Curcumin benefits

How should curcumin benefits be read?

An important distinction: a single measure (a biomarker such as CRP or HbA1c) and an outcome a person feels (pain, depression symptoms) are not the same step of evidence. In addition, the product used in curcumin studies is generally an extract giving curcuminoids at the gram level; the turmeric used in the kitchen is not the same thing (for this difference, see the turmeric page). Safety topics and liver injury reports are in the main guide.

Is curcumin good for knee osteoarthritis pain?

In studies comparing with placebo, a reduction in pain has been reported; however, the evidence is short-term, heterogeneous, and of low to moderate quality.

Wang et al. (2021). This systematic review and meta-analysis of turmeric extracts in adults with knee osteoarthritis included 16 randomized trials lasting up to 16 weeks and covering a total of 1,810 adults. In eleven trials the comparison was with placebo, and in five with NSAIDs (non-steroidal anti-inflammatory drugs). Compared with placebo, turmeric extracts reduced knee pain (SMD −0.82; 95% CI −1.17 to −0.47; I² = 86.23%) and physical function (SMD −0.75; 95% CI −1.18 to −0.33; I² = 90.05%), and showed an effect similar to NSAIDs. Body mass index explained a large part of the variation: improvement is smaller in those with higher BMI. No significant difference between groups was reported in biochemical indicators or imaging results. The adverse event rate was 12% lower than with NSAIDs and similar to placebo. The authors described turmeric extract as a safe and effective option compared with placebo or NSAIDs, and in the same place stated that the evidence from short-term studies is heterogeneous and carries a moderate risk of bias, so there is uncertainty about the true effect. ^[1]

Feng et al. (2022). This meta-analysis, which isolated applications giving curcuminoids alone (not mixed with other active substances), included 15 trials and 1,670 patients. Compared with placebo, the VAS pain score (WMD −1.77; 95% CI −2.44 to −1.09) and the WOMAC pain score (WMD −1.42; 95% CI −2.41 to −0.43) were found to be lower, and curcuminoids were not inferior to NSAIDs. The frequency of adverse events did not increase compared with placebo (RR 1.03; 95% CI 0.69–1.53). The authors emphasize that, because of the low quality of the studies and marked heterogeneity, a “cautious and conservative” recommendation should be made. The study authors declared no conflict of interest. ^[2]

The two meta-analyses are not independent confirmation of each other; it is possible that they include some of the same primary studies. The findings were obtained in adult patients with knee osteoarthritis with use lasting weeks; use lasting years and hip or hand osteoarthritis are not answered by these data. Wang et al. found no significant difference between groups in imaging and biochemical outcomes.

Does curcumin reduce inflammation (CRP)?

CRP is a marker of inflammation in the blood; its reduction does not mean that a disease outcome has improved. In the meta-analysis by Gorabi et al. (2022), in people with autoinflammatory/proinflammatory conditions, there were 9 studies for CRP and 23 for hs-CRP. CRP fell significantly compared with placebo (WMD −3.67 mg/L; 95% CI −6.96 to −0.38; p = 0.02); hs-CRP was also significantly lower than in the placebo group. Subgroup results varied by dose and duration and are not in the same direction for CRP and hs-CRP (a significant effect was seen for CRP with durations longer than 10 weeks and for hs-CRP with durations of 10 weeks or shorter). The authors state that curcumin appears useful in reducing CRP and hs-CRP in proinflammatory conditions but limit the result to these conditions. ^[3] This is a biomarker finding; it does not show that clinical outcomes such as symptoms, disease flares, or cardiovascular events improved in patients.

Does curcumin lower blood sugar?

Bahari et al. (2026) pooled 34 randomized trials (39 treatment arms) examining curcumin/turmeric supplementation in adults with prediabetes or type 2 diabetes. Compared with control groups, fasting glucose fell by 10.15 mg/dL (95% CI −12.59 to −7.72), HbA1c by 0.32 percentage points (95% CI −0.43 to −0.21), and HOMA-IR by 0.46 (95% CI −0.60 to −0.32). In subgroup analyses, the effect varied by health status (type 2 diabetes versus prediabetes), dose, and formulation; the authors state that, because of marked heterogeneity between studies, it should be interpreted carefully and that higher-quality trials are needed. ^[4] The mean reduction of about 0.3 points in HbA1c is a small effect; it is not a finding that replaces drug treatment.

