Does turmeric reduce inflammation and joint pain?

As a supplement, a little, and the evidence behind it is weak. In 15 randomized trials with 1621 people who had osteoarthritis, turmeric extract or curcumin eased pain against placebo by a moderate standardized amount (SMD 0.66, 95% CI 0.43 to 0.88). Across 66 trials in people with a mix of conditions, the same supplements lowered the blood marker CRP by 0.58 mg/L. Every one of these trials tested capsules or extracts. Nobody has tested turmeric as a kitchen spice.

Unsettled. The trials point the same way, and the reviews that graded them call most of the evidence very weak. A 2025 overview of 7 systematic reviews on knee osteoarthritis rated 37 of 48 outcomes as extremely low quality evidence, and only 5 as medium. A review of 69 supplement trials for osteoarthritis found large short-term pain relief from Curcuma longa extract and curcumin, and no supplement at all with a clinically important effect on pain in the long term.

What the trials found

Joint pain in osteoarthritis

The fullest pooling is a 2021 meta-analysis of 15 trials with 1621 adults, which we read in full. The trials ran mostly 4 to 16 weeks with 40 to 400 people each, many of them in India, Iran and Thailand. Against placebo, pain on the WOMAC scale fell by an SMD of 0.66, and physical function improved by an SMD of 0.79 (0.31 to 1.27). An SMD is a standardized score rather than a count of points, so here is the same thing in plain units from one trial.

In Tasmania, 70 people with painful knee osteoarthritis and fluid in the joint took a Curcuma longa extract or placebo for 12 weeks. Pain fell by 9.1 mm more on a 100 mm scale (95% CI 0.4 to 17.8), an interval that comes close to zero. The swelling in the joint, measured by ultrasound, did not change. The extract maker part-funded the trial.

A 2025 network analysis of 39 trials with 4599 people with knee osteoarthritis compared seven supplements. Curcumin helped on some outcomes. Boswellia ranked best for pain and stiffness. The abstract gives no curcumin number to put next to it, and the comparisons are indirect.

Against painkillers

In the 2021 meta-analysis, turmeric extract or curcumin relieved osteoarthritis pain about as much as anti-inflammatory painkillers such as ibuprofen (WOMAC pain SMD 0.04, 95% CI minus 0.18 to 0.25), with fewer adverse events. That rests on few head-to-head trials. The largest, in 367 Thai patients with knee osteoarthritis, found the extract no worse than ibuprofen after 4 weeks, and abdominal pain or discomfort was more common on ibuprofen. It had no placebo arm and lasted only 4 weeks.

Inflammation markers in the blood

A 2023 meta-analysis of 66 trials found turmeric or curcumin lowered CRP by 0.58 mg/L (0.41 to 0.74), TNF-alpha by 3.48 pg/mL and IL-6 by 1.31 pg/mL. IL-1 beta did not change significantly. The trials varied in dose, product and disease, and the abstract gives neither a total of participants nor a quality grade. An umbrella review of 21 meta-analyses, which share many trials, put the CRP drop at 0.87 mg/L (0.59 to 1.14). In 472 adults with type 2 diabetes across 6 trials, hs-CRP fell by 1.06 mg/L (0.16 to 1.97). These are changes in a lab marker. None of these trials measured whether people felt better or got sick less.

Rheumatoid arthritis

A systematic review of spice trials in inflammatory arthritis found only two curcumin trials in rheumatoid arthritis, both low quality, and drew no conclusion about disease activity.

Who should be careful

Not for everyone. Concentrated curcumin can injure the liver. The US Drug-Induced Liver Injury Network described 10 cases linked to turmeric: 5 people were hospitalized and 1 died of acute liver failure, the injury appeared after 1 to 4 months, and it was strongly linked to the gene variant HLA-B*35:01. That is a case series with no denominator, so the risk per user is unknown, and 3 of the 7 products tested also contained piperine.

