The findings
Curcumin, the active compound in turmeric, reduces knee osteoarthritis pain in randomized trials, with pooled effects similar in size to standard pain drugs and fewer stomach complaints in head to head testing. It does not rebuild cartilage. Plain turmeric powder does not match the extracts used in trials.
Curcumin, the active compound in turmeric, reduces knee osteoarthritis pain in randomized trials, with pooled effects similar in size to standard pain drugs and fewer stomach complaints in head to head testing. It does not rebuild cartilage. Plain turmeric powder does not match the extracts used in trials.
What the trials show
The strongest single trial is also the most practical one. Kuptniratsaikul and colleagues randomized 367 patients with knee osteoarthritis to a Curcuma domestica extract at 1,500 mg per day or ibuprofen at 1,200 mg per day for four weeks. Pain, function, and total WOMAC scores improved in both groups, and the extract met the prespecified test for noninferiority against ibuprofen on all three scores. Abdominal pain and discomfort were less frequent in the extract group [1].
Single trials can mislead, so the pooled record matters more. Zhao and colleagues (2024) ran a Bayesian network meta-analysis of 23 randomized trials with 2,175 patients with knee osteoarthritis. Curcumin beat placebo on visual analogue pain scores, total WOMAC, and WOMAC pain, and it cut the use of rescue medication. Curcumin ranked at or near the top of the treatments compared for pain, close to NSAIDs, and it carried a lower risk of adverse events than NSAIDs alone in that analysis [2].
A second network meta-analysis tells the same story with a warning attached. Wai and colleagues (2025) pooled 17 randomized trials of turmeric products for knee osteoarthritis. Every preparation class reduced WOMAC pain against placebo. Bioavailability enhanced preparations cut WOMAC pain by a mean difference of 2.47 points in the pairwise analysis and reached the threshold the authors used for a clinically important difference. The warning: the authors rated the certainty of the evidence as low, and they called for better trials [3].
Two limits run through all of this work. First, no trial shows cartilage regrowth or a change in the course of osteoarthritis. The measured benefit is pain and function. Second, trial quality varies, formulations vary even more, and many trials are short.
The form problem
Turmeric root is only about 2 to 9 percent curcumin by weight, and plain curcumin is poorly absorbed. A spoon of spice in food does not reach the doses or blood levels in the trials above.
Trials use standardized extracts, usually around 500 to 1,000 mg of curcumin per day, often in a delivery system built to raise absorption: a phytosome such as Meriva, a turmeric oil combination such as BCM-95, or a nanoparticle form such as Theracurmin. No trial has randomized these forms against each other on knee pain, so no honest page can name a winning brand or form.
One form choice carries a safety point. Some products pair curcumin with piperine, which raises absorption in part by slowing drug clearing enzymes in the liver. If you take prescription drugs cleared by those enzymes, a piperine product is the riskiest way to buy curcumin. Talk to your doctor or pharmacist first.
What the experts say
Every quote below comes from our expert database, copied word for word, with its source linked. Positions without a verified quote are labeled as positions, not quotes.
Rhonda Patrick takes a phytosome form herself: "I take curcuma zorb from Pure Encapsulations. Which is the Meriva. So it's a curcumin phytosome, which is more bioavailable." (FoundMyFitness Q and A). https://www.foundmyfitness.com/episodes/qa-20-dr-rhonda-patrick
Kathryn Nelson, a medicinal chemist, is the named skeptic. In her 2017 review in the Journal of Medicinal Chemistry she concluded that curcumin is "an unstable, reactive, nonbioavailable compound and, therefore, a highly improbable lead" as a drug. Both readings can hold at once: a compound can fail as a drug lead and still produce a modest pain effect in trials. https://www.newscientist.com/article/2528418-do-turmeric-and-curcumin-have-any-actual-health-benefits/
Derek (More Plates More Dates) includes curcumin in his own daily list, describing himself as "taking 1,200 milligrams of curcumin daily" (YouTube). That dose sits at the top of the common trial range. https://www.youtube.com/watch?v=-EYlit7J1nk
Bryan Johnson takes 1 gram of turmeric at dinner and again before bed, per his published protocol summary, as an anti-inflammatory for brain and joint health (position, no verified quote on file for this item).
Examine.com covers curcumin for osteoarthritis pain as one of the better supported supplement uses, which matches our reading of the pooled trials. Our added layer is the grade below and the form warning above, because the product on a shelf is often not the product in a trial.
The split, plainly: users point to the trial record and their own results, the chemistry critique points to absorption and stability, and neither side claims cartilage repair. On pain alone, the trial record is the stronger argument.
