Skip to content

Printed from https://bottledhope.co/supplements/turmeric/. Last reviewed 2026-10-12. Check the site for the latest version.

Turmeric and curcumin

A yellow spice from a plant in the ginger family, colored by curcumin. It is promoted as a supplement for osteoarthritis, depression and high cholesterol.

Explained simply · Last reviewed

What it is

Turmeric is a yellow spice from a plant in the ginger family. NCCIH, the US agency that studies complementary medicine, says the plant is grown commercially in Southeast Asia and India, and that curcumin, a major component of turmeric, gives the spice its yellow color. It says turmeric has been used in Chinese, Indian (Ayurvedic), Islamic and Thai traditional medicine, and that today it is promoted as a supplement for conditions including osteoarthritis, depression, allergies, itching and high cholesterol.

More about Turmeric and curcumin

NCCIH, the US agency that studies complementary medicine, says the plant is grown commercially in Southeast Asia and India, and that curcumin, a major component of turmeric, gives the spice its yellow color. It says turmeric has been used in Chinese, Indian (Ayurvedic), Islamic and Thai traditional medicine, and that today it is promoted as a supplement for conditions including osteoarthritis, depression, allergies, itching and high cholesterol.

NCCIH says the plant's rhizome, the underground stem, has been used for culinary and cosmetic purposes, among others. It says the health activities of turmeric are commonly attributed to curcuminoids, a group of compounds that includes curcumin.

LiverTox, a US government liver-injury database, states that curcumin makes up between 1% and 6% of whole turmeric extracts by dry weight, and that commercial turmeric and curcumin products vary widely in how much curcuminoid they contain. It notes that curcumin is poorly absorbed by mouth, and that forms using piperine from black pepper or lipid nanoparticles were developed to increase its absorption.

LiverTox lists tablets, liquids, root extracts and teas among the forms turmeric is taken in. Whole-root powder, extracts concentrated in curcuminoids, and enhanced-absorption curcumin are different products, and the evidence below keeps them apart where the sources do.

Also known as: curcumin, curcuminoids, Curcuma longa, turmeric root, Indian saffron

At a glance

Safety: known risks, 6 groups who should avoid it or take extra care and 3 documented interactions. See Safety ↓

Claims made about Turmeric and curcumin, and how strong the evidence is for each:

Each rating follows our published rules and the sources cited for it. We don’t add our own medical judgment. How we rate

Safety

Read this before the ratings below.

