Evidence guide · UK
Turmeric for knee osteoarthritis: what the November 2026 Cochrane review found
A Cochrane review published on 19 November 2026 found turmeric may make little to no difference to knee osteoarthritis pain, function or quality of life over three months. The numbers, the liver-injury reports and what NICE recommends instead.
Review type: research summary. This page summarises published research and product labels, cited below.
Turmeric and its curcuminoids, including curcumin, are widely promoted as supplements for osteoarthritis. A Cochrane review published on 19 November 2026 pooled the randomised trials that compared turmeric extracts or curcuminoids with placebo in osteoarthritis. Its conclusion is modest: turmeric may make little to no difference to pain, function or quality of life over three months, although more people on turmeric rated their treatment a success.
Short answer: in knee osteoarthritis, turmeric lowered pain scores by less than a point on a 0 to 10 scale compared with placebo, which falls short of the 1.5-point change the reviewers treat as clinically meaningful. Every trial lasted three months or less, every trial was in knees, and the certainty of the evidence is low. It is very uncertain whether side effects differ from placebo in short trials, but rare and sometimes serious liver injury has been reported with turmeric supplements, especially high-dose or enhanced-absorption products. Therapeutic exercise and weight management, where appropriate, are core treatments in NICE guidance, which also says to offer a topical NSAID for knee osteoarthritis.
Grades used here: Strong (consistent results from several good-quality human trials), Moderate (good trials, but few or with mixed results), Limited (small or low-quality human studies, or good studies in other groups of people), Insufficient (too little human research to judge) and Evidence of harm.
What the Cochrane review included
Shirinsky and colleagues searched CENTRAL, MEDLINE, five other databases and two trial registers up to 7 November 2025, with no language restrictions. They included randomised trials of oral Curcuma longa extracts or curcuminoids, with or without ingredients that improve absorption, and excluded combination products containing other active herbal compounds. Each eligible trial was screened for trustworthiness with the Cochrane Pregnancy and Childbirth Trustworthiness Screening Tool, and bias was assessed with Cochrane's RoB 2 tool.
Nine trials with 1,245 participants met the criteria. All were in knee osteoarthritis, the mean age of participants ranged from 52 to 62, and 67% were women. Every trial lasted 90 days or less. The products varied: one trial used a formulation with enhanced bioavailability, two used extracts with essential oils, and four used standardised or rhizome extracts with curcuminoid content from 23% to 88%. Four trials compared turmeric with placebo (379 participants), four with non-steroidal anti-inflammatory drugs (673 participants) and one with paracetamol (193 participants). The published summary limits its reporting to the placebo comparison, the review's primary comparison.
What it found against placebo
Pain: over three months, pain on a 0 to 10 scale fell by 1.57 points with placebo and by a further 0.72 points with turmeric (95% CI 1.39 to 0.06 points lower; 4 trials, 349 participants). The reviewers note this did not reach the clinically significant threshold of 1.5 points.
Function and quality of life: on a 0 to 100 function scale, where 0 is best, scores improved by 5.42 points with placebo and 3.5 points more with turmeric, a difference whose confidence interval crosses zero (7.29 lower to 0.29 higher; 348 participants). Quality of life, on a 0 to 100 scale where 100 is best, rose 9.83 points with placebo and 2.19 points more with turmeric (1.56 lower to 5.94 higher; 291 participants).
Treatment success: in two trials with 170 participants, 191 per 1,000 people on turmeric reported treatment success against 107 per 1,000 on placebo (risk ratio 1.78, 95% CI 1.19 to 2.67).
The reviewers graded the benefit evidence as low certainty, downgraded for indirectness and publication bias. Their conclusion: compared with placebo, Curcuma longa may result in little to no improvement in pain, function or quality of life, and may improve participant-reported global assessment of success in people with knee osteoarthritis. They call for trials longer than three months and for trials in hand and hip osteoarthritis.
Side effects in the trials
Adverse events were reported by 226 per 1,000 people on turmeric and 224 per 1,000 on placebo (risk ratio 1.01, 0.68 to 1.51; 4 trials, 379 participants). Withdrawals because of adverse events were 66 per 1,000 on turmeric against 46 per 1,000 on placebo (risk ratio 1.45, 0.61 to 3.43; 3 trials, 321 participants). Three trials reported no serious adverse events, and one reported two in the placebo group that were unrelated to the study. The reviewers are very uncertain about all of these harm estimates, which they graded as very low certainty. Trials of a few hundred people over three months are too small and too short to pick up a rare reaction.
A 2020 placebo-controlled trial
A 2020 trial in the Annals of Internal Medicine randomised 70 people in southern Tasmania with knee osteoarthritis and fluid-related swelling in the joint (effusion-synovitis) to two capsules a day of Curcuma longa extract or placebo for 12 weeks. Knee pain on a 100 mm visual analogue scale fell by 9.1 mm more with the extract (95% CI 17.8 to 0.4 mm lower), and pain on the WOMAC scale also improved, but the extract did not change the volume of joint swelling on MRI or cartilage composition. Adverse events occurred in 39% of the extract group and 53% of the placebo group. The authors list a modest sample size and short duration as limitations. The trial was funded by the University of Tasmania and Natural Remedies Private Limited. A 9 mm change on a 100 mm scale is of the same order as the Cochrane estimate.
Liver injury: rare but documented
The US Drug-Induced Liver Injury Network reviewed its cases from 2004 to 2022 and identified 10 in which turmeric was implicated, all enrolled since 2011 and six since 2017. Eight of the 10 were women and the median age was 56. Five people were admitted to hospital and one died of acute liver failure. Turmeric was confirmed in all seven products analysed, and three of those also contained piperine (black pepper). Seven of the 10 carried the immune gene variant HLA-B*35:01, and the injury typically appeared one to four months after starting. The authors say turmeric liver injury appears to be increasing in the United States, perhaps reflecting usage patterns or more frequent combination with black pepper.
