Reference: JAMA. 2025 May 27;333(20):1804-1812
Practice Point: Metformin might help reduce knee pain and improve function in adults with overweight or obesity and symptomatic knee osteoarthritis.
EBM Pearl: Trials conducted via telemedicine may be convenient but may also be subject to inclusion bias.
Knee osteoarthritis disproportionately affects older adults with overweight or obesity, often resulting in chronic pain and reduced functional capacity. Emerging research is investigating the potential of metformin – a well-established drug in the management of type 2 diabetes – as a possible therapy for knee osteoarthritis. Could this decades-old drug offer a novel approach to addressing joint degeneration and symptom burden in this population?
A recent trial conducted in Australia presents preliminary evidence that metformin may help. This trial, using mostly telemedicine-based methods, evaluated 107 adults > 40 years old (mostly female) with symptomatic knee osteoarthritis and overweight or obesity who were randomized to metformin or placebo for 6 months. Patients were recruited from the community through local and social media. Metformin was given as an extended-release formulation and titrated over 6 weeks from 500 mg once daily up to 2,000 mg once daily. Notably, patients with diabetes, severe knee pain (visual analog scale [VAS] score > 80 mm), and inflammatory arthritis were excluded. There were two onsite or telemedicine visits (at baseline and 6 months). Mean body mass index at baseline was 32.7 kg/m2, and mean VAS pain score was 59-60 mm over the week prior to enrollment (moderate knee pain).
Compared to the control group, pain was lower at 6 months in the metformin group by about 11-12 mm. While this difference was statistically significant (VAS mean change -31 mm vs. -18 mm in placebo, 95% CI -20 to -2 mm), the authors had predetermined that the minimal clinically important difference was 15 mm. Since the 95% CI (the difference between groups we would expect to see outside of this study 95% of the time) includes differences ranging from a 2 point difference to a 20 point difference, the clinical significance is uncertain. That said, the Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) scores for pain, stiffness, and function also improved. What’s interesting is that weight was only marginally reduced with metformin (-1.8 kg vs. -1.2 kg with placebo). Adverse events are what you expect with metformin (i.e., diarrhea and abdominal pain), and they didn’t result in attrition. Only 3 patients quit metformin permanently due to adverse effects.
Of course, we must acknowledge the limitations of conducting a trial over telemedicine. No objective measures of function could be assessed – there were no physical exams! Key outcomes such as weight and arthritic features were via self-report only, and only 38% of participants returned their medication as instructed for the adherence assessment. Bias may also result from community recruitment via social media and those who have the privilege of access to telehealth capabilities. This necessarily restricts the population to those with the financial and social resources to access telehealth. However, metformin showed promise where it counts: less pain and better function. Bonus – metformin is cheap (relatively speaking) compared to other treatment options, and its safety is well-established. Old drugs can offer new hope for millions living with knee pain.
For more information, see the topic Medications for Osteoarthritis (OA) of the Knee in DynaMed.
DynaMed EBM Focus Editorial Team
This EBM Focus was written by McKenzie Ferguson, PharmD, BCPS, Senior Clinical Writer at DynaMed. Of course, we must acknowledge the limitations of conducting a trial over telemedicine. No objective measures of function could be assessed – there were no physical exams! Key outcomes such as weight and arthritic features were via self-report only, and only 38% of participants returned their medication as instructed for the adherence assessment. Bias may also result from community recruitment via social media and those who have the privilege of access to telehealth capabilities. This necessarily restricts the population to those with the financial and social resources to access telehealth. However, metformin showed promise where it counts: less pain and better function. Bonus – metformin is cheap (relatively speaking) compared to other treatment options, and its safety is well-established. Old drugs can offer new hope for millions living with knee pain.