Turmeric and Liver Injury: What the Cases Show

Table of Contents

1. The Short Answer
2. The DILIN Series: Ten Cases, Collected Prospectively
3. The HLA-B*35:01 Finding — What It Is, and What It Is Not
4. The Italian Phytovigilance Cluster
5. Two Australian Cases
6. When the Injury Looks Autoimmune
7. The Counterweight: The USP Review
8. How Both Things Are True at Once
9. Why This Signal Appeared When It Did
10. What to Watch For
11. What Happens After You Stop
12. Practical Rules That Cut the Risk
13. Key Research Papers
14. Connections
15. Featured Videos

The Short Answer

Turmeric supplements can, uncommonly, injure the liver. The injury is idiosyncratic, which is the medical word for unpredictable: it does not track neatly with dose in the way that, say, paracetamol overdose does, it happens to a small minority of users, and there is no test that tells you in advance whether you are one of them.

The typical published picture looks like this. An adult, more often a woman, starts a turmeric or curcumin supplement for joint pain, inflammation or general health. Weeks to a few months later — the latency is usually measured in weeks, not days — they become tired, lose their appetite, feel unwell in a vague way. Blood tests show a hepatocellular pattern, meaning the enzymes released by damaged liver cells (ALT and AST) rise far more than the ones that signal a bile-flow problem. Then jaundice appears. When the supplement is stopped, most people recover over weeks to months. A minority are seriously ill; the literature includes severe cases, and the possibility of a life-threatening outcome is precisely why this topic gets an article rather than a footnote.

The single most consequential variable in that story is how long somebody keeps taking the product after feeling unwell. That is the part a reader controls.

Back to Table of Contents

The DILIN Series: Ten Cases, Collected Prospectively

The Drug-Induced Liver Injury Network is a US multicentre research network that exists to do one job properly: find people with suspected drug- or supplement-related liver injury, collect their information as it happens rather than reconstructing it afterwards, and have a panel of hepatologists adjudicate how likely the suspected agent really was to be the cause. That last step is what separates DILIN from a database of unverified reports. Somebody with expertise looked at each case and asked what else it could have been.

In 2023, Halegoua-DeMarzio, Navarro, Ahmad and colleagues published the turmeric cases in the American Journal of Medicine under the title "Liver Injury Associated with Turmeric — A Growing Problem: Ten Cases from the Drug-Induced Liver Injury Network." Ten adjudicated cases, gathered prospectively.

Several features of that report are worth holding onto:

• The word "growing" in the title is doing work. The network's point was not that ten is a large number in itself — it is not. It was that turmeric had gone from essentially absent in their earlier casework to a recurring entry, and that the timing lines up with turmeric's rise as one of the best-selling botanical supplements in the United States.
• The pattern was predominantly hepatocellular. Liver-cell enzymes rose steeply. This is the pattern that, when severe and accompanied by jaundice, carries the worst prognosis in drug-induced liver injury generally.
• Products were not uniform. Turmeric supplements are not a single substance. They differ in extract strength, in curcuminoid profile, in what else is in the capsule, and in whether an absorption enhancer is present — and the cases reflected that variety.
• The cases were serious enough to need the network's attention, up to and including severe outcomes. A series like this is assembled precisely because the events are not trivial.
• Causality was adjudicated, not assumed. Other causes — viral hepatitis, alcohol, prescription drugs, autoimmune disease, biliary obstruction — are worked through before a case is attributed to a supplement. This is the main reason the series carries more weight than a comparable number of isolated reports.

Ten cases, then, but ten well-characterised cases from a system designed to avoid fooling itself. That is the appropriate weight to give it: not a reason for panic, and not something to wave away.

Back to Table of Contents

The HLA-B*35:01 Finding — What It Is, and What It Is Not

The DILIN report also described a genetic association. Among the turmeric cases in that series, the HLA-B*35:01 variant turned up in a subset of patients who were genotyped — more often than you would expect by chance in the background population.

It is worth being precise about this, because it is the part of the story most likely to get distorted in retelling.

What HLA is. The human leukocyte antigen genes encode the molecules your immune system uses to display fragments of proteins and other chemicals to T cells — essentially the display cases in which the immune system inspects what is inside your cells. Different HLA variants hold different shapes. That is why particular HLA types have been linked, in other drugs, to particular idiosyncratic reactions: a variant that happens to present a drug-derived fragment in an immunogenic way can convert a harmless chemical into something the immune system attacks. HLA-B*35:01 has been reported in association with liver injury from more than one herbal product, which makes the turmeric observation biologically plausible rather than a statistical accident of one small study.

