, and harder exercise helped a bit more, though gentle exercise still helped. The tests were not very strong (low to very low certainty, and people knew they were exercising), and the authors say exercise could sit alongside therapy and medicine, not replace them. If you're struggling, please talk to a doctor.

We read the abstract and the full text of this paper and checked every number on this page against it (20 facts, checked on 2026-10-03). The study design is named above. Cohort and other observational studies are described as linked, not proven. Found an error? Email theis@onestudyaday.com.