Italian phytovigilance reports add 7 cases of acute hepatitis in Tuscany, from spontaneous reports with no denominator. In every case the product was a high-bioavailability, high-dose curcumin formulation. The NIH's complementary health center (NCCIH) says the same thing about the market: many curcumin products are made to be better absorbed, and liver damage has been reported in some people who took them.

NCCIH also says turmeric supplements during pregnancy may be unsafe. That statement is about supplements, not cooking.

One case report describes a 66-year-old man who developed iron deficiency anemia while taking six turmeric extract capsules a day for osteoarthritis. His hemoglobin returned to normal 2 weeks after he stopped. It is a single case, and the proposed mechanism, turmeric binding iron in the gut, has been shown only in mice.

We found no card on turmeric and blood thinners or other medicines. Our interaction database and the NCCIH page named none, which is an open question rather than a reassurance.

What expert bodies say

NCCIH calls the initial evidence for osteoarthritis positive and says higher-quality evidence is needed before anyone can reach definitive conclusions. It also says more research is needed on how the better-absorbed formulations change curcumin's effects. We did not find a position from a European body that we could read directly.

How we searched

What would change this answer

• Long trials. The pain benefit is short term, and no supplement in the 69-trial review held a clinically important effect on pain over the long term. A year-long placebo-controlled trial of a standard turmeric extract would settle more than another round of short ones.
• Better trials of the same question. With 37 of 48 outcomes rated extremely low quality, larger trials with low risk of bias, run outside the companies that sell the extracts, are what is missing. Our registry search found no running trial of curcumin for knee osteoarthritis.
• A denominator for the liver risk. Case series show that liver injury happens and who it clusters in. They cannot say how often. Data on how many people took these products would turn a warning into a risk.
• Any trial of turmeric as food. All the evidence above is about extracts and curcumin capsules. The spice in curry is untested for joint pain or inflammation.

Nutrition values and other findings for the spice itself are on the turmeric page.

Sources

1. ev-turinf-01 · Meta-analysis or systematic review · systematic review and meta-analysis of RCTs · n = 1621
   The meta-analysis results of each subgroup showed that: (1) compared with placebo, Curcuma longa extract and curcumin can reduce WOMAC score-pain (SMD: −0.66, 95% CI: −0.88 to −0.43, P<0.00001; fixed-effects model).
   
   Who: adults with osteoarthritis in randomized trials of Curcuma longa extract or curcumin versus placebo
   Effect: WOMAC pain SMD -0.66 (95% CI -0.88 to -0.43; I2 34%); VAS pain WMD -11.55 (-14.3 to -9.06); WOMAC function SMD -0.79 (-1.27 to -0.31)
   Certainty: GRADE assessed by the authors; trials of 40 to 400 people, mostly 4 to 16 weeks, many from India, Iran and Thailand; high heterogeneity for VAS and function.
   Biosci Rep, 2021 · checked 2026-10-06 · we read the full_text
2. ev-turinf-02 · Meta-analysis or systematic review · systematic review and meta-analysis of RCTs · n = —
   Compared with non-steroidal anti-inflammatory drugs (NSAIDs), Curcuma longa extract and curcumin have similar effects on joint pain, function and stiffness.
   
   Who: adults with osteoarthritis in randomized trials comparing Curcuma longa extract or curcumin with NSAIDs
   Effect: VAS vs NSAIDs WMD -0.34 (95% CI -1.25 to 0.57); WOMAC pain vs NSAIDs SMD 0.04 (-0.18 to 0.25); adverse events lower than with NSAIDs
   Certainty: Equivalence rests on few head-to-head trials; n is the whole review, not the NSAID subgroup.
   Biosci Rep, 2021 · checked 2026-10-06 · we read the abstract
3. ev-turinf-03 · Meta-analysis or systematic review · systematic review and meta-analysis of RCTs · n = —
   We observed that turmeric/curcumin supplementation significantly reduces levels of inflammatory markers, including CRP (WMD: -0.58 mg/l, 95 % CI: -0.74, -0.41), TNF-α (WMD: -3.48 pg/ml, 95 % CI: -4.38, -2.58), and IL-6 (WMD: -1.31 pg/ml, 95 % CI: -1.58, -0.67); except for IL-1β (WMD: -0.46 pg/ml, 95 % CI: -1.18, 0.27) for which no significant change was found.
   
