Berberine: What the Trials Measured and What They Did Not

What berberine is

Berberine is an isoquinoline alkaloid found in the roots, stems, rhizomes, and bark of several plants, including barberry, tree turmeric, and goldenseal. Plants containing it have a long history in traditional Chinese medicine and in Ayurvedic practice, mostly for digestive complaints. Supplements generally sell it as berberine hydrochloride. The modern human research tests it against metabolic markers rather than against how people feel. Trials measure fasting plasma glucose, postprandial glucose, HbA1c, total cholesterol, LDL cholesterol, HDL cholesterol, triglycerides, and insulin resistance calculated as HOMA-IR. Those are lab values. A shift in a lab value isn't the same thing as avoiding a heart attack, and writing as if the two are interchangeable is where a lot of supplement coverage goes wrong.

The bioavailability problem

Here's the fact that shapes everything else about this compound.

Taken by mouth, very little of it reaches the bloodstream.

France's food safety agency, ANSES, reviewed the pharmacokinetic literature in 2019 and concluded that oral bioavailability appears very low, with in vivo studies reporting under 1%. A 2024 study in Clinical Pharmacology and Therapeutics gave berberine to 44 healthy volunteers and measured what came out in urine afterward. The median recovery was 0.003% (mol/mol) of the dose, counting both unchanged berberine and its phase I metabolites. The authors described it as the largest human pharmacokinetic study of berberine to date, and noted that even if phase II metabolites exceeded everything else a hundredfold, bioavailability would still land below 1%.

That doesn't make the trial results meaningless. It changes what they probably mean. If a compound barely reaches the blood, whatever turns up in a trial is more likely to come from what it does in the gut, where the concentration is high. The 2024 authors wrote the same thing: the effects are more likely due to direct intestinal actions than systemic ones.

Mechanisms get proposed anyway. AMP-activated protein kinase activation, increased LDL receptor expression in liver cells, and reduced PCSK9 secretion all appear in the literature, and most of that work was done in cells or in animals. A mechanism shown in vitro doesn't tell you what happens in a person.

What the glucose trials measured

The trial everyone cites is Zhang and colleagues, published in the Journal of Clinical Endocrinology and Metabolism in 2008. It randomized 116 adults with type 2 diabetes and dyslipidemia to 1.0 g of berberine a day or placebo for three months. In the berberine group, fasting plasma glucose went from 7.0 to 5.6 mmol/L, postload glucose from 12.0 to 8.9 mmol/L, and HbA1c from 7.5% to 6.6%. Total cholesterol, LDL cholesterol, and triglycerides all fell. Five participants had mild to moderate constipation.

A smaller trial from Yin and colleagues in Metabolism the same year put 36 people with newly diagnosed type 2 diabetes on 0.5 g of berberine three times a day or on metformin for three months. Both arms showed drops in HbA1c, fasting and postprandial glucose, triglycerides, and insulin resistance. The 2015 meta-analysis that pooled this comparison found no statistical difference between berberine and oral glucose-lowering medicines. The most recent pooling, Zhao and colleagues in the Journal of Nutrition in 2023, found 20 placebo-controlled trials with 1,761 participants. Berberine lowered fasting glucose by 0.52 mmol/L, HbA1c by 4.48 mmol/mol, fasting insulin by 2.36 mU/L, and HOMA-IR by 0.85. The fasting glucose reduction looked larger in women, in people who already had diabetes, and in Asian participants. The authors flagged those subgroup findings as needing replication, since they rest on few studies.

One unit note, because it trips people up. HbA1c reported in mmol/mol is a different scale from the percentage on most lab reports. A change of 4.48 mmol/mol is a fraction of a percentage point.

What the lipid trials measured

Blais and colleagues published the tightest lipid analysis in Drugs in 2023. They included only randomized placebo-controlled trials in adults and found 18, with 1,788 participants, running 4 to 24 weeks. LDL cholesterol fell by 0.46 mmol/L (95% CI 0.62 to 0.30), total cholesterol by 0.48 mmol/L, triglycerides by 0.34 mmol/L, and apolipoprotein B by 0.25 g/L. HDL cholesterol rose by 0.06 mmol/L. The HDL change appeared to differ by sex, up 0.11 mmol/L in women and down 0.07 mmol/L in men.

Those are real differences on a lab report, and they're modest ones.

The review's own conclusion pointed at the hole in the record, asking for large trials that assess clinical outcomes rather than lipid levels alone.

The geography problem

This is the limitation that matters most, and it's structural rather than a flaw in any single study. Almost every randomized trial of berberine for metabolic outcomes was run in China. Fifteen of the 18 trials in the 2023 lipid review (83%) were conducted in mainland China or Hong Kong. The National Center for Complementary and Integrative Health puts it more broadly: most studies of berberine on body weight and cardiovascular risk factors were done in Asian countries, and very few in North America.

That's not a knock on the investigators. It's a question of external validity. Baseline diet differs. Body composition at a given body mass index differs. Background rates of insulin resistance differ. Gut microbiome composition differs, which matters more here than usual, since the leading hypothesis is that berberine acts in the gut. CYP2D6 handles part of berberine metabolism, and the frequency of reduced-function CYP2D6 variants differs substantially between East Asian and European-descent populations. A pooled estimate drawn almost entirely from one region may not travel.

