Tinospora and Liver Injury: The Autoimmune Hepatitis Reports

This is the first article in the Tinospora Benefits set rather than the last, and that ordering is deliberate. On most herbs the safety section is a coda — a list of theoretical cautions appended after the benefits have been described. Here it is the other way round. The strongest, most recent and most clinically documented human literature on Tinospora cordifolia is not about immunity, blood sugar or arthritis. It is a cluster of published hospital case series from India describing liver injury with autoimmune features in people who had been taking guduchi, several of them appearing during and after the extraordinary surge in giloy consumption that accompanied the COVID-19 pandemic.

That is an uncomfortable thing to put at the top of a page filed under “Benefits”. It is also the honest thing to do. When the best-documented human effect of a herb is a harm rather than a benefit, ranking the page by evidence quality means the harm goes first. A reader who came looking for reasons to take giloy is entitled to see the strongest human data before the weakest.

What follows tries to do three things at once and keep them separate: describe what the reports actually said, describe the published objections to them fairly and without deciding the argument, and explain why the disagreement does not resolve into reassurance. The short version is that a real safety signal exists, its details are genuinely contested, and the contest does not make the signal go away — because the population exposed was overwhelmingly made up of healthy people taking a preventive tonic, and a healthy person has nothing to gain that would offset even an uncommon serious harm.

Table of Contents

1. Why This Article Comes First
2. What the Case Series Described
3. The Clinical Picture: Hepatitis With Autoimmune Features
4. Drug-Induced Autoimmune-Like Hepatitis, or Unmasked Autoimmune Hepatitis?
5. Why Causality Assessment in Herbal Liver Injury Is Genuinely Hard
6. The Published Disagreement, in Both Directions
7. What the Objections Do and Do Not Establish
8. The Asymmetry: Healthy People Taking It for Prevention
9. Numbers This Page Will Not Give You
10. Warning Signs, and What They Should Trigger
11. Who Should Not Take Guduchi At All
12. A Reasonable Position to Hold
13. Key Research Papers
14. Connections

Why This Article Comes First

Herbal safety writing has a structural bias that is worth naming out loud. A herb page is usually organised around what the herb is for, because that is what the reader searched for, and the cautions are then bolted on at the end where they read as legal boilerplate. That structure quietly encodes a claim: that the benefits are the substance and the risks are the fine print.

For guduchi that structure would be simply wrong on the evidence. Rank the human literature by how well it was done and how directly it bears on a person swallowing a product, and the order comes out like this:

1. Case series of liver injury in identified patients, from hospital hepatology units, published in peer-reviewed hepatology journals, replicated across multiple centres. This is real clinical human data on real named exposures, and it is the strongest tier of evidence on this plant.
2. A small number of controlled human trials of guduchi or guduchi-containing regimens for specific indications, mostly small, several open-label, most testing multi-herb formulas. Useful, thin, and discussed on the other research page.
3. A large preclinical immunomodulatory literature in cells and rodents. Genuinely substantial in volume, and genuinely not evidence that a person who takes giloy gets fewer infections. See immune claims and the rasayana tradition.
4. Centuries of documented traditional use as a rasayana. Historically important, and not a substitute for either of the first two tiers.

The best-evidenced human effect sits at the top of that list and it is an adverse one. Putting it anywhere other than first would misrepresent the state of knowledge.

What the Case Series Described

Beginning in 2021, hepatology units in India published descriptions of patients who presented with liver injury and whose common recent exposure was a herbal immune-booster product, most often Tinospora cordifolia — giloy — taken as juice, tablets, powder or a home-boiled decoction. The first widely discussed report was a case series in the Journal of Clinical and Experimental Hepatology. A larger multicentre study drawing cases from centres across India followed in Hepatology Communications in 2022.

