Berberine for Blood Sugar: What 46 Trials Show (2026 Evidence Grade)
Yes. Across 46 randomized controlled trials, berberine lowered HbA1c by 0.73 points and fasting glucose by about 15 mg/dL compared with control in adults with type 2 diabetes (PMID 34956436). That is the headline, and it holds up across the largest pooled analyses in the literature. What the headline leaves out is who was studied, how long, and where the effect stops.
What the trials found
The 2021 meta-analysis that produced the 0.73-point figure pooled 46 trials in which berberine was taken alone or added to standard diabetes drugs. Two-hour post-meal glucose fell by 1.26 mmol/L, fasting insulin by 2.05 units, HOMA-IR by 0.71 and BMI by 1.07, with LDL cholesterol down 0.86 mmol/L and HDL up 0.17 mmol/L (PMID 34956436). The multicenter PREMOTE trial gives the cleanest standalone estimate: 409 adults with newly diagnosed type 2 diabetes across 20 Chinese hospitals, 12 weeks, double-blind. HbA1c fell 0.99 points on berberine versus 0.59 on placebo, a difference of 0.4 points, and berberine caused more gastrointestinal side effects than placebo (PMID 33024120). A 2019 meta-analysis of 28 trials with 2,313 patients found smaller but consistent effects: fasting glucose down 0.54 mmol/L, post-meal glucose down 0.94 mmol/L, HbA1c down 0.54 (PMID 30393248).
| Source | Design | Participants | HbA1c change versus control |
|---|---|---|---|
| PMID 34956436 (2021) | Meta-analysis, 46 RCTs | Type 2 diabetes | -0.73 points (95% CI -0.97 to -0.51) |
| PMID 33024120 (2020) | Multicenter RCT, n=409 | New type 2 diabetes | -0.99 vs -0.59 placebo (about -0.4 net) |
| PMID 30393248 (2019) | Meta-analysis, 28 RCTs, n=2,313 | Type 2 diabetes | -0.54 points (95% CI -0.93 to -0.15) |
| PMID 37679692 (2023) | Pilot RCT, n=34 | Prediabetes | 6.40% to 5.43% (branded product, small trial) |
| PMID 40029660 (2025) | Phase 2 RCT, n=113 | Type 2 diabetes | -0.7 points at 1,000 mg twice daily (berberine-ursodeoxycholate drug salt, not the supplement) |
What we do not know
The 28-trial meta-analysis ran subgroup analyses that every berberine buyer should read. The glucose effect became non-significant in trials longer than 90 days, at daily doses above 2 grams, and in patients older than 60 (PMID 30393248). Almost every trial lasted 8 to 16 weeks. None measured retinopathy, kidney disease, heart attacks or mortality, the outcomes diabetes care exists to prevent. Most trials came from Chinese-language databases of mixed methodological quality, and many tested berberine as an add-on to metformin or other drugs rather than alone. The prediabetes result, HbA1c dropping almost a full point in 34 people over 12 weeks, is far larger than anything the big meta-analyses show and comes from a single branded-product pilot (PMID 37679692). It needs replication before it means anything.
Featured Partner
Invest in the Infrastructure Behind Modern Medicine
As healthcare expands beyond hospital walls, the buildings and campuses supporting that shift are generating compelling returns for investors who move early. The Healthcare Real Estate Fund offers qualified investors direct access to a curated portfolio of medical office, outpatient, and specialty care facilities.
Learn More →How it works, as far as the data go
A phase 1 crossover study in 15 healthy men found a single dose of berberine increased glucose-stimulated insulin secretion during a hyperglycemic clamp without raising basal insulin, which the authors tied to the KCNH6 potassium channel in pancreatic beta cells (PMID 34556670). The PREMOTE investigators traced part of the glucose effect to gut bacteria: berberine suppressed deoxycholic acid production by Ruminococcus bromii (PMID 33024120). Both mechanisms are plausible, and both come from small or exploratory work.
Doses used in studies
These are doses reported by the named sources, not recommendations.
- Trials pooled in PMID 34956436 mostly used 0.9 to 1.5 g per day, typically 300 to 500 mg three times daily before meals, as berberine hydrochloride.
- The prediabetes pilot (PMID 37679692) used 500 mg three times daily for 12 weeks.
- The 28-trial meta-analysis (PMID 30393248) found effects disappearing above 2 g per day.
What this means for you
The evidence supports a short-term glucose-lowering effect in people who already have type 2 diabetes, in the range a modest diabetes drug produces. It is not evidence about prevention in healthy people, and it is not evidence about the long term. Berberine interacts with metformin, insulin and sulfonylureas, so the trial populations most likely to see a benefit are also the ones most likely to be on drugs that interact. See the full berberine evidence review for the HYPE Index grades on all 11 claims, the berberine versus metformin comparison, and the HYPE Index methodology.
