Berberine And Fatty Liver: What The Trials Actually Measured
Berberine is the name on the Livpure list with the largest body of human research. Three pieces of it are about fatty liver, and they answer three different questions. Reading them together shows why a single headline is never enough.
Berberine has been tested in people with fatty liver disease. A 2024 pooled analysis of ten randomised trials in 811 patients found lower liver enzymes, lipids and insulin resistance. A 2024 trial of 1,500 mg a day for 12 weeks found lower ALT. A 2026 trial of 1 g a day for six months in 337 people found no change in liver fat or visceral fat measured by CT, though LDL cholesterol was a little lower. Each measured something different, and the Livpure pack prints no amount for berberine and no serving size to work one out from.
- The most-studied name on the list
- The pooled result: ten trials, 811 people
- A 12-week randomised trial at 1,500 mg a day
- A 6-month trial in 337 people: liver fat did not move
- How the results can all be true
- What none of the three covers
- The catch that is not about the liver
- The arithmetic without a serving size
- What to take from it
- Sources
The most-studied name on the list
Where berberine comes from, and what the NIH says about the research on it.
Berberine is the third name on the Livpure list. NCCIH, the National Institutes of Health centre that covers complementary health, describes it as a substance found in a variety of plants, including goldenseal, barberry and Oregon grape. It is a compound with a long history as a yellow dye and a folk remedy, and it is now a supplement ingredient in its own right.
Search suggestions finish the word berberine with the words fatty liver, liver enzymes and liver toxicity, so a reader arriving here has probably already asked the question this post is about. The NCCIH page on berberine says that studies have looked at people with health problems such as diabetes or fatty liver disease, and it says of weight loss that some studies suggest a benefit but that the evidence is not conclusive, noting that many of the studies in one 2022 review had a high risk of bias. It gives no fatty liver findings of its own. That is why the trials themselves matter.
Three of them are set out below, from the broadest to the newest. For each one the questions are the same. How many people took part, at what dose, for how long, compared with what, and what exactly was measured. That last question turns out to matter most.
The pooled result: ten trials, 811 people
What a 2024 meta-analysis of randomised trials found, and what its numbers mean.
In 2024 Nie and colleagues published a systematic review and meta-analysis in the Journal of Translational Medicine. They searched six databases up to September 2023 and pooled ten randomised controlled trials involving 811 patients with non-alcoholic fatty liver disease.
The results are reported as standardised mean differences, which express each change in units of standard deviation so that different measures can sit on one scale. Negative numbers mean a fall. Compared with control, berberine was associated with lower ALT (-0.72), AST (-0.79) and GGT (-0.62), all three liver enzymes. It was associated with lower triglycerides (-0.59), total cholesterol (-0.74) and LDL cholesterol (-0.53), and with lower insulin resistance on the HOMA-IR measure (-1.56) and a lower body mass index (-0.58). HDL cholesterol was the one outcome where the change was not statistically significant. The only adverse events reported were mild gastrointestinal ones.
By the usual statistical convention, a standardised difference around 0.5 is called moderate and 0.8 large, so the ALT and AST figures look substantial. There are caveats to hold against them. This is a pooled analysis of ten small trials, and the abstract does not tell us how the individual trials were designed or how long they ran. The reviewers themselves say clinical evidence remains inconclusive and describe berberine as showing promise as an adjunct therapy. And the outcomes that moved most are blood markers, which is a different thing from what a liver looks like on a scan.
A 12-week randomised trial at 1,500 mg a day
A single double-blind trial in 70 people, with a placebo.
The pooled analysis is made of studies like the one Koperska and colleagues published in 2024 in the Journal of Physiology and Pharmacology. Seventy people with metabolic dysfunction-associated fatty liver disease were randomly assigned one to one to berberine at 1,500 mg a day or placebo for 12 weeks. It was double-blind. The researchers measured body composition, liver blood tests and metabolic markers.
After 12 weeks the berberine group had a statistically significant fall in ALT compared with the control group (P = 0.0105) and in the de Ritis ratio, which compares AST with ALT (P = 0.0011). Total cholesterol changed in favour of berberine (P = 0.0009). Trunk fat fell in both groups, so that could not be credited to berberine. There were no significant differences in other lipid or glucose measures. The authors also report that, within the berberine group, changes in body mass index correlated with changes in ALT and AST.
The authors call berberine a bioactive compound that appears to have a positive effect on this condition, and add that longer studies are needed. That is a reasonable summary, and it is worth noticing what it does not say. It does not say the liver was smaller, less fatty or less scarred. The measures here were enzymes, cholesterol and body composition.
A 6-month trial in 337 people: liver fat did not move
The most recent and largest trial, and the one that measured the liver directly.
