Domain-specific effects of exercise types on cognitive function in healthy older adults: A systematic review with pairwise and Bayesian network meta-analysis.

BACKGROUND: Exercise is increasingly recognized as a promising non-pharmacological strategy for supporting cognitive function in older adults. However, it remains unclear which exercise types are most beneficial for specific cognitive domains. This study compared the effects of different exercise types on cognitive function in cognitively healthy older adults. METHODS: From inception to March 23, 2025, we systematically searched the Web of Science, PubMed, Cochrane Library, SPORTDiscus, and PsycINFO databases, with an updated search performed on 26 June 2026. Eligible studies included healthy older adults without cognitive impairment (≥60 years). Risk of bias was assessed using the Risk of Bias 2 tool. Random-effects models were used to perform pairwise and Bayesian network meta-analyses. RESULTS: The network meta-analysis included 110 randomized controlled trials involving 8130 participants. Mind-body exercise was the most effective exercise type for executive function (SMD 0.54; 95% CrI 0.35, 0.73), processing speed (SMD 0.83; 95% CrI 0.45, 1.26), and attention (SMD 0.97; 95% CrI 0.40, 1.59). Aerobic exercise was most effective in improving global cognition (SMD 1.11; 95% CrI 0.58, 1.65) and language function (SMD 0.31; 95% CrI 0.10, 0.53). For memory function, high-intensity interval training ranked highest (SMD 0.58; 95% CrI 0.21, 0.96). CONCLUSIONS: The most effective exercise types varied across cognitive domains: aerobic exercise for global cognition and language function, mind-body exercise for executive function, processing speed, and attention, and high-intensity interval training for memory function. These findings may inform exercise-based strategies for cognitive health and healthy ageing.