Kratom has caused liver injury in some of the people who used it. That much is established, at the level of evidence the case reports can carry: individual patients, assessed by doctors who ruled out other causes, who became jaundiced after taking kratom and recovered after stopping.
What the evidence cannot say is how often this happens. We found no study that has followed a defined group of kratom users over time and counted liver injuries, so there is no incidence figure, and anyone who gives you one is guessing. Nor do we know who is at risk, or how much it matters what form of kratom, at what dose, for how long.
This is an evidence review. It grades each source by the tier of evidence it represents, from weakest to strongest: anecdote, case reports and surveys, animal studies, human observational studies, controlled trials. A finding from a lower tier is not presented as if it came from a higher one.
What a case report can and cannot show
A case report describes one patient. A good one records what the person took, when the injury began, what other causes were excluded, and what happened after the substance was stopped. That can make it very likely that the substance caused the injury in that person.
It cannot show how common the injury is. To know that, you need a denominator: the number of people who used kratom and did not get ill. Case reports have no denominator by design. A hundred case reports could come from ten million users or from ten thousand, and the reports themselves would look the same.
Hold on to that distinction through everything below. Most of the liver evidence on kratom sits at the case-report tier, which is the second-lowest.
The strongest case evidence: a US registry with formal causality assessment
The most careful single source is the US Drug-Induced Liver Injury Network (DILIN), a research network that has prospectively enrolled patients with suspected drug-induced liver injury since 2003. Each patient is tested to exclude other causes of liver injury, and expert opinion grades how likely it is that a given product was responsible.
In 2020 the network published its kratom cases (Ahmad and colleagues, Drug and Alcohol Dependence). We read the full text. Its findings:
- Eleven cases were attributed to kratom with a causality grade of definite, highly likely or probable, up to 2019. Only three were enrolled in the first ten years. Eight were enrolled in the last three, four of them in 2019.
- Nine of the eleven were men, with a median age of 40.
- The median time from starting kratom to symptoms was 14 days (range 5 to 28). The paper reports that all eleven had symptoms and developed jaundice.
- The pattern of injury varied: two cholestatic (affecting bile flow), six mixed and three hepatocellular (damaging liver cells directly).
- By the network’s severity scale, two cases were moderate, six involved jaundice and hospitalisation, and three showed features of liver failure. None died or needed a transplant, and all eventually recovered, with a median recovery time of 90 days.
- Seven of the eleven reported alcohol use.
- Three products were analysed and contained mitragynine and 7-hydroxymitragynine with no pharmaceutical adulterants or contaminants detected.
The authors put the eleven cases at about 3% of the liver injuries the network attributed to herbal and dietary supplements, and about 1% of all cases it adjudicated. They read the recent rise as a sign that kratom “may be an increasingly frequent cause” of drug-induced liver injury, using the registry as a rough indicator of use in the population.
Note what that number is. It is a share of the cases one research network enrolled. It is not the share of kratom users who develop liver injury, and the paper does not claim that it is.
The study also looked at genes. Seven of eight patients of European descent carried one of two immune-system gene variants (HLA-B*57:01 or HLA-B*44:03), and both variants occurred in the group at a higher frequency than in the general population. With eight people, that is a lead for future research, and the authors themselves describe the genetic analysis as limited by the small sample.
Every published case, gathered together
Two reviews have collected the published case reports. Neither is open access, so for both we read only the abstract, and we report only what the abstracts say.
- Schimmel and Dart, 2020 (Drugs) identified 26 case reports and abstracts, plus 7 DILIN cases, 25 reports in FDA databases and 27 accounts on internet forums. They applied a standard causality scoring method (RUCAM) to the case reports. Symptoms began a median of 20.6 days after starting kratom. On a liver biopsy the injury looked mainly cholestatic, while blood tests showed a mixed picture. They concluded that kratom “likely causes liver injury based on the totality of low-quality human evidence”, and that “it remains unclear which subgroups of users are at heightened risk.”
