Berberine explained: what it does, who it's for, and the evidence

What berberine is

Berberine is a bright-yellow alkaloid found in the bark and roots of several plants, including barberry, Chinese goldthread, goldenseal and Oregon grape. The plants have a long history in Ayurvedic and Traditional Chinese practice. Modern research has mostly studied the isolated compound rather than the whole plant.

Almost everything sold as a supplement, and nearly every clinical trial, uses berberine hydrochloride (HCl). The hydrochloride is simply the salt the molecule is bound to for stability. It is the same active alkaloid. You will not get a meaningful amount from food, because berberine is a medicinal plant compound rather than a dietary nutrient.

Two other forms appear in marketing. Berberine phytosome binds the molecule to a phospholipid carrier, and dihydroberberine is a chemically reduced version. Both are sold on the promise of better absorption, which is worth taking seriously, because plain berberine is poorly absorbed. More on that below.

How berberine works in the body

Most mechanism data come from cells and animals, so treat the following as hypotheses about how berberine behaves, not proof of what it does in a person.

• Cellular energy signaling. In mice and muscle cells, berberine activated AMPK, an energy-sensing pathway, and increased mitochondrial function in skeletal muscle (PMID 32745766). That is a mechanism finding, not a measurement of strength or muscle growth.
• Gut bacteria. A 2020 review argues that part of berberine's effect comes from changing which bacteria live in the gut, partly by acting as a bacteriostatic agent (PMID 32105754). That is the opposite of a probiotic, which adds live bacteria rather than restraining them.
• Gut hormones. In rats and a human intestinal cell line, berberine increased secretion of GLP-1, a hormone involved in appetite and blood sugar (PMID 19945441). A separate lab study traced part of this to a bitter-taste receptor pathway in gut cells (PMID 26206195). Reviews of plant compounds studied for GLP-1 effects describe the work as mostly preclinical (PMID 34981502).

There is a wrinkle. Very little berberine reaches the bloodstream after an oral dose, yet it still appears to affect how some other drugs behave (PMID 42653808). Part of its action may therefore happen in the gut itself, though that remains unsettled.

What human trials actually support

Berberine has been tested in dozens of trials, mostly in adults with blood sugar or lipid markers outside the normal range. Most used 900 mg to 1.5 g a day for one to three months, though some pooled analyses included doses as low as 300 mg a day. The table summarizes the findings that the studies here can support.

The pattern is consistent: modest, measurable changes in lab markers over weeks to months, at doses far above what many capsules deliver. None of this amounts to treating a disease. Results at 1,000 mg do not transfer to a product delivering a fraction of that.

Where the evidence is thin, mixed or absent

Berberine's reputation has outrun its data in several areas. It helps to separate them.

• Weight loss. Results conflict. One meta-analysis of 12 trials found modest reductions in body weight, BMI and CRP (PMID 32690176). Another of 10 trials found smaller BMI and waist changes but no significant change in weight (PMID 32379652), and a third of 12 trials in 849 adults found no significant change in weight or BMI (PMID 32147051).
• "Natural GLP-1" claims. The GLP-1 findings are lab and rodent work. No human data here show berberine raising GLP-1 enough to matter, and nothing equates it to a prescription GLP-1 drug.
• Mood, anxiety and sleep. The support is animal-only: stressed mice showed less depressive-like behavior (PMID 28842306), colitis-model rats less anxiety-like behavior (PMID 40569370), and insomnia-model rats improved sleep measures (PMID 32623439). A 2019 review of the mood research is preclinical and reports no human depression trial (PMID 30450823).
• Infections. Berberine stops parasites growing in a lab dish (PMID 1796883), and reviews describe broad antimicrobial activity in mechanistic terms without a human trial against a fungal, viral or urinary infection (PMID 42381319). In hospitalized adults with acute diarrhea, a 400 mg dose helped with E. coli diarrhea (PMID 3549923), but another trial found no clear benefit against cholera and none for non-cholera diarrhea (PMID 3935203). None of this supports self-treating an infection.
• Menopause. Trials used combination products, so berberine's own contribution is unknown (PMIDs 22313171, 20024591). A 2015 review proposed it based on effects seen elsewhere, not new menopause data (PMID 25785174).
• Energy and focus. Trials that tracked fatigue did so as one item within heart-failure or radiotherapy-toxicity measures, at 1.2–2.0 g and 900 mg a day respectively (PMIDs 12860219, 19757213). They say nothing about energy in healthy people.

Safety, interactions and who should avoid it

In trials, side effects were generally mild and gastrointestinal, and the fatty liver meta-analysis reported only mild GI complaints (PMID 38429794). But "generally mild in trials" is not the same as risk-free for everyone, and several cautions are well supported.

• Medication interactions. A 2026 review found berberine can raise blood levels of some other drugs and add to the effect of blood-sugar-lowering medications (PMID 42653808). Anyone taking prescription drugs should check with a pharmacist or physician first. The berberine interactions guide covers this in more detail.
• Heart rhythm. A 2026 systematic review of 22 studies documented QT prolongation, slowed heart rate and low blood pressure in some cases (PMID 40248925). A case report describes a 92-year-old man who developed a dangerous rhythm abnormality about two weeks after starting a berberine supplement; it resolved after he stopped (PMID 42647909). Case reports cannot establish how common this is, but they justify caution with heart conditions or QT-prolonging drugs.
• Pregnancy and newborns. Lab and animal work shows berberine displaces bilirubin from blood proteins, which is why it is advised against in pregnancy and in newborns (PMID 8513024).
• Liver. The picture is mixed. The LiverTox database says berberine has not been linked to enzyme elevations or clinically apparent liver injury at typical use (PMID 33242239), and a meta-analysis found no significant effect on ALT or AST overall (PMID 32690176). Yet a small phase I trial at 900 mg/day recorded one grade 3 transaminase elevation as possibly treatment-related (PMID 31619442), and in mice bred with the alpha-1 antitrypsin deficiency mutation, berberine worsened liver inflammation and fibrosis (PMID 39298444). That last finding is animal data, but it is a reason for people with known liver risk factors to ask first.

Choosing a product and reading the label

The most common label trap is the headline number. "800 mg" often describes the weight of the extract, not the berberine inside it. An extract standardized to 8% delivers roughly 64 mg of berberine per 800 mg, an order of magnitude below most trial doses. Look for the standardization percentage or a stated berberine content, then compare that figure with the doses in the research above.

Form matters mainly for absorption. In healthy volunteers, a manufacturer-tested berberine phytosome produced about 10 times higher blood berberine levels than unformulated berberine (PMID 34904017). In five healthy men, 100–200 mg of dihydroberberine produced higher blood berberine than 500 mg of plain berberine, though blood sugar and insulin did not differ (PMID 35010998). Both are small studies, and higher blood levels have not been shown to mean better outcomes. A comparison of berberine forms goes through the trade-offs.

Berberine is a supplement, so independent testing for identity and contaminants is a sensible filter. Our berberine topic page collects further reading on dose, timing and quality.

Who berberine is for, and who it isn't

Berberine is most plausibly relevant to adults curious about blood sugar or lipid markers who are working with a clinician and are not on interacting medications. The evidence base was built in that population, at those doses, over weeks to months. Even there, effects are moderate and are not a substitute for diet, activity or prescribed treatment.

For everyone else, the honest summary is modest. Berberine is a real compound with a real, if limited, human evidence base for metabolic markers, a lot of promising lab work, and a few safety signals that deserve respect. Claims beyond that tend to come from mice, rats and petri dishes.