Exercise for Depression and Anxiety: What the Head-to-Head Evidence Actually Says

You have probably heard some version of “exercise is as good as antidepressants.” It gets repeated in podcasts, gym captions, and the occasional doctor’s office. The claim is close enough to true to be useful and loose enough to mislead. So here is the honest version, built from the largest and most recent reviews rather than a single flashy study.

The short answer: regular exercise produces a real, measurable drop in depression and anxiety symptoms for most people who stick with it. The effect is roughly in the same ballpark as therapy and medication for mild-to-moderate cases. But the evidence has a known weakness, and “as good as a pill” is not the same as “a drop-in replacement for your pill.” Both things are true at once. Let’s walk through it.

What the biggest review found

The most comprehensive look to date is a 2023 umbrella review in the British Journal of Sports Medicine by Singh and colleagues. An umbrella review is a review of reviews, which is the closest thing science has to a bird’s-eye view. They pulled together 97 systematic reviews covering more than 1,000 trials and over 128,000 participants, and they applied strict rules so that the benefit could be pinned on the exercise itself rather than on the attention or social contact that often comes bundled with it.

Across all adult populations, physical activity produced a medium-sized reduction in depression (median standardized effect of about −0.43), a similar reduction in anxiety (about −0.42), and a slightly larger one for general psychological distress (about −0.60). In plain terms, a “standardized effect” of −0.4 to −0.6 means the average person who exercised ended up clearly better off than the average person who did not, by a margin you would actually notice in daily life. The authors put it bluntly: these reductions were “comparable to or slightly greater than” what you typically see from psychotherapy and medication, where the equivalent numbers often land in the −0.2 to −0.4 range. (Singh et al., BJSM 2023)

That is the headline. Now the details, because the details are where the practical advice lives.

Dose: harder and shorter beat gentle and endless

Two findings from the umbrella review surprise people.

First, intensity matters. Higher-intensity exercise produced larger benefits for depression (an effect near −0.70) than moderate (−0.56) or low-intensity movement (−0.22). This does not mean you have to sprint until you see colors. It means that effort appears to be part of the active ingredient, and that a brisk walk that leaves you slightly breathless likely beats a slow amble.

Second, and counterintuitively, shorter programs showed larger effects than longer ones. Interventions of 12 weeks or less landed around −0.84 for depression, while programs running 24 weeks or more dropped to about −0.28. (Singh et al., BJSM 2023)

Read that carefully, because it is easy to misinterpret. It almost certainly does not mean exercise stops working after three months. It more likely reflects two boring realities: people drop off as programs drag on, and the biggest mood gains tend to show up early. The takeaway is encouraging, not discouraging. You should expect to feel something within weeks, not be told to grind for half a year before anything happens.

Which type of exercise? The 2024 head-to-head

The umbrella review tells you that exercise works. A 2024 network meta-analysis in The BMJ by Noetel and colleagues went a step further and ranked the options against each other. It combined 218 trials with roughly 14,000 people, all focused specifically on depression. A network meta-analysis is the right tool here because it can compare treatments that were never tested directly against one another, by linking them through their shared comparisons.

Compared with usual care, the effects (measured as Hedges’ g, another standardized scale) lined up like this:

Walking or jogging: about −0.62
Yoga: about −0.55
Strength training: about −0.49
Mixed aerobic exercise: about −0.43
Tai chi or qigong: about −0.42

Every modality helped. The walking-and-jogging and strength-training results matter because they are cheap, require no special skill, and are the kinds of things a 40-year-old with a desk job can actually start on a Tuesday. The authors also noted that yoga and strength training were among the best-tolerated, meaning people were more likely to stay with them. (Noetel et al., BMJ 2024)

Same study, the part everyone quotes: effects tracked the intensity prescribed, reinforcing that vigor is doing real work. And exercise helped about equally whether or not people had other health problems, and regardless of how severe their depression was at the start.

So is it really “as good as a pill”?

Here is where the same BMJ analysis earns its keep, because it placed exercise, drugs, and therapy on one shared scale.

In that comparison, SSRI antidepressants on their own produced a smaller effect (about −0.26) than walking, jogging, yoga, or strength training. Cognitive behavioral therapy on its own came in around −0.55, right alongside the better forms of exercise. (Noetel et al., BMJ 2024)

Two cautions before you flush your prescription, which you should not do without talking to your prescriber.

First, these are indirect comparisons across very different study designs. Exercise trials are notoriously hard to “blind” — you always know whether you have been jogging — and people who sign up for an exercise study often expect to feel better, which inflates results. Drug trials use placebo pills and are far better at controlling that expectation effect, which tends to pull their measured numbers down. So a one-to-one ranking flatters exercise somewhat. It is a reason to take exercise seriously, not a reason to declare it the clear winner.

