Skip to content

Exercise review reports better balance, with uncertainty around the falls estimate

Exercise trials favored better balance in older adults without diagnosed diseases; the abstract's falls result is harder to interpret.

By 100HP editorialAbstract-based explanation checked

Based on the published abstract. The full paper may contain additional methods, results and limitations.

The 30-second takeaway

The review reports better balance and mobility test results after exercise among older adults without diagnosed diseases. Its authors also conclude that falls were reduced, but the abstract uses inconsistent wording and does not clearly define the falls outcome. Better test performance alone does not quantify someone's chance of avoiding a fall.

Keep in mind

The abstract describes falls as a count but reports an odds ratio, with conflicting wording about direction. Control activities and overall follow-up duration are also missing.

THE NUMBERS, WITH CONTEXT

What researchers found

0.92 (95% CI 0.63 to 1.21)

Berg Balance Scale: standardized effect (g)

Exercise versus controls after intervention; exact timing was unspecified. The difference is measured relative to score variability, in standard deviations, not Berg scale points or percentages. Positive favors exercise.

PeerJ, 2025 · Original source ↓

Different outcomes, different meanings

The review assessed balance, timed mobility and scores on a fall-efficacy scale. Each measure answers a different question; changes in these test or scale scores should be distinguished from actual falls.

Schedules remain uncertain

The authors proposed different exercise schedules for fall efficacy and balance using subgroup analyses. Without direct comparative estimates in the abstract, readers cannot judge whether any schedule reliably outperformed another.

CHECK THE ORIGINAL

The original publication

Effects of exercise intervention on falls and balance function in older adults: a systematic review and meta-analysis.

Yu H, Zhong J, Li M et al.
PeerJ · 2025

PubMed ID
41122235
Record checked

AI-assisted research and writing. This explains one selected publication; it is not a complete review of everything known. Our approach.