TOPLINE:

Twins who started metformin had a 40% lower risk for being diagnosed with incident osteoarthritis over an 11-year follow-up period than their untreated co-twins.

METHODOLOGY:

• Researchers conducted a co-twin study in which one twin started metformin and the other did not, to assess whether starting metformin altered the risk for a first diagnosis of osteoarthritis.
• They included Swedish twins aged 35-75 years who were alive on December 31, 2005, had no prior diagnosis of osteoarthritis, and had not used metformin between June and December 2005. The full eligible sample included 52,341 individuals (twins or singletons).
• The main analysis compared 1261 twins who started metformin treatment with their 1261 untreated co-twins. Participants in both groups had a mean age of 61.7 years, and women comprised 39% of the metformin group and 47% of the untreated group.
• The main outcome was incident osteoarthritis in peripheral joints recorded in specialist care.
• Participants were followed up for up to 11 years.

TAKEAWAY:

• Over 11 years, 5.4% of twins treated with metformin developed osteoarthritis compared with 6.3% of untreated twins, corresponding to a 40% lower risk (adjusted hazard ratio, 0.60; 95% CI, 0.39-0.92).
• The 10-year cumulative incidence was 6.7% in metformin-treated twins and 9.7% in untreated twins, corresponding to an adjusted absolute risk reduction of 3.1%.
• Treated twins had a higher BMI than untreated twins (26.1 vs 24.4), yet they had a lower risk for osteoarthritis after adjusting for confounders.

IN PRACTICE:

“Our study indicates that there may be favorable effects of metformin on incident OA [osteoarthritis] diagnoses, at least when OA is diagnosed in specialist care and when we compare to individuals with no treatment,” the authors of the study wrote.

“[The study] observations point to the need of studying BMI-reducing interventions in a general population, with the aim to lower the incidence of OA,” they added.

SOURCE:

The study was led by Karin Magnusson, PhD, Lund University, Lund, Sweden. It was published online on December 18, 2025, in Osteoarthritis and Cartilage.

LIMITATIONS:

Researchers could not adjust for baseline diabetes because primary care diagnoses were missing. Osteoarthritis was likely undercounted because only specialist records were used. Certain confounding factors such as diet or physical activity may have remained unaccounted for.

DISCLOSURES:

The study received funding from multiple sources, including the Greta and Johan Kock Foundation, the Swedish Research Council, and the Osterlund Foundation. One author reported having served as a consultant to Grünenthal Sweden AB and Key2Compliance AB. Another author disclosed being associate editor for statistics in Osteoarthritis and Cartilage. One author reported being a part-time scientific advisor for Arthro Therapeutics.

This article was created using several editorial tools, including AI, as part of the process. Human editors reviewed this content before publication.