Reviews/Osteoporosis Fracture Prevention/Exercise fall and fracture risk reduction
Exercise And Fall Prevention · ostfrac_20 · v2
This corpus asks whether exercise programs actually prevent broken bones in older people, and the trial evidence pulls in opposite directions. The single largest and most rigorous trial [1], which enrolled about 9,800 people, found no reduction in fractures from a targeted exercise program, and if anything a slight (non-significant) increase. A much smaller, non-randomised study [11] did find significantly fewer fractures with exercise. Long-term observational studies that simply follow active versus inactive people [10][14][15] fairly consistently link more activity to fewer hip fractures, but people who exercise tend to be healthier to begin with, so these cannot prove cause and effect. Because the best-designed intervention did not confirm a benefit while the supportive signal comes from a small trial and confoundable cohorts, the panel judged the evidence genuinely unresolved.
The best-powered intervention trial, PreFIT [1] (N=9803; also reported as [2]), found no fracture reduction from targeted exercise versus mailed advice (rate ratio 1.20, 95% CI 0.91 to 1.59). The cluster-randomised OFRA program [5] was null on all fragility fractures combined (HR 0.94, 95% CI 0.80-1.11) but reduced femoral fractures (HR 0.76, 95% CI 0.59-0.99). The small controlled EFOPS trial [11] (N=137) found a significant reduction in low-trauma fractures (RR 0.51, 95% CI 0.23 to 0.97, p=.046). Observational cohorts consistently associate activity with lower fracture, including walking in [10] (HR 0.864, 95% CI 0.762-0.980), though [13] (N=387,307) shows occupational physical activity can raise fracture risk (HR 1.47 in the highest quartile). The contested point: The division is between a protective signal from a small non-randomised trial [11] plus confoundable cohorts [10][14][15], and the null result of the largest dedicated randomised trial [1][2]. Whether the trials showing benefit [5][9] can attribute that benefit to exercise is also disputed, since they bundled DXA, vitamin D, calcium and environmental measures alongside exercise.
All 19 papers behind this claim with their DOIs and what each does to it, the panel’s arbitration, the evidence gap map and the open research questions. Free, and sent to your inbox.
We’ll also send you other claims from Osteoporosis Fracture Prevention, one at a time, and tell you when a release or a pivotal trial changes a grade. Unsubscribe by replying.
One reviewer's only job is to attack the finding. The challenges below were weighed before the grade was set.
The single largest randomised exercise-intervention trial with a fracture endpoint, PreFIT [1]/[2] (N≈9803), found no reduction and a non-significant trend toward MORE fractures with exercise (RR 1.20, 0.91-1.59) — the best-powered interventional evidence contradicts the claim.
The positive interventional result rests on EFOPS [11], which is small (N=137), single-centre, and described as a 'controlled' (not randomised) trial — high risk of bias and not generalisable to 'older adults' broadly.
Where it remains uncertain. The positive evidence is either small and non-randomised [11] or observational and confoundable [10][14][15], while the best-powered randomised test [1][2] is null and even trends toward more fractures.
Aleatory
PreFIT [1][2] reports only 379 of 9803 (3.9%) participants sustained a fracture, lower than anticipated, widening the confidence interval and limiting power to detect a true effect.
Epistemic
The corpus has no randomised trial powered on fractures as a primary endpoint that confirms benefit; [1] is null, [11] is small (N=137) and non-randomised, so the question is under-tested.
Inferential
Supportive cohorts [10][14][15][17][19] are observational; healthier people both exercise more and fracture less, as [6] notes, so association is not causation.
The load-bearing papers from a corpus of 19, each classified by what it does to the claim. Every DOI is taken from the corpus the claim was graded on.
A multicomponent exercise RCT showed no effect on falls and reported no fracture reduction, limiting the mechanistic chain from exercise to fewer fall-related fractures.
“There were no significant effects of the 18-month intervention on distal femur or proximal tibia trabecular bone microarchitecture or falls incidence”
Panel auditInternal validity: medium · scope of inference: Community-dwelling adults aged 60+ with osteopenia or increased falls risk, musculoskeletal surrogate outcomes.Key limitations: no fracture endpoint · low adherence · small sample
Cluster-randomised exercise-containing program null on the primary fracture endpoint but with a femoral-fracture reduction; mixed.
“The incidence of fragility fractures did not differ between the intervention and the control group (HR 0.94; 95% CI 0.80-1.11). However, femoral fractures, the most frequent fracture type, were reduced in the intervention group (HR 0.76; 95% CI 0.59-0.99).”
Panel auditInternal validity: medium · scope of inference: Older rural German residents with prior fracture or older women, in a routine-care prevention program.Key limitations: multifaceted intervention not isolated exercise · low exercise uptake (~30%) · routine-data endpoints
Observational support for a protective association between exercise and fall-related (hip) fracture, though confounded and non-interventional.
“Adjusted models showed that walking for exercise significantly reduced hip fracture risk (HR, 0.864; 95% CI, 0.762-0.980; P = 0.0230)”
Panel auditInternal validity: medium · scope of inference: US women aged 65+, walking-for-exercise exposure and hip fracture.Key limitations: observational not interventional · self-reported exposure · healthy-user confounding
Largest randomised exercise-intervention trial with a fracture primary endpoint is null, directly opposing the claim.
“Screening and targeted intervention did not result in lower fracture rates; the rate ratio for fracture with exercise as compared with advice by mail was 1.20 (95% confidence interval [CI], 0.91 to 1.59)”
Panel auditInternal validity: high · scope of inference: Community-dwelling adults aged 70+ in England under a screen-and-treat exercise strategy.Key limitations: intervention only to at-risk subset · 65% uptake · fracture rate lower than anticipated
4 of 19 papers shown. The full review lists all of them, with the panel’s arbitration →
Claims are graded at each quarterly release, and re-graded between releases only when a pivotal trial lands. Every change is a new version of the claim.
A directional signal is present but the largest, most rigorous exercise-intervention trials are null while the positive trial is small and non-randomised, so the corpus cannot carry a confident 'reduces'.
Release 2026-q4 re-graded every claim from its corpus under protocol v0.19. An earlier release's grade is shown for the record; it came from an earlier protocol and is not a movement.
Assumptions that carry weight but are not themselves established.
The panel records the gap that would most move the claim, and the study that would close it.
Adequately powered RCT of exercise alone with a fracture primary endpoint. The largest trial [1] is null and under-powered given the low 3.9% event rate, and the positive trial [11] is small and non-randomised, so the central interventional question is unresolved. The panel rates this the highest-importance open gap for the claim, and names the study design that would close it: A multi-centre randomised trial of a defined exercise program versus usual care in older adults at fracture risk, powered on incident fractures as the primary endpoint, with no bundled non-exercise co-interventions.
Open question: Does an exercise intervention, isolated from other measures, reduce fractures in older adults? Why it is still open: The largest randomised trial [1] is null with a trend toward more fractures, while the only positive trial [11] is small and non-randomised and benefit programs [5][9] are multifaceted.
Priors is a sceptical first reader, not peer review, and not medical advice.
Release 2026-q4 · graded 21 Oct 2026 · protocol v0.19