News · Fitness & Exercise
One osteoporosis drug came with 17% less new osteoarthritis than the standard alternative
Denosumab and bisphosphonates both treat thin bones. Comparing new users in Japanese insurance records, one of them also came with fewer new osteoarthritis diagnoses over three years, at 2 fewer cases per 100 people.
- One osteoporosis drug came with 17% less new osteoarthritis than the standard alternative.
- In plain terms that is about 2 fewer cases per 100 people over three years.
- Both drugs treat thin bones, so this is a comparison, not a treatment for arthritis.
- People who switched between the drugs showed a weaker signal that could be chance.
- Built from Japanese insurance records rather than a trial, so cause is not established.
Osteoporosis and osteoarthritis get confused constantly, including by people who have one of them. They are different diseases in different tissues, and they are usually managed by different people.
Osteoporosis is a disease in which your bones become weak and are likely to fracture. Osteoarthritis is a type of arthritis that only affects the joints, in which the cartilage breaks down and becomes rough. One is a problem of bone density, the other of the surface where two bones meet.
They also happen to the same people, at the same ages. Writing in the Journal of Bone and Mineral Research, researchers asked whether the drug chosen for the first has any bearing on the second.
Comparing new users of denosumab with new users of oral bisphosphonates, the absolute risk reduction for incident osteoarthritis was 2.0% over three years.
Why anyone thought to look
The idea is not as far-fetched as the two diagnoses make it sound.
Denosumab has demonstrated potential bone- and cartilage-protective effects in preclinical and early trials, and the reason is anatomical. In an arthritic joint the cartilage wears, but the bone directly underneath it is also changing: it thickens, it forms cysts, it grows the spurs a radiologist reports. Bumps of extra bone called bone spurs may grow in the joint area.
A drug that slows bone turnover might therefore slow the bone half of that process. The question was whether an effect visible in laboratory models and short trials shows up in ordinary patients living with two common diseases.
What was compared, and how
Both arms were treated. This is not a drug against nothing.
The researchers ran a sequential nested target trial emulation with propensity score matching based on longitudinal health insurance claims data from Japan, collected between 2015 and 2024, in adults aged 50 or over with osteoporosis.
That method takes routine records and analyses them as if they were the randomized trial nobody has run: eligibility, treatment start and follow-up are all specified in advance, which removes several of the biases that make ordinary database comparisons unreliable. It does not remove the biggest one.
Two comparisons were made: new users of each drug, and people who switched from bisphosphonates to denosumab against those who continued.
The result, and where it held
The new-user comparison was the clearer of the two.
At three years, the absolute risk reduction for incident osteoarthritis was 2.0%, which works out at about 17% fewer new diagnoses. The intention-to-treat analysis produced almost the same figures, 1.9% and 0.84, which is reassuring: results that survive both analyses are less likely to be an artifact of who stayed on treatment.
The switch comparison did not hold. Patients who switched from oral bisphosphonates to denosumab, against those who continued, showed 0.9% with an interval crossing zero.
The authors summarize it carefully: denosumab initiation was associated with a modestly lower risk of incident osteoarthritis than oral bisphosphonate initiation, although the association was weaker in the switch-continue comparison.
The objection that governs everything
Doctors do not choose between these two drugs at random, and the reasons are systematic.
Denosumab is often preferred for patients with impaired kidney function, for whom bisphosphonates are unsuitable, and for those at higher fracture risk. Bisphosphonates are cheap, oral, and the usual first choice for everyone else. Those are two different populations before any drug is given.
Matching on recorded characteristics handles what the claims data captured. It cannot handle a rheumatologist’s impression of frailty, or how much a patient walks, or whether their knees already ached in a way nobody coded.
There is a further wrinkle specific to this outcome. Osteoarthritis in insurance records is a diagnosis somebody wrote down, and how often that happens depends on how often the person sees a doctor. Denosumab requires an injection every six months; bisphosphonates can be collected from a pharmacy. Those are different amounts of clinical contact, and more contact usually means more diagnoses, which would push the result the other way.
What it is worth
Two fewer cases per hundred over three years, in an observational comparison, is not a reason to change a prescription.
There is no cure for osteoarthritis, and it usually gets worse slowly, which is what makes even a modest preventive signal interesting. But osteoporosis drugs are chosen on fracture risk, kidney function, dental status and tolerability, and this finding is nowhere near strong enough to enter that calculation.
Where it does belong is on the list of questions a trial should answer. A hypothesis that started in animal models has now survived a reasonably careful look at real patients, which is the point at which somebody should design the study that settles it.
People also ask
What did the study find?
In the per-protocol analyses, the absolute risk reduction for incident osteoarthritis was 2.0% (95% CI 1.1%-3.2%) in the new-user comparison, with a relative risk of 0.83 (0.73-0.91), and 0.9% (-0.2% to 2.8%) in the switch-continue comparison, with an RR of 0.92 (0.76-1.02).
What is denosumab?
An injected antibody given twice a year that blocks a signal called RANKL, which is what tells bone-eating cells to get to work. Bisphosphonates, the usual alternative, are tablets that poison those cells more directly.
Why would a bone drug affect a joint?
Because osteoarthritis is not purely a cartilage disease. The bone directly beneath the cartilage remodels as the joint deteriorates, and denosumab has shown bone- and cartilage-protective effects in preclinical and early trials. Whether that carries into whole joints in real patients is what this tested.
How large is a 2 percentage point difference?
About two fewer people per hundred developing osteoarthritis over three years, or one avoided case for every fifty people treated. Modest, and not nothing in a condition with no cure.
Why did switching not show the same benefit?
The switch-continue comparison, of people moving from bisphosphonates to denosumab versus staying, gave a smaller estimate whose interval crossed no difference. That may mean the benefit needs to start early, or simply that this comparison had fewer people in it.
Is this a randomized trial?
No. It emulates one from insurance claims, which fixes eligibility, treatment start and follow-up in advance and matches on recorded characteristics. It cannot match on what was never recorded, and denosumab tends to be chosen for patients with worse kidney function or higher fracture risk.
Should this change which drug someone takes?
Not on its own. Osteoporosis drugs are chosen on fracture risk, kidney function, dental health and what a person can tolerate, and osteoarthritis prevention is not currently one of the inputs. This is general information rather than medical advice.