What it is not. It is not a validated clinical test, and nobody in that report suggested it was. Three limits matter:

1. The association was observed in a small subset of a small series. That is how genetic associations in idiosyncratic liver injury usually start, and it is also how many of them fail to replicate.
2. The variant is common; the injury is not. HLA-B*35:01 is carried by a meaningful slice of the general population, and the overwhelming majority of those people take turmeric supplements without incident. Carrying the variant cannot therefore mean "you will be injured." At most it shifts a very small probability.
3. There is no screening programme, and testing yourself would not tell you much. A test whose positive result still leaves you overwhelmingly likely to be fine, and whose negative result does not guarantee safety, is not a decision tool. Watching for symptoms is.

The right way to carry this finding is as a mechanistic clue: evidence that turmeric-associated liver injury is at least partly an immune-mediated, host-specific event, which fits the clinical picture of a delayed reaction in a minority of users. It is not a reason to seek genotyping, and any product or clinic offering to test you for it as a safety service is well ahead of the evidence.

Back to Table of Contents

The Italian Phytovigilance Cluster

One reasonable objection to a single national case series is that it might reflect something local — a particular product, a particular supplier, a particular contaminant in one market. The Italian data answer that objection.

Lombardi, Crescioli, Maggini and colleagues published an analysis in the British Journal of Clinical Pharmacology in 2021 covering acute liver injury following turmeric use in Tuscany, drawing on the Italian Phytovigilance database and pairing it with a systematic review of published case reports. Italy runs a dedicated surveillance system for suspected adverse reactions to herbal and botanical products, which is unusual and useful: it means somebody is collecting these reports in a structured way rather than leaving them scattered across the literature.

Two conclusions follow. First, the signal is not confined to the United States; it appears in an independent surveillance system, in a different regulatory environment, with different products on the shelves. Second, the systematic-review half of that paper put the individual published cases into one place, which is how a scatter of anecdotes becomes something with a describable shape — a typical latency, a typical biochemical pattern, a typical course.

Pharmacovigilance data come with their own well-known limits, and this page should state them rather than lean on the finding uncritically. Spontaneous reporting systems are voluntary, so they under-count; they are also sensitive to publicity, so a period of media attention can inflate reports without any change in the underlying rate. And they cannot give you a rate at all, because nobody knows the denominator — the number of people who took the product and were fine. What they are good at is detecting that something is worth investigating. That is what happened here.

Back to Table of Contents

Two Australian Cases

Luber, Rentsch, Lontos and colleagues reported two cases of turmeric-induced liver injury in Case Reports in Hepatology in 2019. Individual case reports sit lower on the evidence hierarchy than an adjudicated series, and they should be read that way — but they are not worthless, and they serve two purposes here.

The first is geographic. A third continent, a third health system, the same clinical picture. Independent replication of an unusual observation is exactly how a genuine signal distinguishes itself from local noise.

The second is descriptive. Case reports are written by the clinicians who actually managed the patients, and they preserve the details that aggregate analyses flatten out: what the product was, how long it had been taken, what the patient initially thought was wrong, what tests were done, how long recovery took. A reader trying to recognise this problem in themselves learns more from two well-written narratives than from a summary statistic.

The publication date is also part of the story. By 2019 this was being reported as a recognisable entity rather than a curiosity, which is consistent with DILIN's observation that the problem was growing alongside the products' popularity.

Back to Table of Contents

When the Injury Looks Autoimmune

Lukefahr, McEvoy, Alfafara and colleagues described, in BMJ Case Reports in 2018, drug-induced autoimmune hepatitis associated with turmeric dietary supplement use.

Autoimmune hepatitis is a condition in which the immune system attacks liver cells. A handful of drugs and supplements can provoke a picture that looks like it — the same antibodies, the same appearance under the microscope, the same response to steroids — but which is triggered by the agent rather than arising on its own. The distinction is not academic. Genuine autoimmune hepatitis often requires long-term immune suppression. A drug-induced mimic frequently settles once the trigger is removed, sometimes with a course of steroids to speed things along, and may not need indefinite treatment.

This case matters to a general reader for a practical reason: it is a demonstration that turmeric-associated liver injury can present in more than one way. If your only mental template is "supplements cause a sudden chemical poisoning of the liver," an insidious, antibody-positive, autoimmune-looking illness that began three months after starting a capsule will not look like a supplement problem — to you or, without a careful history, to the doctor seeing you. That is why the supplement history matters so much in liver clinics, and why it is worth volunteering rather than waiting to be asked. Many people do not think of a botanical capsule as a drug and do not mention it.