   Who: adults in randomized controlled trials of turmeric or curcumin supplements, mixed health conditions
   Effect: CRP WMD -0.58 mg/L (95% CI -0.74 to -0.41); TNF-alpha -3.48 pg/mL (-4.38 to -2.58); IL-6 -1.31 pg/mL (-1.58 to -0.67); IL-1 beta -0.46 pg/mL (-1.18 to 0.27)
   Certainty: Abstract does not give total participants or quality grading; trials vary in dose, formulation and disease.
   Cytokine, 2023 · checked 2026-10-06 · we read the abstract
4. ev-turinf-04 · Meta-analysis or systematic review · umbrella review of meta-analyses of randomized clinical trials · n = —
   Pooled estimates of 21 meta-analyses revealed that curcumin significantly reduced CRP (weighted mean difference (WMD) = -0.87; 95 % CI: - 1.14, - 0.59, P< 0.001), tumor-necrosis factor-alpha (TNF-α) (WMD = -2.72; 95 % CI: -4.05, -1.38; P< 0.001), interleukin-6 (IL-6) (WMD = -0.97, 95 % CI: -1.40, -0.54; P< 0.001), malondialdehyde (MDA) (Effect size (ES) = -0.81; 95 % CI: -1.39, -0.23, P = 0.006) and pulse wave velocity (PWV) (WMD = -45.60; 95 % CI: -88.16, -3.04, P = 0.036), and increased flow-mediated dilation (FMD) (WMD = 1.64, 95 % CI: 1.06, 2.22, P < 0.001), catalase (CAT) (WMD = 10.26; 95 % CI: 0.92, 19.61, P= 0.03), glutathione peroxidase (GPx) (WMD = 8.90; 95 % CI: 6.62, 11.19, P <0.001), and superoxide dismutase (SOD) levels (WMD = 20.51; 95 % CI: 7.35, 33.67, P= 0.002 and SMD = 0.82; 95 % CI: 0.27, 1.38, P= 0.004).
   
   Who: adults in meta-analyses of randomized clinical trials of curcumin supplementation
   Effect: CRP WMD -0.87 (95% CI -1.14 to -0.59); TNF-alpha WMD -2.72 (-4.05 to -1.38); IL-6 WMD -0.97 (-1.40 to -0.54)
   Certainty: Pooled meta-analyses overlap in the trials they include; registered in PROSPERO (CRD42024539018).
   Prostaglandins Other Lipid Mediat, 2024 · checked 2026-10-06 · we read the abstract
5. ev-turinf-05 · Meta-analysis or systematic review · systematic review with random effects meta-analysis and meta-regression · n = —
   No supplements were identified with clinically important effects on pain reduction at long term.
   
   Who: patients with hand, hip or knee osteoarthritis in placebo-controlled randomized trials of oral supplements
   Effect: effect size >0.80 for pain at short term for Curcuma longa extract and curcumin; no clinically important effect on pain at long term for any supplement; evidence quality very low to high
   Certainty: Few trials and participants per turmeric product; GRADE very low for the overall analysis.
   Br J Sports Med, 2018 · checked 2026-10-06 · we read the abstract
6. ev-turinf-06 · Meta-analysis or systematic review · systematic review and Bayesian network meta-analysis of RCTs · n = 4599
   Curcumin, collagen, ginger, and krill oil also demonstrated benefits in some outcomes.
   
