Exercise improves short-term function in frail older adults but its effect on actual falls remains unproven, meta-analysis finds

A meta-analysis pooled 14 randomised controlled trials (2,719 participants) meeting the Fried frailty phenotype Meta-analysis. Short-term exercise programmes, of 12 weeks or less, improved gait speed, Timed Up and Go performance, and five-times sit-to-stand performance. Evidence certainty was rated low to very low throughout, and falls-incidence data — reported in only two of the 14 trials — were too sparse to pool into a combined estimate.

The review’s own authors caution that these functional-outcome gains have not been robustly linked to an actual reduction in falls.

Services using gait speed or Timed Up and Go improvements as a proxy for falls-prevention success should treat that link as unproven rather than established: the evidence for the functional-outcome improvement is considerably stronger than the evidence that it translates into fewer falls, and better-powered trials reporting falls incidence directly are needed before that gap can be closed.