Exercise and Therapy for Depression: What the 73-Trial Cochrane Review Says, and Doesn't
A January 2026 Cochrane review found exercise performing about as well as therapy for depression, with caveats. Here is what it means and a 13-to-36-session starter plan.
When you are low enough, "go for a walk" can sound like an insult. I have heard it said kindly and I have heard it land badly, and the difference was usually whether the person saying it had ever tried to put on shoes on a day like that.
So it is worth being careful about what the best available evidence actually says. In January 2026, a Cochrane review updated its long-running analysis of exercise for depression, and the headline traveled fast: exercise looks about as effective as psychological therapy. That sentence is mostly right, and it needs some fine print before you build a plan on it.
Why this review carries weight
Cochrane reviews are slow and strict on purpose. Reviewers search widely, apply the same inclusion rules to every study, rate how much each trial can be trusted, and say plainly how certain they are of the result. That is why a Cochrane update is worth more than a single trial that makes a splash.
This one, led by Andrew Clegg and colleagues, pools 73 randomized controlled trials with nearly 5,000 adults who have depression. It updates the 2008 and 2013 versions and adds 35 new trials since the last one. For a field that has been arguing about exercise and mood for decades, that scale matters. Earlier versions concluded the evidence was promising but shaky, and the new one has a lot more to stand on.
Against that, the fine print. In the comparison with no treatment, exercise showed a moderate effect on depressive symptoms, but the reviewers rate the certainty as low. The comparison with psychological therapy rests on about ten trials, and the one with medication on about five. Many of the individual trials were small. And when the analysis was limited to the better-designed trials, the apparent effect of exercise shrank. None of that erases the finding. It tells you the plausible size of the benefit is smaller and less certain than the headline sounds.
Why moderate intensity and 13 to 36 sessions mattered more than "exercise more"
The most practical result is a pattern about dose. As reported from the review, benefits were greatest with moderate-intensity exercise delivered over roughly 13 to 36 sessions. Not the most intense routines, and not a vague intention to be more active.
That has a believable logic to it. Moderate effort is hard enough to count and gentle enough to repeat. Very hard training can be a poor fit for a person with low energy and low motivation, who may quit after two sessions and conclude that exercise does not work for them. And a window of a dozen or more sessions says something about time: whatever helps seems to build over weeks, not after one good workout.
I would hold the exact numbers lightly. A range of 13 to 36 sessions comes from pooling studies that used different programs, so it describes what was tested, not a threshold the body respects. What I take from it is simpler. A modest, structured program, kept going for a couple of months, has a better record than sporadic bursts.
Exercise as a first-line option, not just an add-on
For years the polite position was that exercise is a nice extra alongside real treatment. This review pushes against that for some presentations. If a structured exercise program performs about as well as therapy in mild to moderate depression, then it deserves to be offered as a legitimate starting choice, discussed with a clinician, and not treated as a pep talk.
That shift matters for practical reasons too. Therapy has waiting lists and costs money. Some people cannot take medication or do not want to. A gym membership, a pair of walking shoes or a free class at a community center is, for many, the one option available this month. Having evidence that it counts is useful.
Why the benefit looked largest in young adults and postnatal women
A separate meta-analysis published in February 2026 reported the largest benefits in adults aged 18 to 30 and in postnatal women, with aerobic exercise showing the strongest effect. I have not read that paper in full, and I would treat these subgroup figures as a lead to follow, not a settled fact.
There are several reasons it could be true, and I am guessing at which matter. Younger adults often have more physical capacity and fewer barriers to starting. For women after childbirth, depression has a physical dimension of exhaustion and sleep loss, and a structured routine that also gets someone outdoors with a stroller can quietly address several things at once. As the parent of a young child, I can say that the hardest part of a postnatal routine is not the exercise, it is finding the forty minutes. If it works in that group, the practical implication is to make it small, social and flexible.
Important caveats
This does not mean exercise replaces therapy or medication for everyone. The evidence is strongest for mild to moderate depression. For moderate to severe depression, exercise is better thought of as part of a plan with professional care, not as a substitute for it. People with severe symptoms, a history of mania, or any thoughts of self-harm should be working with a clinician, not with a training schedule.
If you are having thoughts of ending your life, please contact a crisis line right now. In the United States you can call or text 988. If you are elsewhere, your local emergency number or a crisis line will connect you with someone who can help.
Check with your doctor before starting if you have a heart condition, a joint problem or any other reason to be cautious. And be kind about the days that go badly. A missed session is information about the day, not a verdict on the plan.
A starter plan modeled on the range the trials tested
This plan follows the dose pattern in the review, with details filled in by general exercise guidance, since the trials differed in their exact programs. Treat it as a draft to adjust with a clinician.
- Frequency. Three sessions a week. At that rate, 13 sessions take about four to five weeks, and 36 sessions take about twelve.
- Intensity. Moderate: your breathing is up but you can still speak in short sentences, and you could not sing. On a 10-point scale of effort, aim for about 5 or 6.
- Length. Start at 20 minutes if that is what you can do, and build toward 30 to 45. A short session you complete beats a long one you skip.
- Modality. Aerobic work showed the strongest effect in the subgroup reports, so build on brisk walking, cycling, swimming, light jogging or dance. Add two short strength sessions later if you like. Pick the one you will actually do.
- Company. If a friend or a class makes it more likely that you show up, use that. It is part of the intervention, not a distraction from it.
A way to pace it:
- Weeks 1 to 2 (sessions 1 to 6). 20 minutes, easy to moderate. The only goal is to arrive.
- Weeks 3 to 6 (sessions 7 to 18). 25 to 35 minutes at a steady moderate effort.
- Weeks 7 to 12 (sessions 19 to 36). 30 to 45 minutes, mixing in a longer weekend session if you enjoy it.
Write down your mood in one line after each session, for example a number from 1 to 10. Not for judgment, only so that after week six you have something better than memory to look at. Memory is a poor witness when you feel low.
FAQ
Does this mean I can skip therapy?
Not necessarily. Exercise performed about as well as therapy in these trials on average, but averages hide individuals, and most evidence is for mild to moderate depression. If you are already in therapy or on medication, exercise can sit alongside it. Do not stop treatment without talking to your clinician.
Is a walk enough, or does it have to be a workout?
Brisk walking at moderate effort counts as aerobic exercise. The key is the effort and the repetition, not the venue. If you can talk but not sing, you are in about the right zone.
What if I cannot get myself out of the door?
Shrink it until it is nearly silly: put on the shoes and walk to the end of the street. Many people find that starting is the hard part and that continuing is easier. If getting started feels impossible for weeks, that is itself worth telling a doctor.
How soon should I notice a difference?
The trials pooled in this review ran for weeks, and the dose pattern points to benefits building over a couple of months. Some people feel a small lift sooner, especially in sleep or energy. If nothing has shifted after about twelve weeks of steady effort, it is a good moment to ask for additional support.