Bright yellow, bitter and extracted from barberry bark and roots, berberine has become one of France's most sought-after supplements. Social media presents it as a natural alternative to diabetes treatments. Clinical trials do show effects on blood glucose and cholesterol, but mainly in people with disease at high doses. Health authorities remain very cautious. Meta-analyses, mechanisms, side effects, interactions, contraindications and ANSES/EFSA opinions: what is actually known.
Key takeaways
- In 37 type 2 diabetes trials, berberine lowered glycated haemoglobin by an average 0.6 percentage points at 0.9–1.5 g daily.
- It also reduced LDL cholesterol by around 0.25 g/L in eleven often low-quality trials.
- Digestive symptoms are common; it inhibits several liver CYP enzymes and interacts with many treatments.
- ANSES cannot guarantee safety; in 2026, an EFSA draft opinion could establish no safe intake.
What is berberine?
Berberine is an alkaloid, a yellow nitrogen-containing molecule. It occurs in barberries (Berberis vulgaris, Berberis aristata), Chinese goldthread and goldenseal. These plants have long been used in Chinese and Ayurvedic medicine, notably for diarrhoea.
Its glucose effect was noticed when given for diarrhoea to diabetic patients. No health claim is authorised in the EU for these plants. Supplements usually contain an extract standardised to berberine hydrochloride, such as Hexagon® berberine, providing 291mg Berberis aristata extract, including 116mg berberine hydrochloride, per capsule.
Berberine and blood glucose: what do meta-analyses show?
In brief: a real but moderate effect measured in people with diabetes. The largest meta-analysis to date (2022) pooled 37 randomised trials and 3,048 type 2 diabetes patients, including 36 Chinese trials. At 0.9–1.5 g daily, usually for three months, berberine lowered on average:
- fasting glucose by 0.82 mmol/L (around 0.15 g/L);
- glycated haemoglobin (HbA1c) by 0.63 percentage points;
- two-hour post-meal glucose by 1.16 mmol/L (around 0.21 g/L).
Trial quality was judged moderate: randomisation and blinding were often poorly described. HbA1c reduction was greater when baseline exceeded 9% (−0.79 percentage points).
Two trials illustrate these figures. In a 2008 pilot study, 36 newly diagnosed adults received 0.5 g of berberine or metformin three times daily: HbA1c fell from 9.5% to 7.5% with berberine, an effect comparable to metformin. That same year, a placebo-controlled trial in 116 people with diabetes and elevated lipid levels found fasting blood glucose falling from 1.26 to 1.01 g/L with 1 g per day for three months. These findings concern medically supervised patients. Data in people with normal blood glucose are rare.
Does it lower cholesterol and triglycerides?
Yes, in trials. A meta-analysis of eleven randomised trials (874 participants) measured mean reductions of 0.65 mmol/L LDL (around 0.25 g/L), 0.61 mmol/L total cholesterol and 0.50 mmol/L triglycerides (around 0.44 g/L). HDL rose slightly. Authors nevertheless stressed “generally low” methodological quality.
In 2004, 32 high-cholesterol patients had 25% lower LDL and 35% lower triglycerides in three months. This work revealed a mechanism different from statins (below).
| Criterion | Average effect | Evidence |
|---|---|---|
| Fasting glucose | −0.15 g/L | 37 trials, type 2 diabetes |
| HbA1c | −0.63 percentage points | 37 trials, type 2 diabetes |
| LDL cholesterol | −0.25 g/L | 11 trials, low quality |
| Triglycerides | −0.44 g/L | 11 trials, low quality |
| Weight | −2.1 kg | 12 randomised trials |
Berberine and weight loss: a modest effect
Do not rely on berberine to lose weight. A twelve-trial meta-analysis found mean loss of 2.1 kg, around 1 cm less waist circumference and a slight CRP reduction, an inflammation marker. It found no liver enzyme effect (ALT, AST).
Berberine was also studied in fatty liver. In an open-label trial of 184 patients, 1.5 g daily plus lifestyle advice reduced liver fat by 52.7% in sixteen weeks, versus 36.4% with advice alone. Milk thistle is another plant studied here: see our milk thistle article.