Risks

  • LiverTox gives turmeric a likelihood score of A, which it defines as “recently established as a well-documented cause of clinically apparent liver injury”. It states that turmeric “appears to have become the most common cause of clinically apparent, herbal-related liver injury in the United States”, and that the incidence is not known but is “probably very rare, in the range of 1:10,000 to 1:100,000 persons exposed”.
  • LiverTox states that most cases are attributed to highly bioavailable forms of curcumin, made with piperine (black pepper) or lipid nanoparticles to increase absorption, but that cases have also been described with unadulterated ground turmeric powder and with turmeric herbal teas.
  • For the liver injury attributed to highly bioavailable forms of curcumin, LiverTox reports that it typically began 1 to 4 months after starting, ranging from a few weeks to as long as eight months, with fatigue, nausea and poor appetite followed by dark urine and jaundice. It states for those cases that “fatal instances might occur at a rate of 10% of jaundiced cases, particularly if the product is not discontinued promptly”. Of turmeric liver injury overall, it states that most cases resolve within 1 to 3 months of stopping, but that in some the injury is severe and unremitting, leading to acute liver failure and either death or the need for liver transplantation.
  • LiverTox reports a close association between turmeric liver injury and the immune-system gene variant HLA-B*35:01, found in over 70% of cases compared with 10% to 15% of controls, and describes the injury as idiosyncratic, perhaps immunologically mediated.
  • Health Canada’s 2025 safety review of turmeric and curcuminoid natural health products taken by mouth found “a possible link” between their use and the risk of liver injury, which, it states, “in rare cases, could lead to serious outcomes”. It reviewed 12 Canadian cases, all with insufficient clinical information or other possible causes, though a link to the product “could not be ruled out”, and over 60 international case reports. Three deaths were reported, 2 of them attributed to liver injury linked with oral turmeric or curcuminoid products.
  • Health Canada describes the risk as appearing to be idiosyncratic (“not dependent on the dose or duration of use, and unpredictable in onset with risk factors not fully known”), and states that in most cases the injury is reversible once the product is stopped. Its review did not cover turmeric eaten as food, which it states “has not been identified as a safety concern” when used as a culinary spice in typical dietary amounts.
  • Health Canada expects license holders to update the labels of all licensed turmeric and curcuminoid products with warnings giving the signs and symptoms of liver injury, “including yellowing of the eyes or skin, dark urine, nausea, vomiting and stomach pain”, advising consumers to stop using the product and consult a healthcare professional if symptoms occur, and to consult one before use if they have a liver disorder or are taking medications.
  • NCCIH states that liver damage has been reported in some people who took curcumin formulations with increased bioavailability. It also states that conventionally formulated oral turmeric or curcumin, not modified to enhance bioavailability, “is likely safe in the recommended amounts for up to 2 or 3 months”.
  • France’s food safety agency reported in June 2022 that Italy had recorded around 20 cases of hepatitis involving turmeric food supplements, and that its own monitoring scheme had received over 100 reports of adverse effects, including 15 of hepatitis, potentially related to supplements containing turmeric or curcumin.
  • NCCIH reports that oral turmeric can cause nausea and vomiting, acid reflux, stomach upset, diarrhea or constipation, and that curcumin applied to the skin can cause hives or itching. LiverTox describes side effects as uncommon and mild, but states they may include dermatitis and gastrointestinal upset.
  • A 2025 umbrella review of curcumin meta-analyses reports that eight of the reviews it included described adverse events, mainly gastrointestinal symptoms such as bloating, nausea, abdominal pain, diarrhea and constipation, as well as headache, dizziness, rash and hot flushes, and that they reported no serious adverse events. A 2025 review of knee osteoarthritis trials tabulated adverse events without pooling them; its authors describe them as appearing mild, and state that their reliability was limited because trials defined and reported them differently.

Who should avoid it or take extra care

  • Pregnancy: NCCIH states that the use of turmeric supplements during pregnancy may be unsafe.
  • Breastfeeding: NCCIH states that little is known about whether turmeric is safe while breastfeeding in amounts greater than those commonly found in food.
  • People with bile duct disease: France’s food safety agency advises against turmeric food supplements, stating that turmeric stimulates the secretion of bile.
  • Anyone who has had liver injury from turmeric: LiverTox states that re-exposure should be avoided.
  • People with a liver disorder: Health Canada expects the labels of licensed turmeric and curcuminoid products to advise consulting a healthcare professional before use.
  • Anyone taking the medicines listed below, or any regular medicine: France’s food safety agency advises seeking medical advice first, and NCCIH advises anyone who takes any type of medicine to talk with their health care provider before using any herbal product.

Documented interactions

  • Anticoagulants: France’s food safety agency states there is a risk of curcumin interacting with anticoagulants, which may make them less safe or less effective. It advises people taking them not to consume turmeric food supplements without seeking medical advice. Source [1]
  • Cancer drugs: The same agency states there is a risk of curcumin interacting with certain cancer drugs, which may make them less safe or less effective, and gives the same advice. Source [1]
  • Immunosuppressants: The same agency states there is a risk of curcumin interacting with immunosuppressants, which may make them less safe or less effective, and gives the same advice. Source [1]

Safety sources: Source [1]Source [2]Source [3]Source [4]Source [5]Source [6]

What the evidence says

Each claim is rated on its own evidence. Ratings follow the scale on our methodology page.

  • Eases knee osteoarthritis pain Evidence rating: Limited

    People with knee osteoarthritis take it hoping it will ease the pain in their knees.

    A 2025 review found less knee pain with enhanced-absorption curcumin than with placebo, at low to very low certainty; NCCIH calls the early evidence positive but not definitive.

    A 2025 network meta-analysis of 17 randomized trials in adults with knee osteoarthritis kept turmeric products apart by type. Enhanced-absorption curcumin products reduced pain on the WOMAC scale compared with placebo. The reviewers rated that result low certainty from the trials that compared them directly, and very low from the full network.