Is curcumin good for depression?

For the meta-analysis by Fusar-Poli et al. (2020), 930 articles were screened and nine were found eligible; in seven, participants were people with major depressive disorder, and in two, people with depression secondary to a medical illness. For depressive symptoms, Hedges g was −0.75 (95% CI −1.11 to −0.39) in 10 studies with 531 participants, and for anxiety symptoms g was −2.62 (95% CI −4.06 to −1.17) in 5 studies with 284 participants; the authors described the effect as “large.” However, the authors state that curcumin may have improved depressive and anxiety symptoms when added to standard care, that the results should be interpreted cautiously because of the small sample size, and that new trials are needed, especially in Western countries where curcumin is not part of the usual diet. ^[5] This finding does not mean that curcumin could replace depression treatment.

Does curcumin work in ulcerative colitis?

Lang et al. (2015) conducted a multicenter, randomized, placebo-controlled, double-blind study in 50 people with mild-to-moderate active ulcerative colitis who were on mesalamine and had not responded to an additional two weeks of maximum-dose mesalamine (oral and topical). While continuing mesalamine, 3 g of curcumin per day (n = 26) or placebo (n = 24) was given for one month. In the intention-to-treat analysis, clinical remission at week 4 (SCCAI ≤ 2) was seen in 14 people (53.8%) in the curcumin group and in no one in the placebo group (OR 42; 95% CI 2.3–760). Clinical response was 65.3% versus 12.5%, and endoscopic remission was found in 8 of 22 evaluated patients (38%) in the curcumin group and in none of 16 patients in the placebo group. Adverse effects were infrequent and similar between groups. ^[6]

This is a single, small trial lasting one month; the confidence interval is very wide (2.3–760). The result applies only to patients with mild-to-moderate active disease who are already taking mesalamine and are unresponsive to it, and it does not show that it can be used in place of drug treatment. For this reason, the evidence grade is rated at most C.

Does curcumin work in cancer treatment?

Gutsche et al. (2025) systematically reviewed 34 randomized trials (2,580 patients in total) examining adjunctive curcumin treatment in adults receiving cancer treatment; most patients had head and neck cancer, followed by breast, prostate, and colorectal cancers. Results for oral and skin symptoms, pain, weight and body composition, survival, and disease progression were heterogeneous; significant findings were reported for oral mucositis and weight loss. The risk of bias is moderate to high in all studies, and in one study markedly more vomiting was reported in the curcumin group. The authors concluded that, because of heterogeneous results and methodological limits, no clear statement can be made on the efficacy of curcumin treatment in cancer patients. ^[8] So curcumin is not proven to prevent or treat cancer; evidence is limited even for supportive outcomes such as mucositis and weight loss.

Does curcumin protect against Alzheimer’s disease?

Ringman et al. (2012) conducted a 24-week randomized, double-blind, placebo-controlled study in 36 people with mild-to-moderate Alzheimer’s disease; participants received placebo, 2 g per day, or 4 g per day of a standardized curcumin extract. The 30 people who completed the study had a mean age of 73.5 and a mean MMSE of 22.5. One person (8%) in the placebo group and 5 of 24 (21%; three because of digestive complaints) in the curcumin group left the study. A fall in hematocrit and a rise in glucose were seen but were found to be clinically insignificant. There was no difference between groups in clinical or biomarker efficacy; the level of native (unchanged) curcumin measured in plasma was low (7.32 ng/mL). The authors could not demonstrate evidence of efficacy and point to the limited bioavailability of curcumin. The product and placebo were provided by the manufacturing company; the study was supported by foundations and grants from the US National Institute on Aging. ^[7]

This is a single, small trial; no clinical or biomarker benefit was shown in the trial, and plasma curcumin levels remained low. There is no controlled evidence showing a benefit of curcumin in Alzheimer’s disease.

Summary of the evidence

This guide does not recommend a personal dose. Because the products used vary from study to study, the dose or product in one study cannot be carried directly to another product.