In January 2026 JAMA Network Open published a randomised clinical trial led by Lubi Lei and colleagues at 11 hospitals in China. It enrolled 337 adults who were obese, did not have diabetes and had metabolic dysfunction-associated steatotic liver disease, meaning fat in the liver from metabolic causes. A hundred and sixty-nine were assigned to 1 gram of berberine a day and 168 to a matching placebo, for six months. Adherence was about 90 percent in both groups, and the analysis followed the intention-to-treat principle.
The two primary outcomes were the change in visceral fat area, the fat around the organs in the abdomen, and the change in liver fat content, both measured by computed tomography. Neither differed between the groups. Visceral fat area changed by 1.4 percent more in the berberine group, with a 97.5 percent confidence interval of -2.4 to 5.2. Liver fat content differed by 0.9 percent, with an interval of -0.4 to 2.1. Both intervals span zero comfortably, so the fair reading is no detectable effect.
The secondary outcomes told a smaller story. Berberine was associated with larger falls in LDL cholesterol (7.72 mg/dL more than placebo), apolipoprotein B (3.42 mg/dL more) and high-sensitivity C-reactive protein, an inflammation marker (0.072 mg/dL more). Nothing else moved. The rate of adverse events was similar in the two groups. The authors conclude that berberine at 1 gram a day for six months had an excellent safety profile but did not reduce visceral fat area or liver fat content.
What makes this trial useful is its design. It is large, it is blinded, it ran for six months and it asked a direct question about the organ, using a scan rather than a blood test. It is also a different population from the other two: adults with obesity but no diabetes, so it does not contradict them.
About the product behind this blog
Livpure is a liver supplement in 60-capsule bottles with no Supplement Facts panel in anything supplied to us. Six ingredients are named and not one carries an amount. We publish the gap rather than filling it.
Order NowRead the label firstHow the results can all be true
Three studies, three questions, and one honest summary.
| Study | Design | Dose and length | What was measured | What moved |
|---|---|---|---|---|
| Nie 2024 | Meta-analysis of 10 trials, 811 patients | Not stated in the abstract | Enzymes, lipids, insulin resistance, BMI | All fell except HDL |
| Koperska 2024 | Randomised, double-blind, 70 patients | 1,500 mg a day, 12 weeks | Enzymes, body composition, lipids, glucose | ALT, de Ritis ratio and total cholesterol |
| Lei 2026 | Randomised, double-blind, 337 adults | 1 g a day, 6 months | Liver fat and visceral fat by CT | Neither. LDL, apoB and hs-CRP slightly lower |
The apparent disagreement dissolves once you look at the last two columns. The pooled analysis and the 12-week trial found changes in blood markers: enzymes, cholesterol. The 6-month trial measured fat directly, found no change in it, and still found a small change in LDL cholesterol. That is entirely consistent. A compound can lower some blood measurements without reducing the amount of fat in the liver, and only a trial that measures both can tell which of those a study is seeing.
It also explains why a headline that says berberine improves fatty liver, or one that says it does nothing, is wrong. Improved what? Enzymes yes, in the small and pooled trials. Liver fat on a scan, not in the biggest trial. Fibrosis or the biopsy score, not among the outcomes the three abstracts report. A supplement page that quotes only the first row is telling the truth about one column and leaving the rest blank, which is the selective quoting this blog spends most of its time pointing out on other pages.
A second reason for care is population. The pooled analysis and the smaller trial were in people with diagnosed fatty liver disease. The 2026 trial was in people with obesity and fatty liver who did not have diabetes. Results in one group do not carry across automatically to another, and none of them describes a healthy adult with a normal liver, which is who a general-purpose supplement is usually sold to.
What none of the three covers
Four limits worth stating before anyone carries a result to a bottle.
It is tempting to read a set of positive markers and one null scan as a verdict on the whole ingredient. Four things stop that.
The preparation. The abstracts do not describe the berberine product used in each trial, whether that is a particular salt, an extract or a formulation. Products differ, and a result with one preparation does not automatically apply to a capsule from another source.
The length. The longest of the three ran for six months. Fatty liver is a condition measured in years, and none of these studies says what happens at two or five.
The tissue. None of the three abstracts reports a biopsy result. Fibrosis, the scarring that matters most over time, is not among the outcomes they describe.
The person. The participants were adults with fatty liver disease or with obesity and fatty liver, in Poland and China, plus pooled trials whose settings the abstract does not give. That is not a description of a healthy adult with a normal liver, and it says nothing about someone who takes several medicines at once.
None of that is a mark against the trials, which asked reasonable questions and answered them. It is a reminder that a study result is a statement about a specific product, group and measure, and that a supplement label which names an ingredient without an amount, a form or a serving size has removed most of what a reader would need to know before applying it.
The catch that is not about the liver
What berberine does to medicines, in two studies and one NIH page.
