Efficacy and safety of curcuminoids alone in alleviating pain and dysfunction for knee osteoarthritis: a systematic review and meta-analysis of randomized controlled trials

Feng J, Li Z, Tian L, Mu P, Hu Y, Xiong F, Ma X · BMC Complementary Medicine and Therapies

Fifteen trials and 1,670 patients, restricted to curcuminoids given alone. Against placebo the results section reports VAS pain -1.77 (95% CI -2.44 to -1.09), WOMAC total -10.47 (-15.65 to -5.30), WOMAC pain -1.94 (-2.91 to -0.97), WOMAC function -6.36 (-8.94 to -3.78) and WOMAC stiffness -0.54 (-1.03 to -0.05). Only VAS pain and WOMAC total cleared the minimum clinically important difference; of the other three the authors say the effect carries statistical significance only. Against NSAIDs nothing separated — VAS pain -0.30 (-0.63 to 0.04), WOMAC total -0.68 (-3.88 to 2.52), WOMAC pain 0.24 (-0.47 to 0.96), function -0.57 (-3.07 to 1.94), stiffness 0.19 (-0.17 to 0.56) — with every point estimate below the clinical threshold, which is what the authors mean by not inferior. Adverse events did not differ from placebo (RR 1.07; 0.70 to 1.65) and ran lower than on NSAIDs without reaching significance (RR 0.65; 0.41 to 1.03).

Population

15 RCTs restricted to curcuminoids given alone, in knee osteoarthritis

Intervention

Curcuminoids as monotherapy

Comparator

Placebo or NSAIDs

Limitations

The abstract and the results section do not agree. The abstract gives WOMAC total as -7.06, WOMAC pain as -1.42, function as -5.04 and the placebo adverse-event risk ratio as 1.03; the results section gives -10.47, -1.94, -6.36 and 1.07, and reasons about the clinical-importance thresholds from its own figures. Heterogeneity against placebo ran 77% to 87% on VAS pain and on the WOMAC pain and stiffness subscales. The authors restrict to curcuminoid-only interventions to remove co-intervention noise but then describe the pooled trials as low quality, and make only a cautious and conservative recommendation.