Skip to content
Buy This GLP
← Research
Safety

Hip fractures halved, which is two people in a thousand

The fracture hazard ratio is the dramatic number. The falls result, with a weaker ratio, prevents nine times as much in absolute terms.

Wesley Jenkins6 min read
One year, adults 65+ — semaglutide vs DPP-4 drugsfallsHR 0.6633.6%5.4%femoral fractureHR 0.4880.3%0.5%The weaker hazard ratio produces the larger real-world gap.

Bone is the outcome nobody prices and, on the evidence so far, nobody agrees about either — the reviews are set against each other in two large reviews, one answer and one disagreement. A retrospective cohort added falls to the question, which turns out to change what the numbers mean [1].

Adults aged 65 and over with type 2 diabetes and overweight or obesity were matched against people taking DPP-4 inhibitors. Femoral fractures over one year occurred in 0.3% of the semaglutide group against 0.5% of controls, a hazard ratio of 0.488 (95% CI 0.367–0.649). Tirzepatide gave 0.2% against 0.4%, a hazard ratio of 0.452.

The comparison group is where the caution belongs. DPP-4 inhibitors are gentle oral drugs with few side effects, and in an older population a clinician reaching for one instead of a weekly injection that causes nausea and substantial weight loss is often making that choice for someone already frail. Falls and fractures are largely what frailty produces. Propensity matching on recorded characteristics cannot capture the judgment that led to the prescription, and it runs hard in the direction of this result — the same problem described in the pre-procedure hold study and in covered and still not treated, where who gets treated is never random.

One internal detail pushes the other way and deserves its weight. Fracture reduction was more pronounced in people with a BMI of 30 or above, which is what a genuine mechanism would look like if carrying less weight reduces the force of a fall, and is not an obvious artifact of frailty-based prescribing. Fall reduction, by contrast, was consistent across subgroups.

For a reader deciding what to buy, the practical content is thin but not zero. These drugs do not appear to be breaking older people's hips, which was a live concern given how much bone density work has pointed the other way in animals. What they do to bone over a decade remains unmeasured, and a one-year window in a claims database is not where that will be settled — the same limitation that shapes the muscle question in the cost nobody prices, keeping muscle.

Frequently asked

Do these drugs cut hip fracture risk in half?
The hazard ratio was 0.488, but the underlying rates were 0.3% against 0.5% — about two fractures avoided per thousand people per year. The relative figure and the absolute one describe very different sized benefits.
Which result matters more?
Falls, in practical terms. A weaker hazard ratio of 0.663 applied to a much commoner event produced an absolute difference of 1.8 percentage points, nine times the fracture difference.
Could the comparison be misleading?
Possibly. DPP-4 inhibitors are gentle oral drugs often chosen for frailer older patients, and frailty is what causes falls and fractures. Matching on recorded characteristics cannot capture why a clinician chose one drug over the other.

Sources

  1. [1] Chen HY, Wu JY, Chu YH, Chen TW, Huang CF (2026). Lower fall and femoral fracture risks with semaglutide and tirzepatide compared with DPP-4 inhibitors in older adults with type 2 diabetes Osteoporosis International. PMID 42435064

More in Safety