CurcuminTurmeric extract, various delivery systems · 4 claims

A network meta-analysis of 17 studies found all turmeric preparations significantly reduced WOMAC pain. Bioavailability-enhanced curcuminoid preparations gave a mean difference of −2.47 (95% CI −3.27 to −1.67) and achieved a 30% reduction versus placebo, reaching the minimal clinically important difference. Conventional preparations gave −3.17 with a much wider interval. The authors state the certainty of evidence remains low.

The quality picture is the story here. A 2025 critical appraisal examined 25 meta-analyses published between 2016 and 2024. Only nine were preregistered. Doses ranged from 50 mg to 6,000 mg and durations from one day to twelve months. On methodological quality the appraisers rated 19 as very low and 3 as low.

So: many syntheses, pointing the same way, almost all of them methodologically weak. That is a different situation from either strong evidence or no evidence, and the score reflects it.

• MetaWai HS, Pathomwichaiwat T, Suansanae T, Nathisuwan S, Rattanavipanon W. Effect of turmeric products on knee osteoarthritis: systematic review and network meta-analysis BMC Complement Med Ther 2025 · doi 10.1186/s12906-025-05045-z · PROSPERO CRD42023464749
• RevA critical review of systematic reviews and meta-analyses of curcumin for knee osteoarthritis Front Pharmacol 2025 · 25 meta-analyses appraised, 19 rated very low quality
• RevNCCIH: Turmeric, usefulness and safety “Initial evidence is positive; higher-quality evidence is needed”

Curcumin is poorly absorbed by design of its own chemistry. In animal work, 10 mg/kg given intravenously produced a peak serum level of 0.36 µg/mL, while an oral dose fifty times higher produced only 0.06 µg/mL.

In a randomised crossover study in humans using a curcumin-lecithin formulation, investigators could not detect free curcumin in the blood at all — only phase-2 metabolites. This is why every commercial product is built around a delivery system rather than plain extract.

Turmeric in cooking is a spice. Whatever the trials show, they are not showing it for the contents of your spice rack.

• PKBioavailability data summarised in: Bioavailable turmeric extract for knee osteoarthritis, randomized non-inferiority trial versus paracetamol IV vs oral serum comparisons; phase-2 metabolites only in human crossover study

Piperine does increase curcumin bioavailability, which is why it appears in most formulations. The pharmacology literature describes using it as something that “involves huge risk and close monitoring of patients” — because it increases absorption of other drugs and supplements at the same time, and decreases hepatic metabolism.

In other words, the reason it works is that it interferes with how your body clears things. That is fine in isolation and much less fine if you take anything else. It is listed on the label as a feature.

• PKPiperine bioavailability enhancement and associated interaction risk Discussed in bioavailable turmeric extract trial background
• RevNCCIH on curcumin products and piperine content Oral curcumin products vary in content and often contain piperine

Turmeric-associated liver injury follows a distinctive pattern mediated by an immunologic response. An analysis of an Italian plant surveillance database identified 37 cases of acute liver injury among 73 reports relating to Curcuma longa products. The US Drug-Induced Liver Injury Network has published its own analysis.

The uncomfortable part is that bioavailability enhancement and hepatotoxicity may not be separable. The formulations that produce measurable blood levels are also the ones delivering enough compound to matter, and piperine simultaneously slows hepatic metabolism.

• CaseTurmeric hepatotoxicity surveillance data summarised, citing Italian plant database and US DILIN analyses 37 acute liver injury cases among 73 Curcuma longa reports