    For conventional curcuminoid extracts, the estimate against placebo was indirect, because every included trial compared them with painkillers. The review’s certainty table gives it no rating. In this review no preparation improved stiffness significantly compared with placebo. That is too little evidence to show an effect, rather than evidence of none. The reviewers conclude that “the certainty of evidence remains low”.

    A 2025 umbrella review re-graded two earlier meta-analyses that did not keep product types apart. It describes one as pooling turmeric extracts and curcumin together, and the other as pooling turmeric extracts. The first, in osteoarthritis, found significant improvements in WOMAC pain and in WOMAC stiffness, each from 4 placebo-controlled trials in 315 people. The umbrella rated both high certainty. For the other, it rated the pooled knee-pain result very low. It judged both of the earlier meta-analyses critically low in methodological quality.

    The two reviews therefore disagree on stiffness. The network meta-analysis found no significant improvement; the earlier meta-analysis found a significant one, rated high certainty by the umbrella. NCCIH describes the initial evidence for oral turmeric or curcumin as positive, and states that “higher-quality evidence is needed to reach definitive conclusions”. It calls the effect of curcumin ointment on knee pain unclear. This page rates the claim on the lowest certainty, very low.

    Sources: Source [1]Source [2]Source [3]

  • Lowers blood markers of inflammation Evidence rating: Limited

    People take it hoping it will calm inflammation in the body.

    In the largest analysis it graded, a 2025 umbrella review found lower C-reactive protein, TNF-α and interleukin-6 with curcumin or turmeric supplements in adults, but each at very low certainty.

    This claim is about laboratory markers, not symptoms or disease. A 2025 umbrella review graded earlier meta-analyses of curcumin or turmeric supplements. In the largest, in adults, supplements lowered C-reactive protein (45 trials), TNF-α (24 trials) and interleukin-6 (20 trials) compared with placebo. They did not change interleukin-1β significantly (6 trials).

    The umbrella rated all four results very low certainty. Heterogeneity was very high, with I-squared between 88% and 99% for the three that changed. That means the trials disagreed with each other substantially.

    Analyses in particular groups of people were mixed. In rheumatoid arthritis, C-reactive protein fell compared with placebo or standard care (very low certainty). In women with polycystic ovary syndrome it fell compared with placebo in 2 trials of 81 people (low certainty). In adults with delayed-onset muscle soreness, TNF-α fell compared with placebo (very low certainty), but interleukin-6 (low) and interleukin-8 (very low) did not change significantly. In people with fatty liver disease and people with chronic kidney disease, the differences from placebo were not statistically significant. These non-significant results are too little evidence to show an effect either way. The umbrella does not break these results down by type of product.

    Sources: Source [1]

  • Eases symptoms of depression Evidence rating: Limited

    People with low mood or depression take it hoping it will lift their mood.

    A 2026 review of 25 disagreeing trials found lower depression scores with curcumin than placebo, certainty unsettled; another found no significant difference for curcumin alone (very low certainty).

    A 2026 review pooled 25 placebo-controlled trials in adults that measured depression symptoms, and found lower scores with curcumin products than with placebo. Nine trials were in people diagnosed with major depression or schizophrenia. The reviewers class the other 16 as psychologically healthy, though the review’s own table of trials describes many as people with a physical illness such as type 2 diabetes, metabolic syndrome, migraine or heart disease. Several trials added curcumin to an antidepressant or antipsychotic that the placebo group also took.

    The pool mixed turmeric extracts, plain curcumin, and enhanced-absorption and nano-curcumin products, so its result is not specific to any one of them. One trial the reviewers group as turmeric, Setiawati 2017, used an extract of a related plant, Curcuma xanthorrhiza. The trials disagreed with each other substantially (I-squared 94.8%), and in the review’s forest plot 10 of the 25 found no significant difference on their own. The reviewers judged 5 trials at high risk of bias, and their own table describes 4 as open-label.

    Their GRADE table rates the result moderate certainty. Their abstract calls the evidence “moderate-to-low certainty”, and their discussion says certainty “was graded as low to very low”; this page does not settle that. They conclude curcumin “may provide a potential benefit”, but write that whether the changes are clinically meaningful for patients “remains uncertain”.

