Turmeric reduces inflammation

In randomized trials, turmeric and curcumin supplements have lowered some blood markers of inflammation on average, with results that vary by marker, product and population. A 2023 pooled analysis of 66 trials found lower C-reactive protein (CRP), TNF-alpha and IL-6 with supplementation and no significant change in IL-1 beta. Its average CRP difference was 0.58 mg/L.

Those averages rest mostly on small trials, and not every analysis points the same way or reads the evidence as equally firm. A 2024 analysis of 103 placebo-controlled trials across 42 health measures concluded that curcumin can change inflammatory markers and rated its evidence on CRP as high in credibility; its published summary gives no figure for the size of the change. A 2025 review of 28 trials in prediabetes and diabetes found reductions but rated its certainty low for every marker and detected signs of publication bias for CRP. A 2019 review of 19 trials in people with chronic inflammatory diseases found no statistically significant reduction in any marker it pooled. In a blinded, placebo-controlled trial of 70 adults with knee osteoarthritis, knee pain improved with a turmeric extract, while fluid and inflammation inside the knee on MRI did not change, and blood CRP, IL-6 and TNF-alpha did not differ from placebo.

Knee osteoarthritis is where much of the clinical research sits. A 2025 review of 21 trials with 1,705 patients found lower CRP and TNF-alpha with curcumin, and no significant difference in ESR, IL-1 beta, IL-6 or PGE2. A lower blood marker is a laboratory reading; the trials pooled in these reviews were not designed to show whether that change alters how a condition progresses.

The products tested differ widely: ground turmeric, standardised extracts, and formulations built to raise absorption, such as curcumin combined with piperine from black pepper, at a wide range of doses. Curcumin on its own is poorly absorbed and chemically unstable, and a 2017 medicinal chemistry review describes it as a compound that tends to register activity in laboratory tests without a specific target. In the diabetes review, larger CRP reductions appeared with unformulated curcumin and with doses of 1 g a day or more.

On safety, NCCIH describes conventionally formulated oral turmeric or curcumin as likely safe in usual amounts for two to three months, with digestive upset as the common side effect. It reports liver injury in some people who took enhanced-absorption curcumin products, and describes turmeric supplements as possibly unsafe in pregnancy. The NIH liver injury database rates turmeric a well-documented cause of liver injury and the most common cause of herbal-related liver injury in the United States. Most cases involved highly absorbable curcumin products, though cases have also followed plain ground turmeric and turmeric tea, and it estimates such injury as very rare, roughly 1 in 10,000 to 1 in 100,000 people exposed. The EU herbal medicines committee recognises turmeric only for mild digestive complaints, on the basis of long-standing use, and lists no inflammatory use. NCCIH also notes that some herbs and medicines interact.