It also reinforces the immune-mediated mechanism suggested by the genetic association above. A host-specific immune reaction is the kind of thing that can present as hepatitis in one person and as an autoimmune syndrome in another.

Back to Table of Contents

The Counterweight: The USP Review

A safety page that cites only the alarming half of a literature is not being careful; it is being selective in a direction that feels responsible. So here is the other side, given the space it has earned.

In 2026, Akhtar, Barnes, Gardiner and colleagues published a comprehensive review in Pharmaceutical Biology on behalf of the US Pharmacopeia, titled "Rarely reported cases of hepatotoxicity associated with turmeric- and curcuminoid-containing dietary supplements." The US Pharmacopeia is the body that sets public quality standards for medicines and supplements in the United States; reviews like this feed into whether an ingredient carries a cautionary labelling requirement in its monograph, so they are done with some rigour.

The review's central argument is one of proportion. Turmeric and curcuminoid supplements are consumed by an enormous number of people — turmeric has been among the top-selling botanical ingredients in the US market for years, which puts the exposed population in the millions annually. Against that denominator, the total number of well-documented, properly adjudicated hepatotoxicity cases in the world literature is small. The review also does the unglamorous work of asking, case by case, how strong each attribution actually is: whether alternative causes were excluded, whether the product was analysed, whether the exposure was confirmed, and whether the reported product might have contained something other than what was on the label.

That last point deserves emphasis, because it cuts both ways. Multi-ingredient supplements are common, and a capsule sold as turmeric may contain several botanicals, an absorption enhancer, and excipients. Some published "turmeric" cases involve products where another ingredient is at least as plausible a culprit. Adulteration and mislabelling are documented problems in the supplement industry generally. A review that insists on this distinction is not making excuses; it is asking the question that has to be asked before an ingredient is blamed.

Back to Table of Contents

How Both Things Are True at Once

Readers are often handed these two findings as a contradiction — a network says turmeric hurts livers, a pharmacopeia says it rarely does — and told to pick a side. They are not a contradiction. They are answers to different questions.

DILIN answers: when someone presents with unexplained liver injury and a supplement history, can turmeric be the cause? Answer: yes, demonstrably, in adjudicated cases with a consistent pattern.

USP answers: how often does that happen, per person taking these products? Answer: rarely, on the available evidence, given how many people are exposed.

An analogy that holds up reasonably well: peanut allergy. Peanuts cause severe, sometimes fatal reactions. Peanuts are also eaten safely billions of times a day. "Peanuts can cause anaphylaxis" and "peanuts are rarely dangerous" are both true, because the risk is concentrated in a susceptible minority rather than spread evenly across everybody. Idiosyncratic liver injury behaves the same way. The average risk across all users is low, which is what a population statistic measures. The risk to a susceptible individual, who cannot currently be identified in advance, is not low at all.

That structure tells you what to do with the information. Because the risk is not evenly spread, reducing your average dose is only part of the answer — the more useful lever is early detection. You cannot know in advance whether you are susceptible. You can know within a week or two of symptoms starting, if you take those symptoms seriously and get a liver panel. A rare event that is caught early and reverses is a different thing from a rare event that runs for two months.

The broader review by Halegoua-DeMarzio and Navarro in Liver International, on the challenges in identifying and preventing herbal-induced liver injury, is essentially about this problem: the products are unregulated in ways drugs are not, users often do not report them, the denominator is unknown, and product contents are variable. That is the honest state of the field.

Back to Table of Contents

Why This Signal Appeared When It Did

Turmeric has been eaten for millennia and turmeric extracts have been sold for decades, so why do the case reports cluster in roughly the last ten years? Three changes, all of them about exposure rather than about the plant.

1. Popularity. Turmeric went from a spice-rack item to one of the best-selling botanical supplements in Western markets. More users means more of any rare event, even at a constant rate per user. Some of the apparent increase in cases is simply this.

2. Dose. The typical supplement dose climbed. A capsule of standardised 95% curcuminoid extract concentrates the curcuminoids roughly twenty- to thirtyfold relative to the powder before anything else is done to it, and daily regimens of one to two grams of extract are ordinary on the shelf today.