    A 2025 network meta-analysis of nutritional supplements for depressive disorder, 192 trials in all, rated its confidence in the curcumin comparisons for depressive symptoms low to very low. In its forest plot, against placebo, curcumin on its own did not differ significantly, which is too little evidence to show an effect. Curcumin together with an antidepressant did differ significantly from placebo. That comparison is with placebo, not with the antidepressant alone, so its result includes the antidepressant’s own effect, which the same plot shows as significant. The review’s abstract describes these figures as comparisons with antidepressants alone, while the plot presents them against placebo. The two do not agree, and this page reports the plot.

    A 2025 umbrella review graded an earlier meta-analysis in older adults. There, a pooled result from 2 placebo-controlled trials in 136 people was not significant (low certainty), too little evidence to show an effect. NCCIH states that “there isn’t enough evidence to definitively conclude” whether oral turmeric or curcumin benefits health conditions beyond those it lists, which do not include depression. The 2026 review found a statistically significant reduction in scores; the 2025 network meta-analysis found no significant difference for curcumin alone. This page rates the claim on the lowest certainty, very low.

    Sources: Source [1]Source [2]Source [3]Source [4]

  • Eases symptoms of anxiety Evidence rating: Limited

    People who feel anxious or on edge take it hoping it will help them feel calmer.

    A 2026 review of 15 trials, none in an anxiety disorder, found lower anxiety scores with curcumin than placebo, but trials disagreed widely and the reviewers call certainty low, or low to very low.

    A 2026 review pooled 15 placebo-controlled trials in adults that measured anxiety symptoms, and found lower scores with curcumin products than with placebo. In the review’s table of trials, none was in people diagnosed with an anxiety disorder. Three were in people with major depression. The reviewers class the other 12 as psychologically healthy, including people with type 2 diabetes, metabolic syndrome or migraine. Two trials added curcumin to antidepressant medication that the placebo group also took. The pool mixed turmeric extracts, plain curcumin, and enhanced-absorption and nano-curcumin products, so its result is not specific to any one of them.

    The trials disagreed with each other substantially (I-squared 95.7%), and in the review’s forest plot 7 of the 15 found no significant difference on their own. The reviewers judged 3 of the 15 at high risk of bias, and their own table describes 2 as open-label. A statistical test suggested possible publication bias. A follow-up analysis “did not identify any missing (imputed) studies”, and the reviewers report the result “materially unchanged”. Their GRADE table still downgrades anxiety for possible publication bias.

    That table rates the anxiety result low certainty, also downgraded for inconsistency. Their discussion says the “overall certainty of evidence was graded as low to very low, primarily due to inconsistency and potential publication bias, particularly in outcomes related to anxiety”. This page does not settle the two. The reviewers conclude curcumin “may provide a potential benefit”, but write that whether the changes are clinically meaningful for patients “remains uncertain”.

    NCCIH states that “there isn’t enough evidence to definitively conclude” whether oral turmeric or curcumin benefits health conditions beyond those it lists, which do not include anxiety. This page rates the claim on the lowest certainty the review states, very low.

    Sources: Source [1]Source [2]

  • Improves cholesterol and triglycerides Evidence rating: Limited

    People take it hoping it will improve their cholesterol and triglycerides.

    A 2025 umbrella review found lower total and LDL cholesterol and triglycerides and higher HDL in adults, at very low certainty; NCCIH says too little is known to conclude definitively that it helps.

    A 2025 umbrella review graded earlier meta-analyses of curcumin or turmeric supplements. The largest pooled 55 to 59 placebo-controlled trials in adults for each measure. It found lower total cholesterol, LDL cholesterol and triglycerides and higher HDL cholesterol. The umbrella rated all four results very low certainty. I-squared was 95% to 97%, meaning the trials disagreed with each other substantially.

    In the umbrella review, analyses in women with polycystic ovary syndrome were mixed, all rated very low certainty. Several of their results were not statistically significant, which is too little evidence to show an effect either way. This page does not count reviews published in journals that have been delisted from Web of Science or have closed.

