Exercise for mood: how much, and what kind

Less than the fitness industry implies, and almost any kind. Walking did as much for depression as any other exercise studied.² In new mothers, the one group where anyone pinned a dose down, about 80 minutes a week of brisk walking was enough to move symptoms.⁵ Otherwise, how much people were told to do barely mattered.⁴

What the trials measured

Nearly all of this evidence has one shape. People with a diagnosis or with raised symptom scores are assigned to an exercise program or to usual care, and they fill in a depression or anxiety questionnaire before and after.

The results come back as a standardized effect: the gap between the two groups, measured in units of how spread out the scores were. The scale is conventional. About 0.2 counts as small, 0.5 as moderate, 0.8 as large.

By that scale, exercise cut depression by 0.61 and anxiety by 0.47 across the broadest body of evidence available.¹ In adults with a diagnosis the numbers ran higher: 0.97 for depression, which the reviewers called large, and 0.66 for anxiety, which they called moderate.³

Effect size on depression and anxiety, broad evidence vs. diagnosed adultsstandardized mean difference

Depression, broadest evidence-0.61

Anxiety, broadest evidence-0.47

Depression, adults with a diagnosis-0.97

Anxiety, adults with a diagnosis-0.66

There is a hole running through all of it. You cannot blind a person to whether they have been exercising. Confidence in the estimates was rated low for walking or jogging and very low for every other type.² People who know they are in the active group tend to report feeling better, and none of these trials can separate that from the exercise.

What kind

One analysis compared the types head to head in people with major depression, which makes it the best guide there is.² Against active controls, walking or jogging came out on top, at a Hedges' g of -0.62 across 1,210 participants. The minus sign means symptoms fell. Yoga followed at -0.55 across 1,047, strength training at -0.49 across 643, mixed aerobic exercise at -0.43 across 1,286, and tai chi or qigong at -0.42 across 343.²

The ranking matters less than the spread. Every type landed in the moderate range, and the uncertainty around each is wide enough that the order could easily shuffle. Strength training and yoga were the two people stuck with most readily.²

The other reviews agree that type is secondary. In diagnosed adults, aerobic work, resistance work and a mixture of the two all improved symptoms, with none proving superior.³ The broadest review found aerobic exercise had the largest impact, and still concluded that exercise helps in every format tested.¹ The practical reading is that the kind you will keep doing is the right kind.

How much

Dose is where the evidence most contradicts the marketing. Across the strength-training trials the average effect on depression was 0.66, moderate by the scale above, and roughly four people had to train for one extra person to improve.⁴

Then the analysis asked what predicted that effect. The total volume of training prescribed did not. Nor did whether participants were healthy or ill. Nor did whether they actually got stronger.⁴ The mood benefit arrived whether the program was heavy or light, and whether or not it built any strength.

Only one review put a real number on the dose, and it studied women after childbirth. Exercise on its own cut depressive symptoms by 0.52 there, and lowered the odds of postpartum depression by 45%.⁵ For at least a moderate improvement, women needed about 350 MET-minutes a week, a unit that multiplies how hard by how long. The review translates it as roughly 80 minutes of moderate exercise: brisk walking, water aerobics, stationary cycling or resistance training.⁵

Eighty minutes a week is a little over ten minutes a day. In that population, it was enough.⁵

How hard

Intensity cuts both ways, depending on what is being treated. In people with major depression, the effect rose with the intensity prescribed, and walking, jogging, yoga and strength training all did best when done hard.² For anxiety, the broadest review found the reverse: shorter and gentler sessions were tied to the largest reductions.¹ In children and teenagers, mixed activity at moderate intensity did most for depression, while resistance work did most for anxiety.⁶

Put together: harder helps a little for adult depression, moderate suits young people, and for anxiety gentler is at least as good.^1,2,6 Intensity is a dial here, not a requirement.

Where it happens

The setting shows up in the data more clearly than the dose does. Group and supervised programs produced larger reductions in depression than unsupervised ones, and the biggest benefits landed in adults aged 18 to 30 and in women after childbirth.¹

In people with coronary artery disease, exercising in person cut depressive symptoms by 0.55 and anxiety by 1.16 against no exercise at all. Home-based programs did neither.⁷ The type of exercise made no difference to that result. The room did.⁷

That is one specific population, and the home programs may have failed on attendance rather than on biology. But it lines up with the supervision signal in the broader evidence.¹ Turning up somewhere, with other people, looks like part of the dose.

How fast

These trials cannot answer the question most people are actually asking, which is how a walk feels an hour later. They score symptoms at the end of a program that has run for weeks.

Inside that frame, longer was not better. In children and teenagers, programs under 12 weeks beat programs of 12 weeks or more for depression.⁶ In adults, shorter exercise was most strongly linked to falling anxiety.¹ Whatever exercise does for mood seems to show up inside the first three months, and stretching the program out did not add to it.⁶

What remains unknown

The effect is consistent. Its size is not. Trials that blinded allocation or assessment reported smaller reductions in depression than trials that did not, and almost no exercise trial is at low risk of bias.^2,4 The headline number may be too big, and nobody can say by how much.

The broadest review excluded people with chronic physical conditions.¹ That leaves out a great many of the people most often told to exercise for their mood.

Only the review in women after childbirth produced a dose threshold, and it belongs to that population.⁵ There is no equivalent number for anyone else.

Two questions matter most for anyone trying to use this, and both are open. Whether the benefit outlasts the supervised program. And whether exercise at home can be made to work as well as exercise in a room with other people.⁷