Effect of turmeric products on knee osteoarthritis: a systematic review and network meta-analysis

Wai HS, Pathomwichaiwat T, Suansanae T, Nathisuwan S, Rattanavipanon W · BMC Complementary Medicine and Therapies

The only formulation-versus-formulation comparison in the field. Every turmeric preparation beat placebo on WOMAC pain — conventional preparations -3.17 (95% CI -5.50 to -0.83), bioavailability-enhanced ones -2.47 (-3.27 to -1.67), conventional plus an active drug -4.01 (-6.22 to -1.80) — but no comparison between interventions reached significance, so the enhanced products did not out-perform the conventional ones. The SUCRA ranking places the enhanced preparations last of the active interventions on both WOMAC pain and function. Of the placebo comparisons only the enhanced preparation was assessed against the minimum clinically important difference, and it cleared it, at 29.7% change from baseline on pain and 54.4% on function. WOMAC stiffness improved on nothing — not against placebo, not against any other intervention.

Population

17 RCTs of turmeric preparations in knee osteoarthritis (PROSPERO CRD42023464749); the WOMAC pain network is drawn from 7 of them, covering 913 patients

Intervention

Conventional curcuminoid preparations and bioavailability-enhanced curcuminoid preparations, alone or with NSAIDs/acetaminophen

Comparator

Placebo and active drug comparators, compared indirectly

Limitations

Certainty of the network estimates is low or very low for every statistically significant outcome, from imprecision and incoherence; the direct estimates reach moderate at best. The formulation classes pool very different products and doses, and the planned dose subgroup analysis could not be run because the trials measured different analytes. Baseline pain intensity was higher in one comparison than the rest, and excluding it moved the rankings. The network is thin, so the absence of a bioavailability advantage is suggestive rather than settled.