    NCCIH lists high cholesterol among the uses that many studies have examined. It states that “we don’t know enough to definitively conclude if turmeric or curcumin is beneficial for any health purposes”. This page rates the claim on the lowest certainty, very low.

    Sources: Source [1]Source [2]

  • Lowers blood sugar Evidence rating: Limited

    People take it hoping it will bring their blood sugar down.

    A 2025 umbrella review found lower fasting blood sugar and insulin resistance in fatty liver disease and polycystic ovary syndrome, at moderate to very low certainty, but no significant HbA1c change.

    A 2025 umbrella review graded earlier meta-analyses of curcumin, each in a different patient group. This page does not count reviews published in journals that have been delisted from Web of Science or have closed, which leaves two.

    In the umbrella review’s fatty liver disease analysis, compared with placebo, fasting blood sugar fell (low certainty) and insulin resistance fell (very low). HbA1c, a longer-term measure, did not change significantly (very low). In polycystic ovary syndrome, compared with placebo or medication, fasting blood sugar and insulin resistance fell (both moderate). HbA1c, from 2 trials, did not change significantly (low). The non-significant results are too little evidence to show an effect either way.

    The umbrella judged the polycystic ovary syndrome meta-analysis critically low in methodological quality, and the fatty liver one very low. It describes the products in both only as curcumin, and does not break these results down by type of product. This page rates the claim on the lowest certainty, very low.

    Sources: Source [1]

How and when to take it

What the cited sources say about taking it. Ask your doctor or pharmacist what applies to you.

  • Stop and contact your doctor if you notice signs of liver trouble

    NCCIH advises anyone who notices liver damage symptoms “such as fatigue, nausea, poor appetite, dark urine, or jaundice” to stop turmeric or curcumin products immediately and talk with their health care provider. LiverTox states that a severe outcome is more likely if turmeric is continued after symptoms and signs of liver injury appear, and that re-exposure should be avoided because recurrence and more severe injury can occur. Source [1]Source [2]

  • Ask your doctor first if you take anticoagulants, cancer drugs or immunosuppressants

    France’s food safety agency states there is a risk of curcumin interacting with these medicines, which may make them less safe or less effective, and advises people taking them not to consume turmeric food supplements without seeking medical advice. Source [1]

  • The label may not tell you which kind of curcumin is inside

    France’s food safety agency states that supplement labels “seldom specify” whether they contain classic or enhanced-absorption formulations. NCCIH notes that oral curcumin products vary in how much curcumin they actually contain, and often contain substances from other plants, such as piperine from black pepper. Source [1]Source [2]

Which form the evidence used

Supplements are sold in several forms. This is what the sources on this page say about them — which form the research used, and any comparison those sources themselves made. Where they made none, this page does not rank the forms.

  • Enhanced-absorption curcumin (for example with piperine from black pepper, or in lipid, nanoparticle or other delivery forms), What the sources supportDetails

    A 2025 review of knee osteoarthritis trials groups these as bioavailability-enhanced curcuminoid preparations, enhanced “via additives (such as piperine), using matrices or particle size reduction techniques”. Five of the 7 trials in its main pain analysis used them. The reviewers call the comparison with conventional preparations “still inconclusive”, noting that some enhanced-product trials gave lower doses, so this page cannot say either works better. NCCIH states that “more research is needed to understand the impact of bioavailability on curcumin’s effects”, and that “highly bioavailable formulations of curcumin” “may harm your liver”. LiverTox attributes most turmeric liver-injury cases to highly bioavailable forms of curcumin. France’s food safety agency says such formulations “can pose a risk of adverse effects”. Source [1]Source [2]Source [3]Source [4]

  • Conventional curcuminoid or turmeric extracts, What the sources supportDetails

    The same review groups as conventional curcuminoid preparations the extracts made with a hydroalcoholic solution that concentrates curcuminoids. In every included trial of these extracts, the comparison was with painkillers (NSAIDs or paracetamol), not placebo, so the review’s estimate against placebo for them is indirect. France’s food safety agency lists turmeric extracts, enriched with curcumin or not, among the “classic forms”. Source [1]Source [2]

