A randomized, placebo-controlled trial found that oral resveratrol supplementation did not reduce knee pain in adult patients with osteoarthritis compared with placebo.

In the ARTHROL study, published in PLOS Medicine, researchers evaluated 142 adults with painful knee osteoarthritis over 6 months. The participants received either oral resveratrol (40 mg twice daily for 1 week, then 20 mg twice daily) or matched placebo.

The double-blind trial was conducted at three tertiary care centers in France. Among the 649 patients assessed for eligibility, 142 (22%) were recruited. Eligible participants were adults aged ≥ 40 years with knee osteoarthritis per American College of Rheumatology criteria, pain duration ≥ 1 month, pain intensity ≥ 40/100, and radiographic Kellgren-Lawrence grade 1 to 3.

At baseline, participants had a mean age of 61.4 years (standard deviation [SD] = 9.6), were 71% female, and reported mean knee pain of 56.2/100 (SD = 13.5). Mean symptom duration was 8.5 years (SD = 8.2). The mean body mass index was 28.3 kg/m2 (SD = 6.2). Radiographically, 17% had Kellgren-Lawrence grade 1, 32% grade 2, and 51% grade 3.

Adherence was assessed by caplet counts at 6 months. The mean number of remaining caplets was 239.4 (SD = 49.1, n = 58) in the resveratrol group and 222.2 (SD = 27.6, n = 48) in the placebo group.

At 3 months, the mean reduction in knee pain on a 0 to 100 numeric rating scale was –15.7 points (95% confidence interval [CI] = –21.1 to –10.3) in the resveratrol group vs –15.2 points (95% CI = –20.5 to –9.8) in the placebo group (absolute difference = –0.6 points, 95% CI = –8.0 to 6.9, P = .88). Similar results were observed at 6 months: –16.8 points (95% CI = –23.4 to –10.3) vs –17.1 points (95% CI = –23.4 to –10.9), respectively (absolute difference = 0.4 points, 95% CI = –8.4 to 9.1, P = .93).

Key secondary outcomes also showed no statistically significant differences between the groups:

• WOMAC function subscale scores at 3 months: –9.2 (95% CI = –13.0 to –5.4) vs –10.6 (95% CI = –14.3 to –6.8), absolute difference = 1.4 (95% CI = –3.9 to 6.7, P = .59)
• Patient global assessment at 3 months: 1.4 (95% CI = –3.3 to 6.2) vs 1.2 (95% CI = –3.5 to 5.9), absolute difference = 0.2 (95% CI = –5.9 to 6.4, P = .95)
• OARSI-OMERACT response rates at 3 months: 52% vs 50%, relative risk = 1.03 (95% CI = 0.74 to 1.43, P = .86)
• Use of rescue medications (no statistically significant differences in intra-articular injections, analgesics, or nonsteroidal anti-inflammatory drugs [NSAIDs]).

Adverse events were similar between groups. Serious adverse events occurred in 4% (n = 3/71) of the patients who received resveratrol and 3% (n = 2/71) of those who received placebo, with none considered related to treatment. A total of 95 adverse events were reported: 52 in the resveratrol group and 43 in the placebo group.

The researchers noted some study limitations, including being underpowered because of slow enrollment during the COVID-19 pandemic. The study was stopped before reaching the prespecified sample size of 164 participants. The estimated effect size of 0.55 was likely optimistic, as no oral medical intervention for knee osteoarthritis has demonstrated a standardized effect size greater than 0.4.

The researchers suggested the lack of efficacy may have been the result of insufficient bioavailability of trans-resveratrol in joint tissues, despite using an optimized formulation. However, no biological samples were collected to evaluate this hypothesis.

They also noted the population had fairly severe, long-standing symptoms, which may require more comprehensive treatment approaches. At baseline, 63% of the participants reported using non-opioid analgesics, 44% used NSAIDs, and 14% had received intra-articular injections in the previous 3 months.

The substantial placebo response observed (15-point pain reduction) highlighted the challenges of demonstrating superiority to placebo in osteoarthritis trials. When considering a response level of at least 20% reduction in pain, corresponding to the OARSI-OMERACT response criteria, the responder rate was 54% in the resveratrol group and 56% in the placebo group.

A statement of competing interests can be found in the study.