Any honest piece on berberine has to say that it is not an inert ingredient. NCCIH reports that the side effects seen in research are mainly gastrointestinal, such as nausea, abdominal pain, bloating, constipation or diarrhoea. It says berberine may interact with medicines, giving the transplant drug cyclosporine as an example, and that it is likely unsafe for infants and may be unsafe in pregnancy or while breastfeeding.
The clearest human data on interactions come from a 2012 study in the European Journal of Clinical Pharmacology. Healthy men took 300 mg of berberine three times a day for two weeks, and researchers then measured how quickly probe drugs were cleared. Activity of the enzyme CYP2D6 fell, with a ninefold rise in the urinary ratio of dextromethorphan to its product. The ratio for losartan and its metabolite doubled, indicating reduced CYP2C9 activity. Levels of midazolam, a probe for CYP3A4, rose by about 38 to 40 percent. In plain terms, two weeks at 900 mg a day slowed the body’s handling of several probe drugs, some of them ordinary prescription medicines.
That is a reason to tell a pharmacist about berberine if you take a prescription, whatever the amount. It is not a reason to be alarmed by a study in healthy men at a known dose, and our ingredients page carries the interaction record as well. What this post adds is only that the safety record of berberine in the fatty liver trials, mild digestive symptoms and adverse events like placebo, sits beside an interaction record that is real.
The arithmetic without a serving size
What each trial dose would mean per capsule, given that the pack does not say how many capsules make a serving.
The Livpure pack prints no amount for berberine, and no Supplement Facts panel was supplied, so there is no serving size either. The only direction on the label is to take it with a big glass of water every day. A 60-capsule bottle lasts sixty days at one capsule a day and fifteen days at four, and nobody has said which applies.
That still allows some arithmetic, if we treat the serving size as unknown and show the possibilities. The trial doses were 1 gram and 1.5 grams a day. If every milligram of the capsule were berberine, the table shows what each capsule would have to hold.
| Capsules a day | To reach 1,000 mg a day | To reach 1,500 mg a day |
|---|---|---|
| 1 | 1,000 mg each | 1,500 mg each |
| 2 | 500 mg each | 750 mg each |
| 4 | 250 mg each | 375 mg each |
Five other named ingredients share the capsule, so those numbers are ceilings and not estimates, and we are not claiming to know what the capsule holds. The point is that every amount at which berberine has produced a result in these trials is a gram or more a day, and that a label naming it without an amount leaves a buyer unable to say whether the capsule is in that range or a small fraction of it. If it is a small fraction, the trials do not describe it, in either direction: not the enzyme changes, and not the interaction findings either.
What to take from it
Six sentences you can use.
- Berberine has more human research behind it than any other name on this list, including trials in fatty liver.
- Pooled trials and a 12-week trial found lower liver enzymes and some lipids. A large 6-month trial found no change in liver fat or visceral fat on CT.
- Enzymes and liver fat are different measurements, and a compound can move one without the other.
- Reported side effects are mostly digestive, and berberine slowed the clearing of several probe drugs in healthy men at 900 mg a day.
- The trial amounts are a gram or more a day, and this pack prints no amount and no serving size.
- Nothing here says a capsule of this product does what any trial found.
If you want to read the numbers that trial reports carry, ALT and AST: the numbers behind every liver supplement study explains what enzymes are and are not. The ingredients page shows all six names together, and Pictures Are Not Ingredients explains why a berry on a graphic is not an amount. As always, if you take medicine, are pregnant or nursing, or have a liver condition, ask a clinician before adding any supplement.
Sources
Every study, label and guideline named above, with the record it was checked against.
- Nie Q, Li M, Huang C, et al. The clinical efficacy and safety of berberine in the treatment of non-alcoholic fatty liver disease: a meta-analysis and systematic review. J Transl Med. 2024;22(1):225. https://pubmed.ncbi.nlm.nih.gov/38429794/
- Koperska A, Moszak M, Seraszek-Jaros A, Bogdanski P, Szulinska M. Does berberine impact anthropometric, hepatic, and metabolic parameters in patients with metabolic dysfunction-associated fatty liver disease? Randomized, double-blind placebo-controlled trial. J Physiol Pharmacol. 2024;75(3). https://pubmed.ncbi.nlm.nih.gov/39042390/
- Lei L, Wang B, Zhao L, et al. Berberine and adiposity in diabetes-free individuals with obesity and MASLD: a randomized clinical trial. JAMA Netw Open. 2026;9(1):e2554152. https://pubmed.ncbi.nlm.nih.gov/41543854/
- Guo Y, Chen Y, Tan ZR, et al. Repeated administration of berberine inhibits cytochromes P450 in humans. Eur J Clin Pharmacol. 2012;68(2):213-217. https://pubmed.ncbi.nlm.nih.gov/21870106/
- National Center for Complementary and Integrative Health. Berberine: what you need to know. https://www.nccih.nih.gov/health/berberine-and-weight-loss-what-you-need-to-know