Berberine is a plant alkaloid that meta-analyses of randomized trials show lowers fasting glucose and A1C in people with type 2 diabetes — by roughly 0.5–0.9 percentage points of A1C in pooled analyses — and modestly improves triglycerides and LDL cholesterol. Typical doses studied are 900–1,500 mg per day in divided doses. The trials are mostly short, small and conducted in China, long-term safety is not established, gastrointestinal side effects are common, and it interacts with many medications. It is not a replacement for metformin or any prescribed treatment.
What berberine is
Berberine is a bitter, yellow alkaloid found in several plants used in traditional Chinese and Ayurvedic medicine — goldenseal, barberry, Oregon grape and Chinese goldthread (Coptis chinensis). It has been used for centuries against diarrhoea and infection; its interest for metabolic conditions is more recent, driven by a run of clinical trials since the early 2000s.
Supplements almost always contain berberine hydrochloride, usually in 500 mg capsules.
How it appears to work
Berberine has several proposed mechanisms, most demonstrated in laboratory or animal studies:
- Activation of AMPK, an enzyme that acts as a cellular energy sensor. Activating it increases glucose uptake into cells, reduces glucose production in the liver and promotes fat burning. Metformin activates the same pathway, which is the origin of the "natural metformin" comparison.
- Increased insulin sensitivity in muscle and liver.
- Slower carbohydrate breakdown in the gut through inhibition of the enzyme alpha-glucosidase.
- Changes to the gut microbiome, which may contribute to its effects on glucose and lipids.
- Increased LDL receptor expression in the liver, which explains its cholesterol-lowering effect.
Berberine is poorly absorbed — only a small fraction of a dose reaches the bloodstream — which is why doses are large and why some of its effects may be exerted in the gut rather than after absorption.
Evidence by use
| Use | Rating | What the trials show |
|---|---|---|
| Blood glucose in type 2 diabetes | Moderate | Meta-analyses pooling 37–46 randomized trials report reductions in A1C of roughly 0.5–0.9 percentage points and in fasting glucose of about 15 mg/dL versus control, alone or added to standard treatment. Most trials lasted 8–13 weeks and were conducted in China; quality was variable. |
| Cholesterol and triglycerides | Moderate | Consistent reductions in LDL cholesterol, total cholesterol and triglycerides in the same trial populations, of a size smaller than statins but larger than most other supplements. |
| Insulin resistance / prediabetes | Limited | Improvements in HOMA-IR in trials of people with diabetes or metabolic syndrome; little direct trial evidence in prediabetes, and no evidence that it prevents progression to diabetes. |
| Weight | Limited | Small reductions in body weight and BMI (around 1–2 kg) in some trials; effects are modest and inconsistent. |
| Polycystic ovary syndrome | Limited | A few trials suggest improved insulin sensitivity and ovulation, with small samples. |
| Gut infections / diarrhoea | Limited | Older evidence for infectious diarrhoea; not relevant to metabolic use. |
The overall picture: berberine reliably lowers glucose and lipids in short trials. What is missing is the evidence that matters most — large, long, well-conducted trials showing it reduces heart attacks, kidney disease or other complications, and that it is safe over years. No such trials exist.
Doses studied
Most trials used 500 mg two or three times a day (1,000–1,500 mg daily), taken before meals. A few used 900 mg per day. Taking it with meals reduces gastrointestinal side effects. There is no evidence that higher doses add benefit, and the poor absorption means more of a larger dose stays in the gut.
Side effects
Gastrointestinal effects are common: constipation, diarrhoea, bloating, cramping and nausea, affecting a meaningful proportion of trial participants. They are usually mild and improve with lower doses or taking it with food.
Hypoglycemia can occur when berberine is combined with insulin or sulfonylureas.
Berberine can cause jaundice and brain damage (kernicterus) in newborns by displacing bilirubin; it must not be used in pregnancy, while breastfeeding, or in infants.
Interactions
This is the part most product labels omit. Berberine inhibits several liver enzymes — including CYP3A4, CYP2D6 and CYP2C9 — that metabolize a large share of prescription drugs, and it affects the transporter P-glycoprotein. Raised blood levels have been documented for cyclosporine and are plausible for statins, some blood pressure medications, warfarin, some antidepressants and many others. It also adds to the effect of glucose-lowering and blood-pressure-lowering medications.
Speak to a pharmacist before taking berberine if you take any prescription medication. The interaction risk is real and poorly characterized, and a pharmacist can check your specific medicines.
Who should avoid it
Pregnant or breastfeeding women; infants and children; people with liver disease; people on cyclosporine or other drugs with a narrow safety margin; and anyone on glucose-lowering medication unless their prescriber is aware.
Where it appears in supplements
Berberine is sold on its own and is a common component of blood sugar support formulas, often combined with chromium, cinnamon, bitter melon or gymnema. When it appears in a multi-ingredient product, check whether the amount is stated: at 900–1,500 mg per day, berberine is bulky, and a formula that lists it inside a 500 mg "proprietary blend" cannot contain a trial-level dose. Our blood sugar supplement research pages examine this product by product.
Bottom line
For someone with type 2 diabetes, berberine is the supplement with the best chance of producing a measurable change in glucose and lipids. It remains a supplement: unregulated, unstandardized, without long-term safety data, without evidence for complications, and with real interaction risks. It belongs, if anywhere, alongside prescribed treatment with the prescriber's knowledge — not in place of it. The broader comparison of glucose-lowering ingredients puts it in context.
Frequently asked questions
Is berberine as good as metformin?
In a handful of small head-to-head trials, berberine lowered glucose by a similar amount to metformin over about three months. That is not the same as being equivalent. Metformin has decades of large trials showing it reduces complications and is safe long-term, and it is standardized and regulated as a medicine. Berberine has none of that.
What form of berberine is best?
Berberine hydrochloride (HCl) is the form used in almost all trials. Newer forms marketed for better absorption (phytosomes, dihydroberberine) have less human evidence for glucose outcomes.
How long does berberine take to work?
In trials, measurable changes in fasting glucose appeared within a few weeks and A1C changes over about three months, consistent with how A1C works.
Can I take berberine with metformin?
Some trials combined them and reported additional glucose lowering, but the combination also increases the risk of gastrointestinal side effects and low blood sugar. Do not add berberine to any diabetes medication without your prescriber's knowledge.
References
- Oxidative Medicine and Cellular Longevity. The Effect of Berberine on Metabolic Profiles in Type 2 Diabetic Patients: A Systematic Review and Meta-Analysis of Randomized Controlled Trials (2021). PubMed PMID 34956436. https://pubmed.ncbi.nlm.nih.gov/34956436/
- Frontiers in Pharmacology. Glucose-lowering effect of berberine on type 2 diabetes: A systematic review and meta-analysis (2022). PubMed PMID 36467075. https://pubmed.ncbi.nlm.nih.gov/36467075/
- National Center for Complementary and Integrative Health (NCCIH). Diabetes and Dietary Supplements: What You Need To Know (2021). https://www.nccih.nih.gov/health/diabetes-and-dietary-supplements-what-you-need-to-know
- U.S. Food and Drug Administration. Dietary Supplements (2024). https://www.fda.gov/food/dietary-supplements
This page discusses supplement ingredients or categories. It may mention a commercial product, with disclosure. It is general health information, not personal medical advice. Read our editorial policy and medical review policy.