Abstract
BACKGROUND: Dance is a non-pharmacological option for people with cognitive impairment, yet its comparative advantage over other exercise modalities is still uncertain. This study compared and ranked dance versus other exercise in this population. METHODS: A systematic review and network meta-analysis of randomized controlld trials was conducted using five databases. Primary outcomes were cognitive function, psychological wellbeing, and physical performance. Standardized mean differences and p-values were calculated using pairwise and network meta-analysis within a random-effects model. PROSPERO (CRD42024549996). RESULTS: The systematic review included 209 randomized controlled trials (3,773 participants). Pairwise meta-analyses identified several statistically significant effects for dance interventions. In the network meta-analysis, dance showed no statistically significant advantages over other modalities for any outcome. Nevertheless, for global cognition, dance ranked highest on MoCA by the Surface Under the Cumulative Ranking Curve. For working memory and attention outcomes, dance ranked first or second. For physical performance and mood, it generally ranked second or mid-range. For executive function, speed/attention control, and mobility, dance ranked lower. Several favorable Surface Under the Cumulative Ranking Curve results were driven by single dance studies, warranting caution in interpretation. CONCLUSION: Exercise yields broad benefits for individuals with cognitive impairment. Although dance was not superior across all outcomes, it ranked favorably on selected measures and performed competitively, sometimes on sparse evidence. Given its integrative combination of movement, rhythm, cognitive engagement, and social interaction, dance remains a promising, underutilized option for cognitive rehabilitation, warranting wider implementation and more robust head-to-head trials.