Exercise modalities in mild cognitive impairment: a systematic review and network meta-analysis of comparative effectiveness

Background: Mild cognitive impairment (MCI) represents an intermediate stage between normal aging and dementia and is associated with an increased risk of progression to dementia. Exercise is a promising non-pharmacological intervention for cognitive decline; however, the comparative effectiveness of different exercise modalities across cognitive domains remains unclear. Objectives: This study aimed to compare the effects of different exercise modalities on domain-specific cognitive outcomes in patients with MCI and to identify optimal exercise prescription characteristics for the most effective intervention. Methods: This systematic review and network meta-analysis followed PRISMA guidelines and was preregistered in PROSPERO (CRD420251152983). Randomized controlled trials investigating exercise interventions in individuals with MCI were identified through database searches. Exercise modalities included aerobic exercise, resistance exercise, mind-body exercise, dance exercise, combined aerobic and resistance exercise, and multicomponent exercise, with passive and active controls as comparators. Standardized mean differences (SMDs) were calculated using change-from-baseline scores. A frequentist random-effects network meta-analysis was performed using Stata 17. Prespecified subgroup analyses examined exercise prescription characteristics for multicomponent exercise. Results: A total of 55 randomized controlled trials involving 4,603 participants were included. Network meta-analysis demonstrated that all exercise modalities were more effective than passive control in improving global cognitive function, with multicomponent exercise ranking highest (SUCRA = 83.3%). For memory-related outcomes, resistance exercise, mind-body exercise, dance exercise, and multicomponent exercise were superior to passive control, with mind-body exercise ranking highest (SUCRA = 83.9%). For executive-function-related outcomes, aerobic exercise, resistance exercise, dance exercise, combined aerobic and resistance exercise, and multicomponent exercise showed significant benefits compared with passive control, and resistance exercise ranked highest (SUCRA = 73.7%). Subgroup analyses indicated that moderate-intensity multicomponent exercise performed at least five times per week was associated with greater improvements in global cognitive function. Interventions lasting 13-24 weeks showed the largest effect estimates, although intervention period was not a statistically significant moderator. Conclusion: Different exercise modalities exhibit domain-specific cognitive benefits in individuals with MCI. Multicomponent exercise appears most effective for improving global cognitive function, mind-body exercise may preferentially benefit memory-related outcomes, and resistance exercise may provide greater benefits for executive-function-related outcomes. These findings support individualized selection of exercise modalities for individuals with MCI. Systematic review registration: https://www.crd.york.ac.uk/PROSPERO/view/CRD420251152983, Identifier: CRD420251152983.