The general shape of the reports, as described in that literature, was consistent:

• Patients had been taking giloy for weeks to months, usually daily, usually on their own initiative rather than on a practitioner’s prescription, and almost always for prevention rather than for a diagnosed illness.
• They presented with symptoms of hepatitis — fatigue, poor appetite, nausea, discomfort under the right ribs, dark urine, jaundice, itching — and markedly raised transaminases on blood testing, indicating hepatocellular injury rather than a purely obstructive picture.
• A subset had features that pointed towards an autoimmune process: raised immunoglobulin G, positive autoantibodies of the kind used in diagnosing autoimmune hepatitis, and, where biopsy was performed, histology described as interface hepatitis with a plasma-cell-rich infiltrate.
• Some patients were treated with corticosteroids and improved on them, which is the treatment pathway for autoimmune hepatitis and not the usual pathway for a simple toxic injury.
• The illness was not uniformly mild. The reports describe a spectrum, and the severe end of that spectrum is why hepatologists wrote them up.

Two features of the exposure pattern matter as much as the clinical findings. First, the material was almost never botanically authenticated — nobody had tested the powder in the jar. Second, in many cases giloy was not the only thing being taken; the pandemic period saw people consuming several herbal preparations at once, sometimes alongside conventional medicines. Both facts are central to the argument described below.

The Clinical Picture: Hepatitis With Autoimmune Features

It is worth being precise about what “autoimmune features” means, because the phrase is doing real diagnostic work and is easy to read as vaguer than it is.

Autoimmune hepatitis is a defined disease in which the immune system attacks liver cells. Its diagnosis rests on a combination of findings rather than any single test: raised transaminases, raised immunoglobulin G, circulating autoantibodies — classically antinuclear antibody, anti-smooth-muscle antibody, or anti-liver-kidney-microsomal antibody — exclusion of viral hepatitis and other causes, and characteristic liver histology showing inflammation at the interface between the portal tracts and the liver lobule, with plasma cells prominent in the infiltrate. Response to immunosuppression, usually corticosteroids with or without azathioprine, is part of the clinical picture.

The giloy reports described patients who satisfied several of these criteria at once. That is a different and more specific observation than “their liver enzymes went up”. Raised transaminases alone are common and have dozens of causes. Raised transaminases plus raised IgG plus positive autoantibodies plus interface hepatitis on biopsy plus steroid responsiveness is a recognisable syndrome, and the reason the reports attracted attention is that it appeared in people whose only identified exposure was a herbal tonic.

The relevant scaffolding for a reader who wants to follow the blood tests is on the liver function tests page. The hepatitis panel is what excludes the viral causes, and excluding them is a required step before anyone can reasonably attribute an injury to a supplement.

Drug-Induced Autoimmune-Like Hepatitis, or Unmasked Autoimmune Hepatitis?

Within hepatology, the case series prompted a genuine diagnostic argument that has nothing to do with Ayurveda and everything to do with a hard, unresolved problem in liver medicine. There are at least three ways to interpret a patient who develops autoimmune-pattern hepatitis after taking a substance:

1. Drug-induced autoimmune-like hepatitis. The substance triggers an immune-mediated liver injury that mimics autoimmune hepatitis. It typically improves when the substance is stopped, often responds to steroids, and characteristically does not relapse when steroids are withdrawn — because the driver has been removed. Several conventional drugs are established causes of this pattern.
2. Unmasked de novo autoimmune hepatitis. The person was going to develop autoimmune hepatitis anyway, or had subclinical disease, and the substance either coincided with or precipitated its clinical appearance. Such patients need long-term immunosuppression and relapse when it is withdrawn.
3. Coincidence. A common exposure in a population and an uncommon disease will co-occur by chance, and during 2020–21 giloy was an extraordinarily common exposure in India.

Distinguishing these requires long follow-up after steroid withdrawal, which single case series rarely have. The published hepatology commentary on the giloy cases engaged with exactly this question, which is itself informative: clinicians do not conduct a careful nosological debate about cases they consider fabricated. The debate presupposes that the cases are real and asks what they are.

An honest page cannot settle this, and this one will not try. What can be said is that interpretations (1) and (2) both carry a practical implication for the reader — in the first, the herb caused the injury; in the second, the herb is at minimum suspected of precipitating a serious chronic disease in a susceptible person — and only interpretation (3) is exculpatory. The multicentre replication is the main reason coincidence became harder to hold as a complete explanation.