3. Formulation — and this is the underrated one. The last two decades of curcumin product development were an engineering campaign against poor absorption, and it succeeded. Piperine from black pepper blocks the conjugation enzymes that would otherwise clear absorbed curcumin. Phospholipid complexes, submicron colloidal suspensions, solid lipid particles and micellar preparations each solve a different part of the delivery problem. Published comparisons of these formulations against unformulated extract report multiples in the tens — and for some preparations considerably more. Our bioavailability and forms page covers the mechanisms in detail.

Put those together and the consequence is unavoidable: two bottles reading "500 mg curcumin" can deliver very different amounts of curcumin to your liver. The milligram figure on the front of the label describes what went into the capsule, not what reaches your bloodstream, and the enhanced products are designed to widen that gap. Anybody comparing their own intake to a dose in a published trial, or to what their neighbour takes, is comparing numbers that are not on the same scale.

This is also why the spice keeps its clean record while the supplement generates case reports. It is not that the molecule is different. It is that the amount arriving in the liver is different by orders of magnitude.

Back to Table of Contents

What to Watch For

These are the signs that should take you from "I feel a bit off" to "stop the supplement and get a liver panel this week." Individually several of them are unremarkable; appearing together, or persisting, they are worth acting on.

• Jaundice — yellowing of the whites of the eyes first, then the skin. Check in daylight, not under indoor lighting, and ask someone else to look; people notice it in others before themselves.
• Dark urine, the colour of strong tea or cola, not explained by dehydration.
• Pale, clay-coloured or chalky stools.
• Pain, aching or tenderness under the right ribs at the front — the liver sits there, and a swollen one stretches its capsule.
• Itching with no rash, often worse at night, sometimes preceding jaundice.
• Unusual fatigue — not ordinary tiredness but a flattening exhaustion that has built up over days or weeks.
• Loss of appetite, nausea, or a sudden aversion to rich food.
• Unexplained low-grade fever or a rash alongside any of the above, which can point to an immune-mediated reaction.

Timing is a clue. The published cases usually begin weeks to a few months after starting the supplement, not on day one. If you have been taking something for a month and feel steadily worse, the fact that you tolerated the first three weeks is not reassurance — that latency is characteristic.

Take the bottle to the appointment, or photograph the full label including the ingredient list, the extract standardisation, any absorption enhancer and the batch number. "I take turmeric" is a much weaker piece of clinical information than the actual product, and if a batch problem is ever investigated, the batch number is the only thing that connects your case to anyone else's.

Back to Table of Contents

What Happens After You Stop

The reassuring part of this literature is the course after withdrawal. In the published cases, the usual outcome of stopping the supplement is recovery. Liver enzymes fall, jaundice clears, energy returns. The pace is slower than most people expect — the liver regenerates well but not quickly, and a return to normal blood tests commonly takes weeks to months rather than days. Feeling better before the numbers normalise is typical and is not a reason to restart anything.

What a clinician will generally do is straightforward. Take a full history including every supplement, tea, powder and protein blend. Run a liver panel and repeat it to establish the direction of travel. Exclude the other causes: viral hepatitis serology, alcohol, prescription and over-the-counter drugs (paracetamol above all), autoimmune markers, and an ultrasound to check the bile ducts and the liver's appearance. If the picture fits and other causes are excluded, the supplement is stopped and the tests are followed until they settle. A minority of people — particularly those with an autoimmune-looking pattern — are treated with steroids. A small minority are ill enough to need specialist hepatology care.

Two things not to do. Do not rechallenge. Restarting the product to see whether it really was the cause is the one manoeuvre that reliably makes idiosyncratic liver injury worse, and a second reaction is often faster and more severe than the first. Do not substitute another liver supplement. The instinct to take something for the liver is understandable and, in this situation, precisely backwards: the liver needs fewer things to process, not more, and herbal liver injury is a field with several plausible culprits rather than one.

Also worth knowing: if your enzymes were abnormal and turmeric was the cause, that is useful information about you, not just about the product. Idiosyncratic reactions are host-specific. It is reasonable to treat concentrated turmeric extract as something you personally do not take, while continuing to eat the spice.

Back to Table of Contents

Practical Rules That Cut the Risk

None of this requires a complicated protocol. Six rules cover almost all of it.

