In adults with symptomatic knee osteoarthritis and overweight/obesity, 6 months of metformin 2000 mg/day modestly reduced pain and improved function versus placebo; gastrointestinal side effects were more frequent; larger trials are needed.

Background

Obesity-related knee osteoarthritis involves excess joint loading, systemic inflammation, and metabolic dysfunction. Metformin, a first-line therapy for type 2 diabetes, has anti-inflammatory and metabolic effects and may improve osteoarthritis symptoms. Preclinical and observational data suggested potential benefit, prompting this randomized trial.

Patients

• Community-dwelling adults in Victoria, Australia, recruited via telemedicine (June 2021–August 2023).
• Key inclusion: age >40 years, symptomatic knee OA ≥6 months (ACR-adapted criteria), knee pain >40 mm on 0–100 mm VAS, BMI ≥25.
• Key exclusion: severe radiographic OA (Kellgren-Lawrence grade 4), VAS pain >80, inflammatory arthritis, recent significant knee injury/surgery/injections, diabetes requiring glucose-lowering therapy, kidney or liver impairment.
• Randomized: 107 participants; mean age 58.8 years; 68% female; mean BMI 32.7.

Intervention

Metformin extended release, titrated to 2000 mg once daily over 6 weeks, continued for 6 months. Adherence monitored via telemedicine; 80% reached target dose; mean weight change −1.8 kg.

Control

Identical-appearing placebo once daily with the same titration schedule for 6 months; mean weight change −1.2 kg.

Outcome

• Primary: Change in knee pain at 6 months (100-mm VAS; MCID 15 mm).
• Secondary:
  • WOMAC pain (0–500), stiffness (0–200), function (0–1700); MCID ≈12% improvement.
  • Health-related quality of life (AQoL-8D; MCID 0.06).
  • OMERACT–OARSI responder status (validated modified definition).
  • Adverse events.

Study Design

Community-based, randomized, parallel-group, double-blind, placebo-controlled clinical trial conducted via telemedicine. Allocation 1:1 with concealed central randomization (permuted blocks). Intention-to-treat analyses with multiple imputation for missing outcomes. Location: Victoria, Australia.

Level of Evidence

Level I (randomized controlled trial).

Follow up period

6 months (82% completed primary outcome).

Results

• Primary outcome (VAS pain, 0–100 mm):
  • Mean change at 6 months: metformin −31.3 mm vs placebo −18.9 mm.
  • Between-group difference in change: −11.4 mm (95% CI −20.1 to −2.6; P=0.01).
  • Standardized mean difference (effect size): 0.43 (95% CI 0.02 to 0.83).
  • Note: Between-group difference did not reach the prespecified MCID of 15 mm.
• Secondary outcomes:
  • WOMAC pain: metformin −113.9 vs placebo −68.2; adjusted difference −42.4 (95% CI −83.9 to −1.0; P=0.045).
  • WOMAC stiffness: metformin −56.9 vs placebo −26.7; adjusted difference −23.0 (95% CI −40.4 to −5.7; P=0.01).
  • WOMAC function: metformin −426.1 vs placebo −221.7; adjusted difference −179.8 (95% CI −313.0 to −46.6; P=0.009).
  • Quality of life (AQoL-8D): no significant between-group difference (adjusted difference 0.01; 95% CI −0.02 to 0.05; P=0.47).
  • OMERACT–OARSI responder rate: 65% (35/54) metformin vs 45% (24/53) placebo; absolute difference 20%; odds ratio 2.21 (95% CI 0.92 to 5.31). NNT ≈ 5 (exploratory; not statistically significant).
  • Adverse events: no serious events. Any AE: 30% metformin vs 19% placebo. Most common were diarrhea (15% vs 8%) and abdominal discomfort (13% vs 9%); generally mild–moderate.
• Subgroups (exploratory): Benefit on VAS pain observed at 6 months in females and across joint space narrowing grades; no effect at 3 months.

Limitations

• Modest sample size; 18% missing primary outcome at 6 months.
• Telemedicine design: limited objective assessments; weight and clinical features self-reported; adherence largely self-reported (low pill-return rate).
• Between-group pain difference did not meet the VAS MCID of 15 mm, raising questions about clinical significance despite statistical significance.
• OMERACT–OARSI responder difference was not statistically significant and secondary outcomes were not adjusted for multiplicity.
• Potential unblinding from gastrointestinal side effects could bias self-reported outcomes.
• Race/ethnicity not reported, limiting generalizability.

Funding

Funded by Australia’s National Health and Medical Research Council (NHMRC; APP1194829). Additional fellowships supported individual investigators. Funders had no role in study design, conduct, data analysis/interpretation, manuscript preparation, or publication decisions.

Citation

Pan F, Wang Y, Lim YZ, Urquhart DM, Estee MM, Wluka AE, Wolfe R, Cicuttini FM. Metformin for Knee Osteoarthritis in Patients With Overweight or Obesity: A Randomized Clinical Trial. JAMA. 2025;333(20):1804-1812. doi:10.1001/jama.2025.3471. Trial registration: ANZCTR ACTRN12621000710820.