Why Causality Assessment in Herbal Liver Injury Is Genuinely Hard

Herb-induced liver injury is one of the hardest attributions in clinical medicine, and it is worth understanding why, because the difficulty is real and is not an argument invented to protect either side.

• There is no confirmatory test. Attribution is made by exclusion and by pattern, using structured causality instruments that score temporal relationship, the course after stopping the suspect agent, risk factors, concomitant drugs, exclusion of other causes, and whether the agent is already known to be hepatotoxic. The score produces a probability, not a verdict.
• Deliberate re-exposure is unethical. The single most convincing evidence — injury recurring on rechallenge — is something no clinician will induce on purpose. It is occasionally observed accidentally, and when it is, it is powerful.
• The exposure is usually unquantified. A drug case has a dose, a batch and a prescription record. A herbal case has “about a spoon of powder most mornings, from a shop, for a couple of months”, and often no remaining sample to test.
• Polypharmacy is the norm, not the exception. People taking one supplement usually take several, and the pandemic period was extreme in this respect.
• Products are not what their labels say with any reliability. Herbal products have documented problems with species substitution, undeclared ingredients, and contamination with heavy metals or pharmaceuticals. Any of those can be the true hepatotoxin while the label takes the blame — or the reverse.
• Reporting is passive and sparse. Adverse events from supplements are massively under-reported everywhere in the world, because neither the patient nor the clinician necessarily thinks of a “natural” product as a drug exposure worth recording.

That last point cuts in a direction people often miss. Under-reporting means a published case series is a floor, not a ceiling — it tells you cases exist, not how many. And it means that the absence of reports for some other herb is frequently absence of surveillance rather than evidence of safety.

The Published Disagreement, in Both Directions

The giloy reports were contested in print, and the objections deserve to be stated as strongly as their authors made them rather than summarised into a straw man.

The case against attributing the injuries to authenticated Tinospora cordifolia:

• No botanical authentication. The consumed material was not identified by any laboratory method. This is a real methodological gap and the objectors were right to press it.
• Tinospora crispa is a known hepatotoxin and is confused with T. cordifolia in trade. If some patients consumed the wrong species, their injury is not evidence about guduchi. This is the substantive core of the rebuttal and it has its own dedicated page here.
• Concomitant products and drugs. Patients taking several preparations, and in some instances conventional medicines with known hepatic effects, complicate attribution to any single agent.
• Centuries of traditional use without this pattern being described. Classical Ayurvedic use is as a hepatoprotectant, at practitioner-directed doses, in specific preparations, usually within compound formulas — not as an unsupervised daily tonic taken for months by the general population.
• Base rates. Against an exposure numbering in the many millions, a set of case reports does not by itself establish an elevated risk.

The case for taking the signal seriously:

• Replication across independent centres. A single-centre series can be an artefact of one unit’s referral pattern or one clinician’s hypothesis. A multicentre nationwide collection is much harder to explain that way, and this is the single most important development in the story.
• A coherent and specific syndrome. The cases were not a scatter of mild enzyme abnormalities. They shared a recognisable autoimmune-pattern presentation, which is what a real drug-induced immune-mediated injury looks like.
• Biological plausibility that runs the right way. Guduchi is promoted precisely as an immune stimulant, and the reported injury is immune-mediated. A herb marketed for its effect on immune activation producing an immune-mediated organ injury is not a mechanistically strange proposition — it is close to what one would predict. The same logic underlies the long-standing caution about immunostimulant herbs in people with autoimmune disease, discussed on the Echinacea autoimmune cautions page.
• Engagement from within the Ayurvedic research community. Papers in Ayurvedic and integrative-medicine journals took up the question directly, including analyses asking whether a herb classically regarded as hepatoprotective can nonetheless cause liver damage. That is not the behaviour of a field that considers the reports baseless.
• The species objection does not obviously cover the whole set. It is a hypothesis about some unknown fraction of cases. It has not been demonstrated for any specific patient in these series, because the material was gone by the time anyone asked.