1. Cook with it freely. The culinary use of turmeric is not implicated in any of this, and a page that leaves you nervous about a spice jar has failed at its job.
2. Know which product you are taking. Plain powder, standardised extract, or an absorption-enhanced formulation? The third category is a genuinely higher exposure at the same label dose, and it is the category most represented in the case literature.
3. Prefer single-ingredient products. If a capsule contains six botanicals and something goes wrong, nobody — including you — will know which one did it. Single-ingredient products also make the label meaningful.
4. Take it in defined courses with a reason and an end date. "Eight weeks for this joint flare, then reassess" is a plan. "Indefinitely, because it is good for you" is how people end up four months into an exposure they have stopped thinking about.
5. If you are going to take it long-term, check liver enzymes once. A baseline panel before starting and a repeat after a couple of months is a cheap, ordinary test that converts an invisible risk into a visible one. This is especially worth doing if you have any existing liver condition, drink regularly, or take other medicines metabolised by the liver.
6. Stop at the first unexplained symptom, and say what you were taking. Stopping early is most of what separates a good outcome from a bad one in this literature, and the supplement history is the piece of information clinicians most often have to dig for.

For the groups with a specific reason to skip the supplement altogether — existing liver disease, transplant medicine, gallbladder obstruction and several others — see who should avoid turmeric supplements.

Back to Table of Contents

Key Research Papers

1. Halegoua-DeMarzio D, Navarro V, Ahmad J, et al. (2023). Liver Injury Associated with Turmeric—A Growing Problem: Ten Cases from the Drug-Induced Liver Injury Network [DILIN]. Am J Med. — PubMed PMID: 36252717 (the source of the ten-case series and of the HLA-B*35:01 observation reported in this article)
2. Akhtar N, Barnes J, Gardiner P, et al. (2026). Rarely reported cases of hepatotoxicity associated with turmeric- and curcuminoid-containing dietary supplements: a comprehensive review by USP. Pharm Biol. — PubMed PMID: 42364655
3. Lombardi N, Crescioli G, Maggini V, et al. (2021). Acute liver injury following turmeric use in Tuscany: An analysis of the Italian Phytovigilance database and systematic review of case reports. Br J Clin Pharmacol. — PubMed PMID: 32656820
4. Luber RP, Rentsch C, Lontos S, et al. (2019). Turmeric Induced Liver Injury: A Report of Two Cases. Case Reports Hepatol. — PubMed PMID: 31214366
5. Lukefahr AL, McEvoy S, Alfafara C, et al. (2018). Drug-induced autoimmune hepatitis associated with turmeric dietary supplement use. BMJ Case Rep. — PubMed PMID: 30206065
6. Halegoua-DeMarzio D, Navarro V. (2025). Challenges in herbal-induced liver injury identification and prevention. Liver Int. — PubMed PMID: 39136211
7. Shoba G, Joy D, Joseph T, et al. (1998). Influence of piperine on the pharmacokinetics of curcumin in animals and human volunteers. Planta Med. — PubMed PMID: 9619120 (why an enhanced formulation is a different exposure from the same milligrams of powder)
8. Kanai M, Otsuka Y, Otsuka K, et al. (2013). A phase I study investigating the safety and pharmacokinetics of highly bioavailable curcumin (Theracurmin) in cancer patients. Cancer Chemother Pharmacol. — PubMed PMID: 23543271
9. Cheng AL, Hsu CH, Lin JK, et al. (2001). Phase I clinical trial of curcumin, a chemopreventive agent, in patients with high-risk or pre-malignant lesions. Anticancer Res. — PubMed PMID: 11712783
10. Sharma RA, McLelland HR, Hill KA, et al. (2001). Pharmacodynamic and pharmacokinetic study of oral Curcuma extract in patients with colorectal cancer. Clin Cancer Res. — PubMed PMID: 11448902
11. Burgos-Morón E, Calderón-Montaño JM, Salvador J, et al. (2010). The dark side of curcumin. Int J Cancer. — PubMed PMID: 19830693

PubMed Topic Searches

1. PubMed: Turmeric induced liver injury
2. PubMed: Herbal and dietary supplement hepatotoxicity in DILIN
3. PubMed: HLA-B*35:01 and drug-induced liver injury
4. PubMed: Idiosyncratic drug-induced liver injury and recovery
5. PubMed: Curcumin supplement adverse events

Back to Table of Contents

Connections

• Turmeric Safety and Drug Interactions
• Who Should Avoid Turmeric Supplements
• Turmeric Drug Interactions
• Lead in Turmeric
• Turmeric for Liver Health
• Bioavailability and Forms
• Turmeric Benefits
• Turmeric and Digestive Health
• Turmeric
• Javanese Turmeric
• Culinary and Spice Herbs
• Herbs

Back to Table of Contents