Second, you do not have to choose. The same analysis found that exercise combined with an SSRI (about −0.55), and aerobic exercise combined with psychotherapy (about −0.54), performed as well as the strongest single treatments. Movement is an add-on that costs little and can stack with whatever else you are doing. (Noetel et al., BMJ 2024)

What about anxiety specifically?

Most of the head-to-head firepower is aimed at depression, but anxiety has its own evidence. The umbrella review’s anxiety effect (around −0.42) sits right next to its depression number, so the general story holds: regular activity meaningfully lowers anxiety symptoms.

For the strength-training crowd, there is a dedicated meta-analysis. Gordon and colleagues, in Sports Medicine in 2017, pooled randomized trials of resistance exercise and found a small-to-moderate reduction in anxiety symptoms (about −0.31), in both healthy people and those managing chronic illness, regardless of sex. (Gordon et al., Sports Medicine 2017) Lifting is not just for depression or vanity; it has its own anti-anxiety signal.

Why it works: the mechanisms, minus the hand-waving

“Endorphins” is the answer everyone gives and the least interesting part of the real picture. The more substantive mechanisms, drawn from reviews of human and animal work, include a few converging threads.

BDNF. Exercise raises brain-derived neurotrophic factor, a protein that supports the growth and survival of neurons, especially in the hippocampus — a brain region that tends to be smaller in depression and that responds to antidepressant treatment. Contracting muscles even release signaling molecules that reach the brain and nudge BDNF production. (Phillips et al., Frontiers in Physiology 2023)
Inflammation and stress hormones. Regular activity lowers chronic low-grade inflammation and tends to reduce cortisol, both of which are linked to depression. (Kandola-adjacent overview, PubMed 2022)
Behavior and self-efficacy. Exercise is, structurally, the opposite of depression’s withdrawal-and-avoidance loop. It gets you out of the house, into a routine, and gives you small, repeated wins. That behavioral piece is not a consolation prize — it may be a large part of the effect.

The honest summary is that no single mechanism explains it. Several plausible pathways point the same direction, which is part of why the clinical effect is fairly robust even though the biology is not fully nailed down.

The catch you should know about

Now the part the gym captions leave out. The quality of the underlying trials is, on average, not great. In the umbrella review, the large majority of included reviews were rated “critically low” on the standard quality tool. In the BMJ network meta-analysis, confidence in the findings was rated low for walking and jogging and very low for most other modalities. (Noetel et al., BMJ 2024)

This is not a fringe complaint. The 2013 Cochrane review — Cochrane being the most conservative referee in medicine — concluded that exercise is moderately more effective than no treatment, but that when you restrict the analysis to the most rigorous, least-biased trials, the effect shrinks to something small. Cochrane also found exercise was no more effective than medication or psychological therapy, though that comparison rested on only a handful of small trials. (Cooney et al., Cochrane 2013)

So the spread of plausible truth runs from “moderately effective, on par with standard treatments” to “real but modest once you account for bias.” Both ends of that range still land on the same practical conclusion. The downside risk of trying exercise is close to zero, and the upside is a treatment-grade effect with a long list of side benefits no antidepressant offers.

What to actually do

If you strip the research down to instructions a busy person can follow:

Pick something you will repeat. Walking, jogging, lifting, and yoga all have the strongest evidence, and they all clear the bar. The best modality is the one you will still be doing in week six.
Push the intensity a little. Effort appears to be part of the medicine. Aim for “I’m slightly out of breath,” not “I’m strolling.”
Expect results in weeks, not months. The data suggest the early weeks carry the biggest mood payoff. If you feel nothing after a few weeks of genuine consistency, that is useful information to bring to a clinician.
Stack it, don’t swap it. Exercise works alongside therapy and medication, and the combination tested as well as anything. If you are already on treatment, add movement rather than replacing what is working.
Treat severe depression and anxiety as a both/and. The evidence is strongest for mild-to-moderate symptoms. If you are in a deep hole, exercise is a support, not a substitute for professional care, and the motivation to move is exactly the thing depression steals first. Lower the bar — a ten-minute walk counts.

The cleanest way to hold all of this: exercise is one of the few interventions with a credible claim to rival standard treatments for common depression and anxiety, it is free, and the worst realistic case is that you got fitter while finding out. That is an unusually good bet. Just don’t let the strength of the bet talk you out of the rest of your care.

Sources

Produced with AI under BRICK30’s editorial standards. For educational and informational purposes — not a substitute for professional medical or mental-health advice. See our Disclaimer.