Effects of exercise interventions on EEG alpha power across different populations: a systematic review and
Fujie Liu1, Yifei Yang1, Yingying Cao1,2
1School of Physical Education, Shaanxi Normal University, Xi'an, China.
Objective:
The objective of this study was to systematically evaluate the to evaluate the effects of exercise interventions on electroencephalogram (EEG) alpha-band power across different populations.
Methods:
PubMed, Embase, Scopus, the Cochrane Library, and Web of Science were searched from inception to June 25, 2026. Randomized controlled trials comparing an exercise intervention with a control condition and reporting EEG alpha-band power were included. Standardized mean differences were calculated as Hedges' g. Random-effects models using restricted maximum likelihood estimation and Knapp-Hartung adjustment were applied. Heterogeneity, prediction intervals, subgroup effects, intervention-duration meta-regression, influence diagnostics, recording-condition sensitivity analyses, and small-study effects were examined.
Results:
Thirteen randomized controlled trials involving 497 participants were included. Exercise interventions were associated with greater EEG alpha-band power than control conditions (Hedges' g = 1.23, 95% CI [0.44, 2.02], p = 0.005), although heterogeneity was considerable (I2 = 91.56%, τ2 = 1.49) and the 95% prediction interval included the null value (-1.55 to 4.01). No significant moderation effects were observed for alpha-power metric, population type, or exercise paradigm, and intervention duration was not associated with effect size in chronic-exercise studies (β = 0.103, p = 0.142). Leave-one-out analyses yielded consistently positive results. The pooled estimate remained positive and statistically significant when the analysis was restricted to studies with clearly defined resting-state conditions (Hedges' g = 1.21, 95% CI [0.70, 1.72], p < 0.001) and reduced heterogeneity (I2 = 69.46%). Egger's test was not significant (p = 0.465), though publication bias could not be excluded. Confidence in the evidence was limited by risk-of-bias concerns, inconsistency, and imprecision.
Conclusion:
Exercise interventions were associated with greater EEG alpha-band power on average; however, the considerable heterogeneity and prediction interval crossing the null indicate substantial uncertainty across study settings.
Systematic Review Registration:
https://www.crd.york.ac.uk/PROSPERO/myprospero, CRD420251179061.
