This review pooled 33 trials of exercise in children and adolescents with a formal ADHD diagnosis, mostly aged 8 to 14. Ten trials tested a single exercise session and 23 tested longer programs. Studies of children with autism, Tourette or epilepsy were left out. About 36% of studies reported that children were also taking medication.
The authors looked at how the amount of exercise related to the size of the benefit. Exercise improved three thinking skills: switching between tasks, holding information in mind and holding back impulses. The pattern was not a straight line, so more exercise was not always better.
For a single session, only the gain in holding back impulses was clear. The other two were too uncertain to rule out no effect. For longer programs, all three skills showed clear gains. Swimming, then ball games, did best for task switching. Martial arts, then rope skipping, did best for memory. Running did best for impulse control.
Quality was rated low for memory and task switching and moderate for impulse control. Most trials did not fully report exercise doses, so doses were estimated. A test hinted at possible publication bias for long-term task switching, though a correction left the result stable. Most programs took place in a lab.
The authors call their dose findings preliminary and a starting point for future exercise programs for children and adolescents with ADHD.
Talk to a clinician about your own or your child's situation.
The numbers Stats and what the terms mean
The statistics behind the plain-language claims above.
- Pan, 2025: task switching after a single session at 270 METs, SMD = 0.48 (95% CrI = −0.43, 1.36).
- Pan, 2025: holding information in mind after a single session at 170 METs, SMD = 1.27 (95% CrI = −1.61, 3.89).
- Pan, 2025: holding back impulses after a single session at 130 METs, SMD = 1.27 (95% CrI = 0.28, 1.67).
- Pan, 2025: task switching with long-term exercise at 1100 METs per week, SMD = 1.20 (95% CrI = 0.21, 2.09).
- Pan, 2025: holding information in mind with long-term exercise at 1300 METs per week, SMD = 0.82 (95% CrI = 0.08, 1.62).
- Pan, 2025: holding back impulses with long-term exercise at 2500 METs per week, SMD = 0.74 (95% CrI = 0.21, 1.29).
- Pan, 2025: test for possible publication bias in long-term task switching, Egger's test p < 0.01.
- Pan, 2025: the same test for long-term memory and impulse control, Egger's tests p > 0.1. What the terms mean:
- METs A measure of exercise dose, here estimated from how hard, how often and how long children exercised.
- SMD The size of the difference between groups in standard units. Roughly, 0.2 is small, 0.5 is moderate and 0.8 is large.
- 95% CrI The range the true effect most likely falls in. When it includes zero, the result is too uncertain to rule out no effect.
- p How likely a result this strong would be by chance alone. Smaller means less likely.
- Egger's test A check for whether the pattern of published studies looks lopsided, a sign of publication bias.
Source: Pan Q, Zheng S, He P. Beyond binary comparisons: a Bayesian dose-response meta-analysis of exercise on executive function in children and adolescents with ADHD. Pediatric Research (2025). doi:10.1038/s41390-025-04325-1