  • Turmeric polysaccharide extracts, What the sources supportDetails

    A third group in that review is polysaccharide-rich extracts, described as containing a “low amount of curcuminoid content”. Two placebo-controlled trials measured pain with them on a visual scale, and the reviewers rated that result low certainty. The review reports this group had the highest percentage of adverse events among the preparations it tabulated. That table, not pooled by the reviewers, covers 12 of the 17 trials. It records 20 adverse events among 66 people given these extracts (30.3%), against 41 among 236 given placebo (17.4%) and 24 among 260 given enhanced-absorption curcumin (9.2%). It sums counts across event types, so one person can count more than once. No source cited here compared these extracts directly with curcumin products. Source [1]

  • Ground turmeric root (the spice) and turmeric tea, Details

    None of the preparation groups in the 2025 knee osteoarthritis review is plain spice powder. France’s food safety agency lists turmeric rhizome powder among the classic forms. LiverTox states that, while most liver-injury cases are attributed to highly bioavailable forms of curcumin, “cases also have been described with use of unadulterated ground turmeric powder and even with turmeric herbal teas”. Source [1]Source [2]Source [3]

Also documented elsewhere on this site

Other supplements we cover that document an interaction with the same thing. This is a pointer to what those pages say, not a statement about taking them together — we have no source reviewing these in combination.

Sources

  1. Effect of turmeric products on knee osteoarthritis: a systematic review and network meta-analysis
    BMC Complementary Medicine and Therapies, 2025 · Meta-analysis · Primary source · Certainty: very low
  2. Curcumin and multiple health outcomes: critical umbrella review of intervention meta-analyses
    Frontiers in Pharmacology, 2025 · Systematic review · Primary source · Certainty: very low
  3. Turmeric
    National Center for Complementary and Integrative Health, NIH, 2025 · Health authority summary · Primary source
  4. Effects of curcumin on depression, anxiety, and stress symptoms in adults: a GRADE-assessed systematic review and dose–response meta-analysis
    Food & Nutrition Research, 2026 · Meta-analysis · Primary source · Certainty: very low
  5. Comparative efficacy and tolerability of nutraceuticals for depressive disorder: A systematic review and network meta-analysis
    Psychological Medicine, 2025 · Meta-analysis · Primary source · Certainty: very low
  6. Turmeric
    LiverTox: Clinical and Research Information on Drug-Induced Liver Injury, National Institute of Diabetes and Digestive and Kidney Diseases, 2025 · Health authority summary · Primary source
  7. Summary Safety Review - Turmeric and Curcuminoids for Oral Use - Assessing the Potential Risk of Hepatotoxicity
    Health Canada, 2025 · Health authority summary · Primary source
  8. Adverse effects associated with the consumption of food supplements containing turmeric
    Agence nationale de sécurité sanitaire de l'alimentation, de l'environnement et du travail (ANSES), 2022 · Health authority summary · Primary source

Page history

Show all 6 changes to this page
  • — Depression now reports a 2026 review of 25 trials that found lower depression scores with curcumin than placebo, with its certainty and heterogeneity caveats; rating unchanged, limited, because a cited 2025 review rates curcumin alone very low certainty. New claim: anxiety symptoms, rated limited; that 2026 review rates its certainty low in its GRADE table and low to very low in its discussion, and this page records very low.
  • — Rewritten to the content style guideline; ratings and sources unchanged.
  • — Source audit. The blood sugar and blood lipids summaries no longer report the umbrella review's grading of a 2022 type 2 diabetes meta-analysis, because its journal has since been delisted from Web of Science and has closed. No rating changed.
  • — Published. All five claims, the safety section (including Health Canada's 2025 safety review) and every source passed the site's source and wording reviews over two revision rounds. The owner's decisions on uncited sources are logged in docs/08-open-decisions.md.
  • — Safety section adds Health Canada’s 2025 summary safety review, which found a possible link between oral turmeric or curcuminoid products and liver injury. No rating changed.
  • — First drafted. Knee osteoarthritis pain, inflammation markers, depression, blood lipids and blood sugar all rated limited, each on very low GRADE certainty reported by the cited reviews. Safety section leads on the LiverTox likelihood score A for clinically apparent liver injury.