Both columns above are legitimate. The literature has not converged, and anyone telling you it has — in either direction — is going beyond what has been published.

What the Objections Do and Do Not Establish

Here is where the argument turns, and it turns on a point that is easy to miss because the exculpatory version sounds so much more reassuring than it is.

Suppose the species-misidentification objection is entirely correct. Suppose every reported case of autoimmune-pattern hepatitis was caused by Tinospora crispa or some other adulterant, and authenticated Tinospora cordifolia is blameless. What follows for a person standing in a shop holding a jar of giloy powder?

Nothing reassuring. That scenario says the product category is contaminated with a known hepatotoxic look-alike, that the substitution is invisible in dried cut stem or milled powder, and that the consumer has no way to tell which plant they bought. An adulteration defence clears the plant. It does not clear the bottle — and since what people buy is bottles, the practical risk is unchanged or worse. A hazard you cannot identify by inspection is harder to avoid than one you can.

This is not a rhetorical trick; it is the standard structure of herbal-safety reasoning and it recurs across unrelated plants. The two possible worlds are:

1. Authenticated T. cordifolia can trigger immune-mediated liver injury in susceptible people. Then guduchi carries an idiosyncratic hepatic risk and should be treated like any drug with that property: a reason, a duration, and monitoring.
2. The commercial supply is unreliable and sometimes contains a hepatotoxic relative. Then guduchi products carry a hepatic risk and should be treated like any drug with that property: a reason, a duration, and monitoring — plus a supplier who identity-tests.

The two roads lead to nearly the same advice. Which is why the unresolved scientific question, while genuinely interesting, changes the practical answer far less than the debate’s heat suggests.

The Asymmetry: Healthy People Taking It for Prevention

The final piece is about who was exposed, and it is the reason this page exists in the form it does.

Risk is not evaluated in isolation. It is weighed against benefit, and the benefit side depends entirely on who is taking the substance and why. Consider two people:

• A patient with a serious diagnosed illness taking a drug with a known small risk of severe liver injury, under monitoring, because the alternative is worse. Here even a meaningful hepatic risk can be rational. Cancer chemotherapy and long-term immunosuppression are accepted on exactly this logic.
• A healthy adult taking a daily tonic to avoid getting ill, where no controlled human trial has shown that the tonic prevents any infection. Here the benefit side of the ledger is not small — it is unquantified, because the trial that would quantify it has not been run. Any non-trivial risk of serious harm is therefore difficult to justify.

During 2020–21, essentially the entire exposed population was the second kind of person: well, frightened, unsupervised, taking giloy daily for months at unmeasured doses in the hope of prevention. That is the least favourable possible risk–benefit setting. An adverse effect that would be an acceptable price for a treatment is not an acceptable price for a hope.

It is also the setting most likely to reveal an uncommon idiosyncratic reaction, simply because the denominator was enormous and the duration was long. Both things are true at once: the exposure pattern made the signal visible, and it made the signal matter more.

Numbers This Page Will Not Give You

Readers reasonably want a figure — one in how many, how many cases in total, what dose, how long before onset. This page declines to supply several of those, and says so explicitly rather than quietly omitting them, because a confident number here would be a guess dressed as data.

1. No incidence rate. Calculating one requires a numerator of all cases and a denominator of all users. The numerator comes from passive reporting, which under-counts by an unknown factor; the denominator was never measured, because giloy was bought informally and also harvested from garden fences. Neither figure exists.
2. No total case count is asserted here. The published series are described qualitatively, as sets of patients collected at identified centres over identified periods. Any total quoted from memory would be a different number depending on which reports were counted and how overlapping series were handled.
3. No threshold dose or safe duration. Idiosyncratic immune-mediated liver injury is characteristically not dose-dependent in the way a direct toxin is, and the exposures were unmeasured. There is no published safe-dose figure and inventing one would be worse than admitting there is none.
4. No latency window. Onset in the reports spanned weeks to months. That is a description, not a rule, and it should not be read as “safe for the first N weeks”.
5. No fraction of cases attributable to species substitution. This is the central contested quantity and it is unknown, for the simple reason that the consumed material was not tested.

