Adherence and effectiveness of multicomponent exercise fall prevention programmes across delivery formats in community-dwelling older adults: a systematic review and multilevel meta-analysis.
BACKGROUND: Multicomponent exercise is widely implemented for fall prevention in older adults, but the role of exercise adherence and delivery format in determining intervention effectiveness remains inadequately understood, particularly in analyses that account for within-study outcome dependence. METHODS: We searched five databases to 23 November 2025 for RCTs of multicomponent exercise in adults aged ≥60 years. Fall outcomes were synthesised using three-level random-effects meta-analysis and meta-regression, with adherence evaluated as continuous/categorical factors and as moderator alongside delivery format, accounting for within-study effect-size dependence. RESULTS: Forty-one publications were included, comprising 36 distinct cohorts and involving 20,202 participants. Multicomponent exercise programmes significantly reduced fall incidence (IRR = 0.776, 95% CI 0.641-0.939), risk of experiencing at least one fall (RR = 0.914, 95% CI 0.858-0.974), and fall-related injury incidence (IRR = 0.754, 95% CI 0.663-0.858). Programmes achieving adherence ≥75% were independently associated with greater reductions in fall incidence (IRR = 0.501, 95% CI 0.336-0.747) and fall risk (RR = 0.821, 95% CI 0.697-0.967). Mean adherence was highest in combined individual and group delivery, lower in single-format programmes, though including a group component may enhance adherence. CONCLUSIONS: Multicomponent exercise fall prevention programmes are effective for fall prevention in community-dwelling older adults, whereas evidence for their protective effects on fall-related injury outcomes remains limited. Adherence is an important factor influencing programme implementation and intervention effectiveness.