Trial evidence that osteoporosis drugs prevent fractures in men is very uncertain

Design

Cochrane systematic review and meta-analysis of 17 randomised trials

Population

4,132 men aged 50 years or older

Primary outcome

Hip, symptomatic vertebral and other fractures

Effect

Bisphosphonate vs placebo: vertebral RR 0.49 (0.14–1.74); hip RR 0.73 (0.06–8.51); very low certainty

This Cochrane review included 17 randomised trials with 4,132 men aged 50 or over, testing bisphosphonates, teriparatide-type drugs, denosumab and romosozumab. Most osteoporosis trials have enrolled women, so evidence in men has been extrapolated.

For bisphosphonates against placebo at up to two years, fractures were rare and estimates imprecise: hip fracture 2 vs 3 per 1,000 (RR 0.73, 95% CI 0.06–8.51), symptomatic vertebral fracture 4 vs 8 per 1,000 (RR 0.49, 0.14–1.74), and other fractures 16 vs 21 per 1,000 (RR 0.78, 0.42–1.45), all very low certainty. Bisphosphonates probably did not increase adverse events (RR 1.06) or serious adverse events (RR 0.95). Fracture effects of the other drug classes were equally uncertain.

This does not show the drugs fail in men. It shows the trials in men were too small and too short to measure fractures, and most were designed around bone density. Treatment decisions in men at high fracture risk still rest on bone density responses and extrapolation from women, which should be explained honestly. In rheumatology, where men on long-term glucocorticoids are common, fracture risk assessment remains essential.

• Assess fracture risk in men on long-term glucocorticoids and in those with previous fragility fractures
• Explain to men that fracture-reduction evidence comes mainly from trials in women
• Treat men at high fracture risk, recognising that benefit is extrapolated
• Correct vitamin D deficiency and address falls risk alongside drug treatment

Why it matters

Men make up a large share of fractures and glucocorticoid users, yet trials have not been large enough to show whether treatment prevents their fractures.

Don't overread it

Very uncertain is not the same as ineffective — the trials were too small to detect fracture reduction.

The statistics, in plain English

A hip fracture risk ratio of 0.73 with an interval from 0.06 to 8.51 is compatible with a large benefit or a large harm, which is what happens with 4 hip fractures in 1,635 men. 'Very low certainty' means the true effect may be substantially different from the estimate.

Read the rest in the app

You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.

Get Daily Doctor — free

Scan to keep reading on your phone. No account needed to start.