Our grade and rationale
Longevity Lab Report grades curcumin B on evidence and B- on worth it, at $15 to $30 a month for a standard extract.
The B rests on the size of the trial base for a supplement: more than 100 randomized trials across uses in umbrella reviews, the strongest signal in knee osteoarthritis pain and post-exercise recovery, a noninferiority result against ibuprofen, and two network meta-analyses pointing the same way. It stops short of A because trial quality is uneven, samples in several areas are small, certainty ratings are low in the newest pooled work, and the absorption problem means results depend on a form most labels describe poorly.
The worth it B- reflects broad use, a real effect, and a mid range cost. It is the best supported joint supplement in our database, and it is still a pain intervention, not a repair.
How we grade: https://longevitylabreport.com/methodology Full record: https://longevitylabreport.com/ingredients/curcumin
Safety
- Stomach upset is the most common side effect, mainly at high doses.
- Use caution with blood thinners and antiplatelet drugs. A bleeding risk is theoretical, but it is a reason to ask, not guess.
- Piperine products can change how some prescription drugs are cleared. See the form section above.
- Liver injury is a rare reported risk with medicinal turmeric and curcumin products. Australia's Therapeutic Goods Administration advises that the risk may be higher with products that have enhanced absorption or bioavailability and/or higher doses, advises people with current or previous liver problems to avoid these medicinal products, and advises anyone taking them to stop and seek medical advice on symptoms such as jaundice, dark urine, nausea, vomiting, unusual tiredness, weakness, abdominal pain, or loss of appetite. The TGA notes the risk does not appear to relate to turmeric used in typical dietary amounts as a food.
- Turmeric powders have had lead contamination findings in the past. A tested extract is the safer buy.
- This page is information, not medical advice. Knee pain that follows an injury, locks, swells fast, or comes with fever needs a clinician, not a supplement trial.
Frequently asked questions
How much curcumin should I take for knee pain? The common trial range is 500 to 1,000 mg per day of a standardized extract. The ibuprofen comparison used 1,500 mg per day of a turmeric extract for four weeks. Whole turmeric powder does not reach these amounts of curcumin.
Is curcumin better than glucosamine for joints? On trial evidence, yes. Our database grades curcumin B and glucosamine C+. Glucosamine has a larger trial count and a weaker result, including a null result in its largest independent trial.
Limitations: what we do not know
- The newest network meta-analysis rated certainty as low. Better, larger, independent trials could move this grade either way.
- Most trials are short, and long term daily safety data are thinner than the marketing suggests.
- Formulations differ so much that a result for one extract does not prove a result for the bottle in your hand.
- Pain scores are subjective. Blinding helps, but no pooled result here rests on an objective joint measure.
- Several positive trials have industry ties to the extract tested. That does not erase the result, and it does lower confidence.
Sources
- Kuptniratsaikul V, et al. Efficacy and safety of Curcuma domestica extracts compared with ibuprofen in patients with knee osteoarthritis: a multicenter study. Clinical Interventions in Aging. 2014;9:451-458. Trial size n = 367, four weeks, as reported in the study.
- Zhao J, et al. Efficacy and safety of curcumin therapy for knee osteoarthritis: a Bayesian network meta-analysis. Journal of Ethnopharmacology. 2024;321:117493. 23 randomized trials, 2,175 patients. https://nutrition-evidence.com/article/13129/efficacy-and-safety-of-curcumin-therapy-for-knee-osteoarthritis-a-bayesian-network-meta-analysis
- Wai, et al. Effect of turmeric products on knee osteoarthritis: a systematic review and network meta-analysis. BMC Complementary Medicine and Therapies. 2025;25:292. 17 studies, certainty of evidence rated low by the authors. https://pmc.ncbi.nlm.nih.gov/articles/PMC12309109/
- Longevity Lab Report ingredient record: curcumin (evidence grade B, worth it B-, dose, benefits, safety, expert positions), last updated 2026-10-17. https://longevitylabreport.com/ingredients/curcumin
- Australian Government, Therapeutic Goods Administration. Medicines containing turmeric or curcumin: risk of liver injury. Safety alert, 2023. 18 reports of liver problems received to 29 June 2023, including two severe cases, one of them fatal.
Changelog
- 2026-11-11: Published. Panel review passed 2026-11-11 (research/article-panel-review-2026-11-11.md); Humanizer 94 the same day (research/article-humanizer-2026-11-11.md).
- 2026-11-12: Safety section gains the Australian Therapeutic Goods Administration's liver-injury alert for medicinal turmeric and curcumin products, added after publication.
Supplements mentioned
Grades are research summaries, not medical advice. Decisions about your own health belong with your clinician.