How does it work? AMPK and LDL receptors
Two mechanisms are well described in the laboratory. First, berberine inhibits mitochondrial respiratory complex I, like metformin. The cell senses an energy shortage and activates AMPK, its “energy sensor”, improving insulin action. These results come from cells and rodents.
Second, it stabilises LDL receptor messenger RNA in the liver, which then takes up more circulating LDL cholesterol. Observed in hamsters and human cells, this differs from statins. Other possibilities, such as a gut microbiota action, remain under study.
Why is its bioavailability so low?
Berberine enters blood very poorly. Absolute bioavailability in rats is only 0.68%. Some absorbed material is pumped back into the intestine by P-glycoprotein. In patients taking 1.5 g daily, measured blood concentration was only around 7 ng/mL.
Effects may therefore partly involve the gut and liver, where it concentrates in animals. Derivatives such as dihydroberberine are better absorbed in rodents, but human comparisons remain rare. For formats, see hard and soft capsule forms.
Which doses have been studied?
Most trials used 0.9 to 1.5 g daily, often divided. ANSES identifies proven pharmacological effects from 400mg daily in adults, without excluding lower-dose effects. EFSA data show recommended Berberis aristata supplements provide a median 274 to 445mg daily depending on source: Berberis aristata preparations are the products concerned.
What are berberine's side effects?
Most are digestive: constipation, diarrhoea, nausea and abdominal pain. In the 2008 pilot, 34.5% had temporary digestive symptoms. ANSES also mentions hypoglycaemia and hypotension risk.
EFSA's draft adds more concerning points. Berberine is genotoxic in vitro, without enough animal data to resolve uncertainty. No adequate repeated-dose toxicity study exists for berberine alone. Two trial cases of elevated transaminases were recorded with uncertain association. Maternal and offspring toxicity was observed in rodents.
Drug interactions: the critical point
This is a well-documented human risk. In healthy men, two weeks of berberine (300mg three times daily) clearly inhibited three enzymes eliminating many medicines. In kidney transplant recipients, ciclosporin concentration was 29% higher with berberine (0.2 g three times daily) than without.
| Target | Measured effect | Affected examples |
|---|---|---|
| CYP2D6 | Strongly reduced activity | Dextromethorphan |
| CYP2C9 | Clearly reduced activity | Losartan |
| CYP3A4 | Midazolam exposure +40% | Midazolam, ciclosporin |
| Glucose, blood pressure | Possible additive effects | Diabetes medicines, antihypertensives |
Anyone on regular medication should discuss it with their doctor or pharmacist first.
Who should not take berberine?
ANSES recommends the following groups avoid berberine supplements:
- children and adolescents;
- pregnant or breastfeeding women;
- people with diabetes;
- people with liver or heart disorders.
For pregnancy and breastfeeding, a specific reason exists: in vitro, berberine displaces bilirubin from albumin around ten times more than phenylbutazone. In a jaundiced newborn, this could promote severe neurological complications, kernicterus. French labels must advise pregnant women against these products.
What do ANSES and EFSA say?
In 2019, following a DGCCRF request, ANSES concluded that “the safety of these food supplements cannot currently be guaranteed”. It could only establish an indicative toxicological value of 0.1mg daily for a 60 kg adult, probably exceeded by many supplements. Belgium limits this alkaloid family to 10mg daily.
Following a Commission request in 2022, EFSA submitted a 2026 draft opinion for public consultation on thirteen berberine-containing plant parts. It concluded data could establish no safe intake. At writing, the final opinion was unpublished; the Commission would then decide possible measures.
Frequently asked questions
Can berberine replace metformin?
No. Direct comparisons involve small trials, and ANSES advises diabetic people against it, notably because of hypoglycaemia and interactions. Never change diabetes treatment without medical advice.
Is berberine dangerous for the liver or pancreas?
Neither ANSES nor EFSA's draft mentions pancreatic effects among recorded adverse effects. For liver, a twelve-trial meta-analysis found no enzyme increase, but EFSA notes two transaminase elevations and ANSES advises against use in liver disorders.
Are berberine patches effective?
We found no human trial evaluating patches for weight or glucose. Published skin-route work concerns formulation or animals.
How long did trials last?
