Effects of Exercise Modalities on Multidimensional Balance Performance in Healthy Older Adults: A Systematic Review and Network Meta-Analysis of Randomised Controlled Trials

Background: Exercise improves balance in older adults, but which exercise modality is most effective for specific balance demands remains unclear. This study compared exercise modalities for overall and domain-specific balance performance in healthy older adults and examined whether effects varied by fall risk, setting, and dose. Methods: For this systematic review and Bayesian network meta-analysis (NMA), five databases were searched from inception to Dec 22, 2025. RCTs evaluating structured exercise in healthy adults aged ≥ 60 years with eligible balance outcomes were included. Screening, data extraction, and risk-of-bias assessment were performed independently in duplicate. Bayesian arm-based multilevel NMA estimated Hedges’ g with 95% credible intervals (CrIs); certainty was assessed via GRADE. PROSPERO: CRD420261292985. Findings: 423 RCTs (33 901 participants; 2546 effect sizes) were included. Compared with control, 18 of 19 modalities improved overall balance performance; stretching showed no clear benefit. The largest overall effects were for perturbation training (PERT; Hedges’ g 0·66, 95% CrI 0·49–0·82), sensory training (SEN; 0·55, 0·42–0·68), dance (DN; 0·54, 0·45–0·64), and agility stepping (AS; 0·53, 0·39–0·67). Modality effectiveness varied by domain: Tai Chi (TC) led for global balance, PERT for static balance, Pilates (PL) for dynamic balance, AS and aquatic exercise (AQ) for proactive balance, and balance-specific training (BS), AQ, and conventional resistance training (CRT) showed credible benefits for reactive balance. PERT showed a larger effect in participants at increased fall risk (1·58, 0·81–2·23), whereas SEN had the largest estimate in home-based programmes (0·88, 0·42–1·31). Dose-response patterns were non-linear and exploratory. Interpretation: Exercise improves balance performance in healthy older adults, but the preferred modality varies by balance domain and delivery context. PERT is most favourable for overall and static balance, especially with increased fall risk, whereas SEN suits home-based settings. Modality rankings and dose-response findings should be interpreted cautiously.Funding: National Key R&D Program of China (2024YFC3607301).