   Who: adults with knee osteoarthritis in randomized trials of seven supplements versus placebo
   Effect: curcumin benefit in some WOMAC or VAS outcomes; Boswellia WOMAC pain MD 10.58 (95% CI 6.45 to 14.78); no supplement raised adverse events versus placebo
   Certainty: Indirect comparisons; the abstract gives no curcumin-specific numbers.
   Nutrients, 2025 · checked 2026-10-06 · we read the abstract
7. ev-turinf-07 · Meta-analysis or systematic review · overview of systematic reviews (AMSTAR 2, ROBIS, GRADE) · n = —
   Among 48 outcomes assessed, evidence quality was mostly low, with 5 medium-quality, 6 low-quality, and 37 extremely low-quality evidences.
   
   Who: systematic reviews and meta-analyses of curcumin for knee osteoarthritis up to January 2025
   Effect: 7 reviews; 5 moderate, 6 low and 37 extremely low quality outcomes of 48 (GRADE); high risk of bias in 4 reviews (ROBIS)
   Certainty: Newest synthesis (2025), outside the local corpus snapshot; abstract only.
   Front Pharmacol, 2025: A critical review of systematic reviews and meta-analyses of curcumin for knee osteoarthritis · checked 2026-10-06 · we read the abstract
8. ev-turinf-08 · Meta-analysis or systematic review · systematic review and meta-analysis of RCTs · n = 472
   Curcumin supplementation was associated with lower hs-CRP concentrations (MD = -1.06 mg/L; 95% CI -1.97 to -0.16; p = 0.030; I2 = 17.3%), although sensitivity analyses involving unit-ambiguous or statistically reconstructed data increased uncertainty.
   
   Who: adults with type 2 diabetes in randomized trials of oral curcumin or curcuminoids
   Effect: hs-CRP MD -1.06 mg/L (95% CI -1.97 to -0.16; I2 17.3%); TNF-alpha Hedges g -1.20 (-1.98 to -0.42; I2 84.9%; 5 RCTs, 475 people)
   Certainty: Small evidence base; sensitivity analyses with unit-ambiguous data widened uncertainty.
   Nutrients, 2026 · checked 2026-10-06 · we read the abstract
9. ev-turinf-09 · Meta-analysis or systematic review · systematic literature review of RCTs · n = —
   No conclusion on the impact of curcumin supplementation on RA activity could be drawn due to low-quality studies.
   
   Who: patients with rheumatoid arthritis in randomized trials of spice supplementation
   Effect: no conclusion on curcumin and rheumatoid arthritis activity (2 low-quality trials)
   Certainty: No pooled estimate; low-quality trials.
   Nutrients, 2020 · checked 2026-10-06 · we read the abstract
10. ev-turinf-10 · Randomized controlled trial(s) · randomized, double-blind, placebo-controlled trial · n = 70
    CL improved VAS pain compared with placebo by -9.1 mm (95% CI, -17.8 to -0.4 mm [P = 0.039]) but did not change effusion-synovitis volume (3.2 mL [CI, -0.3 to 6.8 mL]).
    
    Who: 70 participants with symptomatic knee osteoarthritis and ultrasonography-defined effusion-synovitis, southern Tasmania
    Effect: VAS pain -9.1 mm (95% CI -17.8 to -0.4); effusion-synovitis volume 0.65 tsp (-0.06 to 1.4; metric: 3.2 mL, -0.3 to 6.8), no change; 2 capsules daily for 12 weeks
    Certainty: Single center, modest size, 12 weeks; part-funded by the extract maker.
    Ann Intern Med, 2020: Effectiveness of Curcuma longa Extract for the Treatment of Symptoms and Effusion-Synovitis of Knee Osteoarthritis · checked 2026-10-06 · we read the abstract
11. ev-turinf-11 · Randomized controlled trial(s) · randomized multicentre noninferiority trial · n = 367
    However, the number of events of abdominal pain/discomfort was significantly higher in the ibuprofen group than that in the C. domestica extracts group (P=0.046).
    