Usually three months. In the 2008 medically supervised diabetes pilot, glucose fell from week one; HbA1c reflects glucose with a lag of several weeks. This timing says nothing about healthy-person effects.
Does berberine lower blood pressure?
ANSES lists hypotension as an adverse effect. Caution is therefore needed with low pressure or antihypertensive treatment.
Can it be combined with resveratrol or NAD+?
These combinations are popular, but they have only been investigated in preliminary human research, with no robust trial. To learn more, read our articles on the myths and realities of resveratrol and on ways to increase NAD+.
Are other plants studied for sugar?
Yes, such as gymnema, temporarily blocking sweet taste on the tongue. Its effects and limits are detailed in gymnema and sugar cravings.
Conclusion
Berberine is among the few plant extracts tested in dozens of glucose and cholesterol trials. These mainly involve diabetic or high-lipid patients at high doses, with uneven methods. It interacts with many treatments, and neither ANSES nor EFSA could define a safe dose. If considering it, discuss it with your doctor first, especially with medication.
Sources
- Xie W et al. (2022). Glucose-lowering effect of berberine on type 2 diabetes: a systematic review and meta-analysis. Frontiers in Pharmacology, 13, 1015045. doi:10.3389/fphar.2022.1015045
- Dong H et al. (2013). The effects of berberine on blood lipids: a systemic review and meta-analysis of randomized controlled trials. Planta Medica, 79(6), 437-446. doi:10.1055/s-0032-1328321
- Yin J, Xing H, Ye J (2008). Efficacy of berberine in patients with type 2 diabetes mellitus. Metabolism, 57(5), 712-717. doi:10.1016/j.metabol.2008.01.013
- Zhang Y et al. (2008). Treatment of type 2 diabetes and dyslipidemia with the natural plant alkaloid berberine. Journal of Clinical Endocrinology & Metabolism, 93(7), 2559-2565. doi:10.1210/jc.2007-2404
- Kong W et al. (2004). Berberine is a novel cholesterol-lowering drug working through a unique mechanism distinct from statins. Nature Medicine, 10(12), 1344-1351. doi:10.1038/nm1135
- Asbaghi O et al. (2020). The effect of berberine supplementation on obesity parameters, inflammation and liver function enzymes: a systematic review and meta-analysis of randomized controlled trials. Clinical Nutrition ESPEN, 38, 43-49. doi:10.1016/j.clnesp.2020.04.010
- Yan HM et al. (2015). Efficacy of berberine in patients with non-alcoholic fatty liver disease. PLoS One, 10(8), e0134172. doi:10.1371/journal.pone.0134172
- Turner N et al. (2008). Berberine and its more biologically available derivative, dihydroberberine, inhibit mitochondrial respiratory complex I: a mechanism for the action of berberine to activate AMP-activated protein kinase and improve insulin action. Diabetes, 57(5), 1414-1418. doi:10.2337/db07-1552
- Chen W et al. (2011). Bioavailability study of berberine and the enhancing effects of TPGS on intestinal absorption in rats. AAPS PharmSciTech, 12(2), 705-711. doi:10.1208/s12249-011-9632-z
- Guo Y et al. (2012). Repeated administration of berberine inhibits cytochromes P450 in humans. European Journal of Clinical Pharmacology, 68(2), 213-217. doi:10.1007/s00228-011-1108-2
- Wu X et al. (2005). Effects of berberine on the blood concentration of cyclosporin A in renal transplanted recipients: clinical and pharmacokinetic study. European Journal of Clinical Pharmacology, 61(8), 567-572. doi:10.1007/s00228-005-0952-3
- Chan E (1993). Displacement of bilirubin from albumin by berberine. Biology of the Neonate, 63(4), 201-208. doi:10.1159/000243932
- ANSES (2019). Use of berberine-containing plants in food supplements (opinion on the safety of using plants containing berberine, request 2018-SA-0095). anses.fr
- EFSA (2026). Draft scientific opinion on safety of berberine-containing plant preparations: stakeholder presentation of 22 April 2026. efsa.europa.eu
Food supplements do not replace a varied, balanced diet or a healthy lifestyle. Ask a healthcare professional for advice if you are taking medication, pregnant or breastfeeding.