Refusing these numbers is not evasion. The clinically actionable content of this literature does not depend on any of them: it is that a serious immune-mediated liver injury has been repeatedly reported in association with a product taken by well people for no proven benefit, and that this is enough to change behaviour without a rate.

Warning Signs, and What They Should Trigger

Anyone taking guduchi, in any form, should know what early liver injury feels like. It is unglamorous and easy to attribute to something else, which is exactly the problem.

• Unusual fatigue that is disproportionate to activity and does not improve with rest.
• Loss of appetite, nausea, or a new aversion to food.
• Discomfort, fullness or ache under the right ribs.
• Dark urine — tea- or cola-coloured — and pale stools.
• Yellowing of the eyes or skin. By the time jaundice is visible, injury is well established.
• Generalised itching without a rash.
• Easy bruising or a tendency to bleed, which reflects impaired clotting-factor synthesis.
• Confusion, drowsiness or altered behaviour — a late and urgent sign requiring immediate emergency assessment.

What to do, in order: stop the product; keep the container, because it is the only physical evidence of what was actually consumed and may be testable; get liver blood tests promptly, and tell whoever orders them exactly what you were taking, including anything herbal, since transaminases are meaningless without the exposure history; and do not restart it to see whether the symptoms come back. Self-rechallenge is the one experiment that could confirm causality and it is also the one that can produce a far more severe second injury.

The corollary for anyone who intends to take guduchi anyway: baseline liver blood tests before starting, repeat testing if use continues beyond a few weeks, a defined reason and a defined end date rather than open-ended daily use, and authenticated material from a manufacturer that performs identity testing.

Who Should Not Take Guduchi At All

These are cautions drawn from the nature of the reported injury and from the herb’s own claimed mechanism. They are reasoning from what is known, not quotations from a regulatory monograph.

• Anyone with existing liver disease of any cause — viral hepatitis, fatty liver disease, alcohol-related liver disease, cirrhosis, or unexplained abnormal liver enzymes. A liver with reduced reserve is the wrong place to run an idiosyncratic-risk experiment.
• Anyone with diagnosed autoimmune hepatitis or another autoimmune disease. The reported injury is immune-mediated and the herb is sold as an immune stimulant; the combination is the least attractive one available.
• Anyone taking a drug with recognised hepatic toxicity — including anti-tuberculous therapy, methotrexate, azathioprine, certain antifungals and antiretrovirals, and regular high-dose paracetamol. Adding an agent implicated in liver injury to a regimen already monitored for liver injury confounds the monitoring as well as compounding the risk. See acetaminophen overdose for how narrow the hepatic margin can be.
• Transplant recipients and anyone on immunosuppressive therapy. An immunostimulant claim directly opposes the therapeutic goal, and organ rejection is a catastrophic failure mode.
• Anyone on immune checkpoint inhibitors for cancer, where immune-mediated hepatitis is already a recognised treatment complication that would become impossible to attribute.
• Pregnancy and breastfeeding. Not because harm has been shown, but because adequate human safety data do not exist — which is a statement about absent evidence, not about safety.
• Children, for the same reason, and with the added consideration that dosing has never been established for them.
• Anyone who cannot verify what species they have bought. Given the substitution question, this is a substantive caution rather than a formality.

A Reasonable Position to Hold

Pulling the threads together, here is a position that is defensible on the published literature and does not require deciding the contested questions:

1. The cases are real and have been replicated across centres. They are not an invention and not an attack on a traditional system of medicine.
2. Whether authenticated T. cordifolia is the cause remains genuinely open. The species-substitution hypothesis is plausible, was raised for good methodological reasons, and has not been demonstrated for the reported patients.
3. Either way, the consumer-facing conclusion barely changes, because the substitution scenario indicts the supply chain the consumer actually buys from.
4. The benefit side of the ledger for preventive use is not merely small but unestablished, which is what makes even an uncommon serious harm decisive rather than merely notable.
5. Short, purposeful, monitored use of authenticated material in a person with a healthy liver is a different proposition from open-ended daily consumption of unverified powder by a well person for prevention. The second is what happened at scale in 2020–21, and it is the pattern this page is arguing against.