The trials are also small and short. Eighteen lipid trials with 1,788 participants works out to under 100 people per trial on average, and the longest ran roughly six months. The 2008 diabetes trial lasted three months. That's long enough to move a lab value and nowhere near long enough to say anything about outcomes across years.

What the trials did not measure

Naming the gaps is more useful than restating the pooled numbers.

No trial counted heart attacks, strokes, or deaths. These studies were built around blood markers, and a marker stands in for the thing you actually care about.

No trial followed people for more than about six months. What happens with daily use across years hasn't been measured.

No trial compared berberine against a GLP-1 medication. Those drugs act on different receptors entirely, and the comparison circulating online has nothing behind it.

No trial mapped a dose response across a wide range, since nearly all of them landed between 0.9 and 1.5 g a day in divided doses.

What's in the capsule

One more gap sits between the trials and the shelf. A 2018 analysis in the Journal of Dietary Supplements bought 15 berberine supplements from U.S. vendors, pulled berberine out of three capsules from each product, and measured it by liquid chromatography tandem mass spectrometry. Average content came to 75% of the label claim, with a standard deviation of 25% and a range from 33% to 100%. Nine of the 15 products (60%) fell outside the 90% to 110% band that the U.S. Pharmacopeial Convention applies to pharmaceuticals.

Cost didn't predict potency. That's worth pausing on, because it means a shopper can't spend their way around the problem, and it means the trials were almost certainly testing material of uneven strength.

Safety and interactions

Across the 27 trials pooled in 2015 by Lan and colleagues, covering 2,569 patients, no serious adverse reaction was reported. The 2023 lipid review reached the same conclusion for the trials that reported harms.

The common complaints are gastrointestinal. NCCIH lists nausea, abdominal pain, bloating, constipation, and diarrhea.

In the lipid review, gastrointestinal adverse events turned up in 12 of the studies and ran more often in the berberine group than in the placebo group, at 2% to 23% versus 2% to 15%.

The interaction profile deserves more attention than it usually gets. Berberine inhibits CYP3A4 and CYP2D6 in the intestinal wall, and those enzymes clear a large share of prescription medicines. It also inhibits P-glycoprotein. A trial in 52 kidney transplant recipients found that berberine taken at 0.2 g three times a day for three months left cyclosporine trough concentrations 29.3% higher than in 52 recipients taking cyclosporine alone. In a six-patient pharmacokinetic sub-study, cyclosporine exposure rose 34.5% and clearance fell 40.4%. The authors attributed the effect to CYP3A4 inhibition.

NCCIH also advises against berberine during pregnancy and breastfeeding, and in infants, because exposure has been linked to a buildup of bilirubin that can damage the brain.

Frequently asked questions

Is berberine approved to treat any condition?

No. It's sold as a dietary supplement, and supplements aren't reviewed by the FDA for effectiveness before they reach the shelf. The trials described here measured lab values, mostly in adults who already had a metabolic diagnosis.

Why do people compare it with prescription weight management drugs?

The comparison is marketing. Pooled analyses of body weight find average losses under a kilogram, and one pooling found none at all. NCCIH says there isn't enough rigorous evidence to know whether berberine affects weight.

Does poor absorption mean it can't do anything?

It means the mechanism probably runs through the gut rather than the blood. That doesn't settle the effect size either way, and it's a reason to be careful with claims built on cell studies that used concentrations people never reach in plasma.

How long did the trials last?

Most ran 8 to 12 weeks. The 2023 lipid review covered trials from 4 to 24 weeks, and the 2008 diabetes trial covered three months.

What if I already take prescription medication?

That's a conversation for a pharmacist or prescriber rather than a page like this. CYP3A4 and CYP2D6 inhibition has been documented in people, and the cyclosporine result shows the effect can be substantial.

---

This article is for general information only and does not constitute medical advice. Statements about dietary supplements have not been evaluated by the Food and Drug Administration and are not intended to diagnose, treat, cure, or prevent any disease. Talk with your health care provider before using berberine, especially if you take prescription medication.

Sources

• Lan J, Zhao Y, Dong F, et al. Meta-analysis of the effect and safety of berberine in the treatment of type 2 diabetes mellitus, hyperlipemia and hypertension. J Ethnopharmacol. 2015;161:69-81.
• Blais JE, Huang X, Zhao JV. Overall and sex-specific effect of berberine for the treatment of dyslipidemia in adults: a systematic review and meta-analysis of randomized placebo-controlled trials. Drugs. 2023.
• Zhao JV, Huang X, Zhang J, et al. Overall and sex-specific effect of berberine on glycemic and insulin-related traits: a systematic review and meta-analysis of randomized controlled trials. J Nutr. 2023.
• Bloecher R, et al. Sex-dependent effects of CYP2D6 on the pharmacokinetics of berberine in humans. Clin Pharmacol Ther. 2024.
• Funk RS, Singh RK, Winefield RD, et al. Variability in potency among commercial preparations of berberine. J Diet Suppl. 2018;15(3):343-351.
• National Center for Complementary and Integrative Health. Berberine and weight loss: what you need to know.