An analysis of Italian phytovigilance reports described seven cases of acute hepatitis in Tuscany up to September 2019, all associated with high-bioavailability formulations and high doses of curcumin or curcuminoids, and identified 23 further published cases. The US National Center for Complementary and Integrative Health says conventionally formulated oral turmeric or curcumin is likely safe in recommended amounts for up to two or three months, that liver damage has been reported with some formulations designed for higher absorption, and that turmeric supplements may be unsafe in pregnancy. It advises stopping turmeric products immediately and seeing a healthcare provider if symptoms such as fatigue, nausea, poor appetite, dark urine or jaundice appear.
Knee osteoarthritis pain: Limited
Four small placebo-controlled trials show a pain reduction smaller than the clinically meaningful threshold, with low-certainty evidence and no trial longer than three months. That supports a Limited grade for a small short-term effect on knee pain. For hip and hand osteoarthritis, and for use beyond three months, the grade is Insufficient because there are no trials in the review.
What NICE and the NHS recommend instead
NICE guideline NG226 says the core treatments for osteoarthritis are therapeutic exercise and weight management where appropriate, along with information and support. It recommends offering a topical NSAID for knee osteoarthritis and considering an oral NSAID if topical treatment is ineffective or unsuitable, taking account of other risks. NICE says not to offer glucosamine. The NHS osteoarthritis page says NICE recommends that supplements should not routinely be offered on the NHS. NG226 does not make a recommendation on turmeric.
Our view
- Turmeric is not a substitute for therapeutic exercise, weight management or a topical NSAID. The best trial evidence suggests any pain benefit is small.
- If you want to try it, the trials ran for up to three months. Without a noticeable difference by then, the evidence does not support continuing for knee pain.
- Avoid high-dose, enhanced-absorption curcumin products, including those with black pepper extract, unless a doctor or pharmacist has agreed it suits you. The Italian liver-injury cases and the NCCIH safety advice both point to these formulations.
- Stop and seek medical advice if you develop fatigue, nausea, dark urine or yellowing of the skin or eyes while taking turmeric.
- Ask a pharmacist before combining turmeric supplements with regular medicines, and do not use them in pregnancy unless your doctor or midwife agrees.
UK rules
In the UK, turmeric capsules are usually sold as food supplements. Food supplements are not medicinal products and should not be sold with a claim to prevent, treat or cure a medical condition, including a claim that turmeric treats arthritis. If you take regular medicines, check with your GP or another healthcare professional before starting a turmeric supplement. If a food supplement makes you unwell, stop it and contact a healthcare professional; concerns about its safety, labelling or claims can be reported to your local authority, which reports food safety incidents to the Food Standards Agency.
FAQ
Does turmeric help knee osteoarthritis?
Only a little, on current evidence. The November 2026 Cochrane review found turmeric lowered pain by 0.72 points on a 0 to 10 scale more than placebo over three months, below the 1.5-point threshold it treats as clinically meaningful, with low-certainty evidence. More people on turmeric rated their treatment a success.
Is turmeric as good as ibuprofen for knee pain?
The Cochrane review included four trials against NSAIDs and one against paracetamol, but its published summary reports only the placebo comparison, so it does not give a pooled answer on that question here. For knee osteoarthritis, NICE says to offer a topical NSAID and to consider an oral NSAID only if topical medicines are ineffective or unsuitable.
Is turmeric safe for the liver?
It is very uncertain whether side effects in short trials differ from placebo, but rare cases of serious liver injury have been reported, including one death in a US case series. Reports cluster around high-dose and enhanced-absorption products, some containing black pepper extract. Stop and seek advice if you notice symptoms such as dark urine or jaundice.
How long should I try turmeric for?
The trials lasted up to 90 days, and the reviewers call for longer studies. There is no evidence on longer use for osteoarthritis.
Sources
These are the documents this page relies on.
- Shirinsky I, Shirinsky V, Filatova K, et al. Curcuma longa (turmeric) or its active ingredients for osteoarthritis. Cochrane Database Syst Rev. 2026;11:CD014683 (published 19 November 2026). doi:10.1002/14651858.CD014683.pub2Cochrane Database of Systematic Reviews (PubMed 43051256)
- Wang Z, Jones G, Winzenberg T, et al. Effectiveness of Curcuma longa extract for the treatment of symptoms and effusion-synovitis of knee osteoarthritis: a randomized trial. Ann Intern Med. 2020;173(11):861-869Annals of Internal Medicine (PubMed 32926799)
- Halegoua-DeMarzio D, Navarro V, Ahmad J, et al. Liver injury associated with turmeric: a growing problem. Ten cases from the Drug-Induced Liver Injury Network [DILIN]. Am J Med. 2023;136(2):200-206The American Journal of Medicine (PubMed 36252717)
- Lombardi N, Crescioli G, Maggini V, et al. Acute liver injury following turmeric use in Tuscany: an analysis of the Italian Phytovigilance database and systematic review of case reports. Br J Clin Pharmacol. 2021;87(3):741-753British Journal of Clinical Pharmacology (PubMed 32656820)
- Turmeric: usefulness and safety (last updated April 2025)National Center for Complementary and Integrative Health, NIH
- Osteoarthritis in over 16s: diagnosis and management (NG226), recommendationsNational Institute for Health and Care Excellence
- Osteoarthritis: treatment and supportNHS
- Food supplements guidanceGOV.UK