Comparative efficacy of exercise interventions for fall prevention in older adults with cognitive impairment: protocol for a systematic review and network meta-analysis

Jinsheng Lu, Shaojuan Huang, Yu Zeng, Danqin Li, Shaoying Song, S Y Chen, Jingli Chen, Q J He, Jinyue Huang, Yao Hu

Introduction Older adults with mild cognitive impairment (MCI) or dementia have a substantially elevated risk of falls compared with cognitively intact peers, and the consequences of falls in this population are more severe. Multiple exercise modalities, including aerobic exercise, resistance training, balance training, mind-body exercise, multicomponent programs, dual-task training, and exergaming, have been investigated for fall prevention, yet the comparative efficacy of these interventions in cognitively impaired populations remains unclear. Existing evidence relies on pairwise meta-analyses that compare only two interventions at a time, limiting the ability to determine which exercise type is most effective. We will conduct a systematic review and network meta-analysis (NMA) to compare the efficacy of different exercise interventions for fall prevention in older adults with MCI or dementia. Methods We will include randomized controlled trials (RCTs) comparing any structured exercise intervention (aerobic exercise, resistance training, balance training, mind-body exercise, multicomponent exercise, dual-task training, or exergaming) with each other or with control conditions in adults aged 60 years or older with MCI or dementia. The co-primary outcomes will be the number of fallers (odds ratio) and the rate of falls (incidence rate ratio). Secondary outcomes will include balance, mobility, cognitive function, fear of falling, and exercise-related adverse events. We will search PubMed, Embase, Cochrane CENTRAL, Web of Science, CINAHL and three Chinese-language databases without language restrictions. Risk of bias will be assessed using the Cochrane Risk of Bias 2 (RoB 2) tool. Data will be synthesized using a frequentist random-effects NMA, with a Bayesian NMA as a sensitivity analysis. Treatments will be ranked using P-scores, and the certainty of evidence will be assessed using the Confidence in Network Meta-Analysis (CINeMA) approach. Discussion This NMA aims to provide a synthesis of the comparative efficacy of exercise interventions for fall prevention in cognitively impaired older adults. Should the available evidence prove sufficient, the resulting treatment rankings may help inform evidence-based exercise prescription in clinical practice. Systematic review registration https://www.crd.york.ac.uk/PROSPERO/view/CRD420261303641 , PROSPERO ID: CRD420261303641.