None of this makes guduchi uniquely dangerous among herbs, and nothing here should be read as implying that Ayurvedic medicine is uniquely unsafe. It makes guduchi a herb whose best human evidence is a safety signal — which is worth knowing before, rather than after, taking it daily for six months.

Key Research Papers

Every reference below is given as a PubMed topic search rather than a direct article link, deliberately. Author attribution in this literature is easy to get wrong — several of the reports are short communications, letters and author replies published close together in the same journals — and a misattributed citation on a safety page is worse than a search that lands the reader on the whole set.

1. Herbal immune-booster liver injury during the COVID-19 pandemic. The case series in the Journal of Clinical and Experimental Hepatology (2021) that opened the discussion, describing patients with liver injury whose common exposure was a herbal immune booster, predominantly Tinospora cordifolia, with autoimmune features in a subset. PubMed search
2. Multicentre nationwide study of giloy-associated liver injury. The larger Indian multicentre collection published in Hepatology Communications (2022), which is the replication that made the signal difficult to dismiss as a single-centre artefact. PubMed search
3. The autoimmune-hepatitis-versus-drug-induced debate in these cases. Editorial and correspondence in the hepatology literature arguing whether the picture represents drug-induced autoimmune-like hepatitis or unmasked de novo autoimmune hepatitis. PubMed search
4. Responses, rebuttals and author replies. The exchange over botanical authentication, concomitant products and species confusion, including replies from the original authors and commentary from Ayurvedic researchers. PubMed search
5. Can a classical Ayurvedic hepatoprotectant cause liver damage? Analyses from the Ayurvedic and integrative-medicine literature, including in the Journal of Ayurveda and Integrative Medicine (2023), examining the apparent contradiction between guduchi’s traditional hepatoprotective reputation and the reported injuries. PubMed search
6. Tinospora crispa hepatotoxicity. The literature on the related species with documented liver toxicity, which is the substrate of the misidentification argument. PubMed search
7. Drug-induced autoimmune-like hepatitis as a category. Reviews describing the syndrome, how it is distinguished from classical autoimmune hepatitis, and its behaviour after steroid withdrawal — the framework the giloy debate is being conducted inside. PubMed search
8. Causality assessment in herb-induced liver injury. Methodological literature on structured causality instruments applied to herbal and dietary supplement hepatotoxicity, and on why attribution is so difficult. PubMed search
9. Herbal and dietary supplement hepatotoxicity generally. The wider field, including registry-based studies showing the rising share of acute liver injury attributed to supplements. PubMed search
10. Under-reporting of herbal adverse events. Pharmacovigilance literature on why supplement-related harms are systematically under-captured, which is why a published series is a floor rather than a total. PubMed search

Connections

• All Herbs
• Tinospora / Guduchi — the main herb page, with traditional use, compounds, forms and dosage.
• Which Tinospora? Species, Adulteration and Identity — the substitution argument in full, and why it does not reassure.
• Immune Claims and the Rasayana Tradition — the benefit side of the ledger, and why it is unquantified.
• Metabolic, Arthritis and Other Research — the remaining claims at their real evidence tier.
• Autoimmune Hepatitis — the disease the reported cases resembled, including how it is diagnosed and treated.
• Cirrhosis — what unresolved chronic liver injury can progress to.
• Acetaminophen Overdose — the best-characterised drug-induced liver injury, useful as a contrast case.
• Nephrology & Hepatology — the wider liver and kidney disease library.
• Liver Function Tests — what transaminases, bilirubin and albumin actually measure.
• Hepatitis Panel — the viral testing that must be done before any injury is blamed on a supplement.
• Echinacea: Safety and Autoimmune Cautions — the same immunostimulant-versus-autoimmunity problem on another herb.
• Ashwagandha — the other major Ayurvedic rasayana, which has its own reported liver-injury signal.
• Milk Thistle in Hepatitis — a herb whose liver claims have actually been trialled, for comparison.
• Toxins — herbal product contamination, species substitution and heavy metals.

Back to Table of Contents