Mood benefit peaked at 801 MET-minutes a week and stopped climbing
In short
In a meta-analysis of 31 randomised controlled trials (2,247 participants) on post-stroke depression, exercise produced a modest improvement in depressive symptoms (SMD -0.15, 95% CI -0.23 to -0.07, P<0.01), and restricted cubic spline modelling found the dose-response relationship was non-linear, with the greatest apparent benefit at about 801 MET-minutes per week. Subgroup effects were larger for resistance training (SMD -0.55), 1–2 sessions per week (SMD -0.41) and sessions of 30 minutes or less (SMD -0.44).
That exercise helps mood is a familiar claim. What this meta-analysis adds is how much — and the answer is not a straight line. Improvement peaked at roughly 801 MET-minutes per week and did not grow beyond it.
The authors searched PubMed, Web of Science, Embase, Scopus and the Cochrane Library and assembled 31 randomised controlled trials across 27 publications, 2,247 participants. The pooled effect was SMD -0.15 (95% CI -0.23 to -0.07, P<0.01) with moderate heterogeneity (I² 41.9%). Restricted cubic splines then mapped volume against symptom change as a curve rather than a slope.
How many minutes is 801 MET-minutes?
A MET-minute is intensity (METs) multiplied by time (minutes), so 801 converts differently depending on how hard the work is. At roughly 4 METs — moderate effort — that is a little over 200 minutes a week; at 8 METs it is around 100 minutes. Thirty minutes of walking five or six days a week lands there, and so does three 40-minute lifting sessions. The point of the paper is that this is not a large volume.
And nothing in this curve supports doing more than that for mood. That is what non-linear means in practice: the relationship between volume and mood was not one where more kept paying.
Why did less, shorter and rarer win?
The subgroup analyses point the same way as the curve. Larger effects were observed with 1–2 sessions per week (SMD -0.41, -0.67 to -0.15), sessions of 30 minutes or less (SMD -0.44, -0.70 to -0.19) and a 9–12 week intervention (SMD -0.16, -0.27 to -0.05). Not with more frequent or longer sessions.
The largest subgroup effect came from the modality. Resistance training: SMD -0.55 (-0.85 to -0.25), more than three times the pooled effect of -0.15. Short, infrequent, loaded sessions — which is the shape most lifting sessions already have.
Three things need stating honestly. First, the pooled effect is small at SMD -0.15 — the authors call it modest. Second, subgroup analyses are exploratory, so the -0.55 for resistance training should not be read as a settled prescription. Third, the population is post-stroke depression patients inside rehabilitation programmes. If you are being treated for depression or have a history of stroke, exercise is an addition to treatment rather than a replacement for it, and when and how hard to start is a decision to make with your clinician. Resistance training, anxiety and sleep are covered separately in resistance training for anxiety and sleep.
Where "more is better" breaks
Blood pressure, inflammation, mood — recent dose-response work keeps producing the same shape. Most of the improvement arrives between zero and a moderate dose, and above that the curve lies down. The same conclusion turned up in the exercise dose for blood pressure and the exercise dose for inflammation. For health markers, unlimited returns are the exception, not the default.
The useful fairlift point is that different goals have different optimal doses. The weekly volume that raises a Muscle Index score and the weekly volume that helps mood are not the same number. The second one is much smaller and saturates much earlier. So in a week where you have to cut volume, keeping one or two 30-minute sessions is well justified — that is exactly the size of the condition with the largest effect in this data.
The reverse also holds: this data does not support adding volume for mood's sake. If the curve flattens past 801 MET-minutes, training above that has to be justified by another goal — strength, hypertrophy, sport. Collapsing two purposes into one volume number means neither gets managed properly.
Frequently asked questions
How much exercise helps mood?
In an analysis of 31 randomised controlled trials (2,247 participants) on post-stroke depression, the greatest apparent benefit occurred at about 801 MET-minutes per week. At moderate intensity around 4 METs that is a little over 200 minutes a week; at 8 METs it is roughly 100 minutes.
Does more exercise help more?
Not in this data. Restricted cubic spline modelling found a non-linear relationship in which benefit peaked near 801 MET-minutes per week and did not increase above that point.
Which type of exercise had the largest effect?
Resistance training in the subgroup analysis, at SMD -0.55 (95% CI -0.85 to -0.25) — over three times the pooled effect of SMD -0.15. Subgroup analyses are exploratory, so this should not be read as a settled prescription.
How long and how often should sessions be?
Larger effects were observed with 1–2 sessions per week (SMD -0.41), sessions of 30 minutes or less (SMD -0.44) and an intervention lasting 9–12 weeks (SMD -0.16). More frequent or longer sessions did not produce better results.
Can this be applied directly to treating depression?
No. The population is post-stroke depression patients in rehabilitation, and the pooled effect is small at SMD -0.15. Exercise is an addition to treatment rather than a replacement, and timing and intensity should be set with a clinician.
Source: PubMed