STRENGTH TRAINING AND DEPRESSIVE SYMPTOMS: WHAT THE RANDOMIZED TRIALS MEASURED

Most of what gets written about strength training after 50 is about muscle, bone, and falls. A smaller but growing body of randomized trials has measured something else: depressive symptoms. Here's what the meta-analyses pooled, how large the effects were, how much confidence their own authors placed in them, and where the evidence for adults over 50 is thinner than the headline numbers suggest.

The Meta-Analysis That Started the Conversation

Gordon and colleagues published the first quantitative synthesis of resistance training and depressive symptoms in JAMA Psychiatry in 2018. They pooled 33 randomized clinical trials with 1,877 participants, comparing resistance training (947 people) with a nonactive control condition (930 people). Every trial used a validated measure of depressive symptoms.

Resistance training was associated with a significant reduction in depressive symptoms, with a moderate mean effect size of 0.66 (95% CI 0.48 to 0.83). The authors reported a number needed to treat of 4.

Three findings in that paper matter as much as the headline number:

• The effect did not depend on getting stronger. Whether a trial produced a significant strength gain was not significantly associated with the size of the reduction in depressive symptoms.
• It did not depend on volume or health status either. Total prescribed training volume, and whether participants were healthy or had a physical or mental illness, were not significantly associated with the effect.
• Better-designed trials found smaller effects. Trials that blinded allocation and/or assessment reported smaller reductions. Heterogeneity was high (I² of 76%), and the authors called for better-quality trials that blind both and compare resistance training against other established treatments.

That last point is the one to hold onto. The 0.66 is a pooled average across trials of mixed quality, and the authors themselves flagged that the more rigorous trials pointed lower.

The Larger Network Meta-Analysis

In 2024 Noetel and colleagues published a systematic review and network meta-analysis in the BMJ, and it answered a narrower question: what happens in people who meet clinical cut-offs for major depression? It included 218 unique studies, 495 trial arms, and 14,170 participants.

Compared with active controls such as usual care or a placebo tablet, the reductions in depression were:

• Walking or jogging: Hedges' g −0.62 (95% credible interval −0.80 to −0.45)
• Yoga: g −0.55 (−0.73 to −0.36)
• Strength training: g −0.49 (−0.69 to −0.29), from 22 studies and 643 participants
• Mixed aerobic exercise: g −0.43 (−0.61 to −0.24)
• Tai chi or qigong: g −0.42 (−0.65 to −0.21)

The authors reported that effects were proportional to the intensity prescribed, and that strength training and yoga appeared to be the most acceptable modalities, meaning participants tolerated them well.

The confidence ratings are the part the summaries usually leave out. Only one included study met the Cochrane criteria for low risk of bias. Using the CINeMA grading tool, the authors rated confidence as low for walking or jogging and very low for every other treatment, strength training included. The paper also carries a published correction (BMJ 2024;385:q1024). The figures above are from the current PubMed abstract.

The authors concluded that these forms of exercise could be considered alongside psychotherapy and antidepressants. They did not frame exercise as a replacement for either.

What About Adults Over 60?

Both of the large analyses above pooled adults of all ages. Two smaller meta-analyses looked specifically at people 60 and older, and their numbers disagree with each other by a wide margin.

Setayesh and Mohammad Rahimi (Geriatric Nursing, 2023) pooled 11 randomized controlled trials with 868 participants aged 60 or older. Resistance training was associated with improvements in depression (SMD −0.38, 95% CI −0.62 to −0.14) and a rise in circulating brain-derived neurotrophic factor, or BDNF (mean difference 0.73 ng/ml, 95% CI 0.04 to 1.42). The authors wrote that further research is needed to confirm the findings and investigate the mechanisms.

Khodadad Kashi and colleagues (Biological Research for Nursing, 2023) pooled 21 studies with 1,610 adults aged 60 or older and reported a much larger depression effect (SMD −1.13). The same analysis found no benefit for vitality, the physical component score, total quality-of-life score, or six-minute walk distance. The authors described their own results as "preliminary evidence" and wrote that more proof is needed to draw solid conclusions.

An effect of −0.38 in one analysis and −1.13 in another, both drawn from adults 60 and older, is a spread that says the size of the effect in this age group is not pinned down.

A 2026 network meta-analysis in Frontiers in Psychiatry (Qin and colleagues) compared 11 non-drug interventions across 83 trials and 6,646 adults aged 55 and older. For depression, dance ranked highest, and resistance training ranked second (SUCRA 67.5%), just ahead of cognitive behavioral therapy (65.6%). The authors cautioned that heterogeneity means the rankings should be interpreted cautiously, and they recommended a personalized, stepped-care approach. Their test for publication bias in the depression analysis was also significant (Egger's intercept 2.40, p = 0.029).

Anxiety, Briefly

Gordon's group ran a companion analysis on anxiety (Sports Medicine, 2017): 16 articles, 922 participants, mean age 43. Resistance training reduced anxiety symptoms with a smaller effect, Δ 0.31 (95% CI 0.17 to 0.44), and heterogeneity was low. Effects were larger in healthy participants (0.50) than in participants with a physical or mental illness (0.19). The mean age of 43 means this one tells us little about adults over 50 specifically.

What the Evidence Doesn't Cover

I searched PubMed for meta-analyses of resistance training and depression restricted to older adults and found the two described above. I could not find a trial in trained masters athletes, meaning people who already lift or train regularly, testing whether adding or changing resistance training shifts depressive symptoms. Nearly all of this evidence comes from people starting from little or no structured training.

None of these papers establishes how resistance training would produce the effect. The BDNF finding is an association measured in blood, not a demonstrated mechanism, and Gordon's analysis found the effect didn't track with strength gains at all.

Depression is a medical diagnosis. None of this research tells an individual reader whether they have it, how severe it is, or whether to change a treatment. Those are conversations for a physician or mental health professional, and the trials above were designed to add to care, not substitute for it.

The Protocol

• What the trials compared: supervised resistance training programs against nonactive controls, or against active controls such as usual care, in the larger network analysis.
• What the effect did not appear to depend on: in Gordon's 2018 analysis, total prescribed volume or whether participants got measurably stronger.
• Intensity: Noetel's analysis reported effects proportional to the intensity prescribed, across exercise types.
• Tolerability: strength training was among the best-tolerated modalities in that same analysis.
• Confidence: low to very low by the authors' own grading, with better-blinded trials pointing to smaller effects.

For the bodyweight strength movements WPP builds a program around, see Battle Hard After 50's strength chapter.

If you are having thoughts of harming yourself, call or text 988 in the United States to reach the Suicide & Crisis Lifeline. If you are being treated for depression, talk with your prescriber or clinician before changing anything about that treatment.

Battle Hard. — Will Power