    Who: 367 primary knee osteoarthritis patients with a pain score of 5 or higher, multicentre, Thailand
    Effect: WOMAC total, pain and function noninferior to ibuprofen at week 4 (P=0.010, 0.018, 0.010); abdominal pain or discomfort more frequent on ibuprofen (P=0.046)
    Certainty: Only 4 weeks; no placebo arm.
    Clin Interv Aging, 2014: Efficacy and safety of Curcuma domestica extracts compared with ibuprofen in patients with knee osteoarthritis · checked 2026-10-06 · we read the abstract
12. ev-turinf-12 · Laboratory or animal data · case series · n = 10
    Turmeric causes potentially severe liver injury that is typically hepatocellular, with a latency of 1 to 4 months and strong linkage to HLA-B*35:01.
    
    Who: adjudicated turmeric-associated liver injury cases enrolled in the US DILIN, 2004 to 2022
    Effect: 5 hospitalized, 1 death from acute liver failure; HLA-B*35:01 allele frequency 0.450 vs 0.056-0.069 in population controls; latency 1 to 4 months
    Certainty: Case series without a denominator, so the absolute risk is unknown; 3 of 7 tested products also contained piperine.
    Am J Med, 2023: Liver Injury Associated with Turmeric, A Growing Problem: Ten Cases from the Drug-Induced Liver Injury Network · checked 2026-10-06 · we read the abstract
13. ev-turinf-13 · Laboratory or animal data · case series from pharmacovigilance with systematic review of case reports · n = 7
    In all cases, hepatotoxicity was associated with Curcuma longa formulations with high bioavailability and high dosage of curcumin/curcuminoids.
    
    Who: Tuscan cases of acute hepatitis reported to the Italian Phytovigilance system up to September 2019, plus 23 published cases
    Effect: 7 regional cases plus 23 published cases; positive dechallenge in most; most published cases also took other medicines
    Certainty: Spontaneous reports; causality by RUCAM and WHO-UMC, no denominator.
    Br J Clin Pharmacol, 2021 · checked 2026-10-06 · we read the abstract
14. ev-turinf-14 · Laboratory or animal data · case report · n = 1
    Turmeric was associated with significant iron deficiency anemia, consistent with the binding of available iron in the gut and the prevention of absorption.
    
    Who: a 66-year-old man taking turmeric extract for an osteoarthritis flare
    Effect: hemoglobin stayed below 12 on turmeric; returned to normal 2 weeks after stopping
    Certainty: Single case; mechanism (iron binding in the gut) shown in mice only.
    Cureus, 2019 · checked 2026-10-06 · we read the abstract
15. ev-turinf-15 · Position of an expert body · government health information page · n = —
    The initial evidence is positive; higher-quality evidence is needed to reach definitive conclusions, and more research is needed to understand the impact of bioavailability on curcumin’s effects.
    
    Who: general public
    Effect: —
    Certainty: Agency summary of meta-analyses, no own analysis.
    NIH National Center for Complementary and Integrative Health, Turmeric: Usefulness and Safety · checked 2026-10-06 · we read the section
16. ev-turinf-16 · Position of an expert body · government health information page · n = —
    Many curcumin products with increased bioavailability are on the market, and liver damage has been reported in some people who have consumed these bioavailable formulations.
    
    Who: adults using oral turmeric or curcumin supplements
    Effect: —
    Certainty: Agency safety statement.
    NIH National Center for Complementary and Integrative Health, Turmeric: Usefulness and Safety · checked 2026-10-06 · we read the section
17. ev-turinf-17 · Position of an expert body · government health information page · n = —
    The use of turmeric supplements during pregnancy may be unsafe.
    
    Who: pregnant women
    Effect: —
    Certainty: Agency safety statement; applies to supplements, not culinary amounts.
    NIH National Center for Complementary and Integrative Health, Turmeric: Usefulness and Safety · checked 2026-10-06 · we read the section