- Calicdan and colleagues, 2026 (Journal of Addiction Medicine) systematically reviewed 31 studies describing 32 cases. Most were single-patient reports from the United States, mostly in adult men. Use of other substances and other illnesses was frequent, and only 7 of the 32 cases (22%) fully reported the dose, form, frequency and duration of kratom use. The injury was predominantly cholestatic and in most cases improved after kratom was stopped. In 4 cases liver function did not improve and the patient went on to a liver transplant. The authors note that formal causality assessment was reported inconsistently, but that many reports supported an association by excluding other causes and, in some cases, by rechallenge: the injury returned when kratom was taken again.
The four transplants are the most serious outcome in this literature. They come from case reports, so they show that kratom-associated liver injury can be severe enough to need a new liver. They say nothing about how likely that is.
The World Health Organization’s Expert Committee on Drug Dependence, reviewing kratom in October 2021 in a report published in March 2022, put the position in one sentence: “A number of cases of kratom-associated liver toxicity have been documented.”
What studies of whole groups of users show
The only way to measure risk is to study a defined population, including the people who did not get ill. Here the evidence is thin.
A cross-sectional study in Thailand, 2026 (Scientific Reports), compared 285 traditional kratom users, most of them long-term (235 had used it for five years or more), with 296 non-users from the same community in Surat Thani province. We read the full text. After adjusting for age, sex, body mass index, smoking and alcohol, it found no statistically significant differences in any liver marker: the liver enzymes AST, ALT and alkaline phosphatase, bilirubin and plasma proteins. Average liver enzyme and bilirubin values in both groups sat within normal reference ranges. This is tier-4 evidence, human and observational.
It found no difference in that population, and it is limited in three ways that matter:
- It is a snapshot. The authors say themselves that a cross-sectional design “precludes the establishment of a temporal sequence or definitive causal inference.”
- It looks at averages in current users. A rare reaction that strikes a few people within weeks of starting would barely move a group average, and anyone who had such a reaction may well have stopped using kratom before the survey reached them.
- It studied traditional use in one Thai community. Exposure was measured as the number of leaves used, which the authors call semi-quantitative. In the published case reports, by contrast, the form and dose were often not recorded at all: only 7 of the 32 cases in the 2026 review gave full exposure details. Whether findings in one setting carry to the other is not known.
Two industry-funded trials in our search gave kratom to healthy volunteers and measured liver tests. They are the only controlled evidence we found, and they point in different directions.
The larger and more relevant one gave dried kratom leaf powder alone (Huestis and colleagues, 2026, Therapeutic Drug Monitoring). It was randomised, double-blind and placebo-controlled, in 116 volunteers who had not used kratom recently: 49 on kratom at one of four escalating dose levels and 67 on placebo, first as a single dose and then daily for 15 days. We read the full text. During the 15-day dosing phase, raised ALT, a liver enzyme, was among the most common adverse events. It was reported at each of the three higher dose levels, most often at the highest, where it affected 3 of 12 participants (25%), against 1 of 60 on placebo (1.7%). Three participants were withdrawn for rising liver enzymes: one at the second-lowest dose, whose ALT climbed over four daily doses, and two at the highest dose with raised ALT and AST. The rises resolved after they stopped. Across everyone given kratom, AST or ALT went above three times the upper limit of normal in three participants and above five times in three, and none above ten times. The authors report that no case met Hy’s law, the standard regulatory marker for a drug likely to cause serious liver injury, and that some placebo participants also had raised liver enzymes of unclear cause. There were no serious adverse events, and the authors concluded that the product was “safe and well tolerated at the dose ranges tested”. The company that makes the product paid for the study, every author received consultant fees from it, and one author is affiliated with it.
The smaller one gave a liquid blend of kava and kratom (2024, Cureus). About 40 healthy adults per dosing period took it for six consecutive days. Only the lowest dose was compared with a placebo, in a randomised crossover; the middle and highest doses were given open-label, with no placebo. Its authors report no significant changes in liver or kidney function. Its adverse-event listings record mildly raised AST and ALT in one participant, during the placebo period. The trial was carried out by a contract research organisation hired by the drink’s maker; one author is a consultant to that company and another is its employee. Because the product combined kratom with kava, it cannot separate the effect of one from the other.
