We reviewed eight evidence syntheses and population datasets published between 2018 and 2024, among them an umbrella review covering 97 systematic reviews and 1,039 trials, then pooled their reported effect sizes into a single set of estimates for how exercise affects depression, anxiety, and psychological distress in adults. Together the sources represent roughly 1.65 million participant records, with some overlap among the trial-based reviews.

The question carries more weight in the fitness industry each year. In the American College of Sports Medicine’s Worldwide Fitness Trends survey, Exercise for Mental Health rose to #6 for 2026, up from #8 in both 2024 and 2025, and the same report found that 78% of exercisers name mental or emotional well-being as their primary reason for training. Survey rankings measure how widely professionals see a practice adopted; they say nothing about whether it works. That is what the clinical literature is for.

Our inclusion criteria favored meta-analyses of randomized trials for treatment effects and meta-analyses of prospective cohorts for prevention effects, supplemented by the largest cross-sectional dataset on exercise and self-reported mental health days. Because the syntheses differ enormously in size, we weighted each by the square root of its participant count, which keeps the largest umbrella review from mechanically overwhelming the smaller clinical meta-analyses. Effect sizes are reported as standardized mean differences (SMD), where 0.2 is conventionally small, 0.5 moderate, and 0.8 large, and we converted pooled effects into a modeled number needed to treat (NNT) so the figures read in plain terms. In the sections below, we break down the headline effects, how they vary by exercise type, and how much activity it takes to see them.

Key Exercise for Mental Health Statistics

In the table below, we break down our pooled estimates for each major mental health outcome alongside the range of values reported across the syntheses we reviewed.

The Key Exercise for Mental Health Statistics, 2026

Outcome	Published Range	Our Pooled Estimate	Modeled NNT or Equivalent
Depressive symptoms, general adult population	SMD −0.43 to −0.66	SMD −0.45	4
Depressive symptoms, clinical depression	SMD −0.42 to −0.95	SMD −0.63	3
Anxiety symptoms	SMD −0.41 to −0.42	SMD −0.42	4
Psychological distress	SMD −0.60	SMD −0.60	3
Risk of developing depression or anxiety, active vs. inactive adults	25% to 26% lower	25% lower	About 1 in 4 fewer new cases
Poor mental health days, exercisers vs. non-exercisers	1.49 fewer days per month	18 fewer days per year	43% lower burden

• Our pooled analysis places exercise in the moderate effect range for every symptom outcome, with the largest effects in adults who already meet clinical thresholds for depression.
• Across the studies we reviewed, roughly one in three adults with clinical depression who take up a structured exercise program is expected to see a meaningful symptom improvement they would not otherwise have had.
• Our model indicates that prevention and treatment effects are similar in scale: regularly active adults carry about a 25% lower risk of developing depression or anxiety in the first place.

Effect of Exercise on Depression by Exercise Type

In the table below, we break down the pooled antidepressant effect by exercise modality, combining network meta-analysis estimates against active controls with a separate meta-analysis of resistance training trials.

The Effect of Exercise on Depression by Exercise Type, 2026

Exercise Type	Pooled SMD	Effect Size Class	Modeled NNT	Notes From the Literature
Walking or jogging	−0.62	Moderate	3.0	Largest effect; greater at vigorous intensity
Strength training	−0.58	Moderate	3.2	Well tolerated; stronger in women and younger adults
Yoga	−0.55	Moderate	3.3	Well tolerated; stronger in men and older adults
Mixed aerobic exercise	−0.43	Small to moderate	4.2	Consistent across trial populations
Tai chi or qigong	−0.42	Small to moderate	4.3	Stronger in men

The strength training figure is the equal-weighted mean of two estimates that measured slightly different things: a network meta-analysis restricted to participants with major depression (g = −0.49) and a meta-analysis of 33 resistance training trials spanning healthy and clinical adults (Δ = 0.66). We kept both because neither alone represents the population a typical gym serves. The modality ranking should be read with care. Confidence ratings in the network meta-analysis were low for walking or jogging and very low for the other arms, largely because few trials blinded assessors, and the resistance training review found smaller effects in trials that did. What holds up across every analysis is direction and rough magnitude: each modality produced a clinically relevant effect, and the gap between the top and bottom of the table is smaller than the gap between any of them and no exercise. For older adults weighing yoga or tai chi, the same formats also carry a fall-prevention benefit we covered in Balance, Falls, and Training That Prevents Them.

Depression Risk by Weekly Activity Volume

In the table below, we break down modeled depression risk at different weekly activity volumes, expressed as minutes of brisk walking or its equivalent.

The Depression Risk by Weekly Activity Volume, 2026

Weekly Brisk Walking Equivalent	Activity Volume (mMET-h/wk)	Modeled Relative Risk	Modeled Risk Reduction
None	0.0	1.00	0%
30 minutes	1.8	0.91	9%
75 minutes	4.4	0.82	18%
150 minutes	8.8	0.75	25%
225 minutes	13.2	0.72	28%
300 minutes	17.6	0.71	29%
450 minutes	26.4	0.71	29%

We built this curve by fitting a saturating exponential model to the published dose-response points from a meta-analysis of 15 prospective cohorts covering 191,130 adults, which reported 18% lower risk at half the recommended activity volume and 25% lower risk at the full recommendation. The fitted model reproduces both anchor points exactly and plateaus near a 29% to 30% reduction, consistent with the original authors’ observation that gains beyond the guideline were minor. The shape matters more than any single row: the first 75 minutes a week deliver about 70% of the total available benefit. Cross-sectional data on 1.2 million Americans point the same way, with the lowest mental health burden among people exercising 30 to 60 minutes per session three to five times a week, and a worse burden than no exercise at all among those logging more than three hours a day. Since only a minority of adults reach the full aerobic guideline, as our activity guidelines report shows, the steep early section of this curve is where most of the population-level opportunity sits.

Exercise for Mental Health in the ACSM Fitness Trends Rankings

In the table below, we break down where Exercise for Mental Health has placed in ACSM’s Worldwide Fitness Trends survey since it entered the top 10.

The Exercise for Mental Health ACSM Rankings, 2026

Trend Year	Rank	Change	Context
2024	#8	Entered top 10	Ranked #1 in Australia’s regional survey
2025	#8	No change	Framed around stress, anxiety, and sleep
2026	#6	Up 2	78% of exercisers cite mental well-being as top motive
2027	#9	Down 3	Renamed Exercise for Mental Health and Well-Being

The rankings come from ACSM’s annual survey of fitness professionals, roughly 2,000 respondents for the 2026 edition and more than 1,400 for 2027, who rate about 50 candidate trends by how widely they see them adopted. The 2026 climb to #6 lined up with the trend placing in the top three among facility owners, program managers, and athletic trainers. The 2027 report, released October 20, 2026, moved it to #9 as group fitness classes and balance, flow, and core training surged past it, though the trend held a top 10 position for a fourth consecutive year. We read the dip as crowding from adjacent categories rather than waning interest: Pilates, yoga, and group classes are among the formats the clinical literature associates with mental health benefit, so some of that demand is now being counted under other names.

Request a copy of this report

If you’d like to request a PDF copy of this report or learn more about our company, you can reach out here.