Read together, the trials show that kratom leaf can raise liver enzymes in some healthy people within days to two weeks of daily use, more often at the highest dose tested but not only there, and that in the leaf trial the rises went away after stopping. The leaf trial had 12 people at its top dose. Neither trial can say how often this happens in the wider population, and neither was designed to detect the kind of rare, serious injury seen in the case reports.
What animal studies add
Animal studies (tier 3) can show whether a substance is capable of damaging the liver, and at what doses. They cannot show what happens in people.
In one 2015 study, male rats were given a standardised kratom leaf extract by mouth for 28 days at 100, 200 or 500 mg per kilogram of body weight (Frontiers in Neuroscience). There were six rats per group. The results section reports damage to the liver, kidneys and lungs at the two higher doses, including proliferation of bile ducts in the liver, although the paper’s discussion says toxic effects were seen at all doses. The blood tests did not follow the same pattern: AST was significantly raised only at the lowest dose, and ALT did not differ significantly at any dose. These are high doses of a concentrated extract given to small groups of rats, so the result tells you the extract can harm a rat liver at those exposures and little more.
Putting the tiers together
| Question | Best evidence | Tier | What it shows |
|---|---|---|---|
| Can kratom cause liver injury? | DILIN registry, case reports with exclusion of other causes | 2 | Yes, in some people |
| How severe can it be? | Case reports, systematic review | 2 | Usually recovers after stopping; four reported transplants |
| How often does it happen? | No study measures this | None | Unknown |
| Do long-term users have abnormal liver tests on average? | One Thai cross-sectional study | 4 | Not in that population, after adjustment |
| Does kratom raise liver enzymes in healthy volunteers? | One placebo-controlled trial of leaf powder, industry-funded | 5 | In some people within 15 days, most often at the top dose; reversible in the trial |
| Who is at risk? | DILIN gene findings in eight people | 2 | Unknown; one lead to follow up |
| Is the extract capable of harming the liver? | One rat study, six per group | 3 | Tissue damage at 200 and 500 mg/kg (discussion says all doses); enzyme results inconsistent |
Both of the claims you will meet most often go further than this evidence. “Kratom destroys your liver” turns a set of case reports into a general risk, when nobody has measured the risk. “Kratom is safe for your liver” leans on studies that were never designed to catch a rare reaction, and ignores eleven registry cases, four transplants and the enzyme rises in the leaf trial.
The honest summary is narrower than either. Liver injury from kratom is documented, it is usually reversible, it has occasionally been severe, and how likely it is remains unknown.
How we found these papers
Europe PMC was searched on 17 October 2026 with the query (TITLE_ABS:(kratom OR mitragynine OR "Mitragyna speciosa")) AND (TITLE_ABS:(liver OR hepatic OR hepatotoxicity OR hepatotoxic OR hepatitis OR "drug-induced liver injury" OR DILI OR cholestasis OR cholestatic OR jaundice OR transaminase OR aminotransferase OR ALT OR AST OR "liver function" OR hepatocellular)), which
returned 109 records. A first version of the query, without the terms aminotransferase, ALT, AST and
“liver function”, returned 106 and missed the kratom leaf trial; the independent fact-check found
it. Seventeen of the records are individual
case reports with a liver term in the title. We did not cover those one by one: the two reviews
above collect them, and the DILIN series applies a consistent causality method to its own cases.
We read in full: the DILIN series (the authors’ manuscript in PubMed Central), the Thai cross-sectional study, both trials and the rat study. The two case-report reviews are not open access, and we read their abstracts only.
One 2025 review of case reports of acute adverse effects (Smallets and colleagues, Frontiers in Pharmacology) was read and is not relied on here. The paper declares no funding for this work. Its authors disclose that their consulting group “has been engaged by processors or distributors of kratom products in various litigation matters and advisory roles”, and that one author has served as an expert in kratom litigation. We set it aside because it is a secondary review and the case reports it draws on are covered above; the two trials, which are also tied to industry, are primary data, and their funding is stated beside them.
This page is not medical advice. For reference, the 2020 review by Schimmel and Dart lists the common presenting signs and symptoms in the cases it reviewed as abdominal discomfort, jaundice (yellowing of the skin or eyes), itching and dark urine.