CoChat

Berberine-Drug Interactions: Mechanisms, Clinical Relevance and Risk Stratification-A Narrative Review

Open access

The DOI matches its Crossref record; 120 of the 120 references it cites were found in a registry or an index; 145 of the 244 sources its claims cite were checked for support.

Abstract

Background: Berberine, an isoquinoline alkaloid present in Berberis spp., Coptis chinensis and Hydrastis canadensis, is among the most widely consumed metabolic-health supplements, popularized as “nature’s Ozempic”. Concurrent, often undisclosed use with prescription drugs is common in older adults, yet berberine is far from inert. Objective: To synthesize the evidence on berberine as a perpetrator of supplement–drug interactions, propose a four-axis mechanistic taxonomy, with product quality treated separately as a modifier of exposure rather than as a mechanism, and derive a clinically actionable risk-stratification framework. Methods: Structured narrative review, prepared per the SANRA quality criteria; PubMed/MEDLINE, Scopus, Web of Science and Embase were searched up to May 2026. Results: Despite very low systemic exposure (oral bioavailability 0.68% in rats; low ng/mL plasma concentrations in humans), high luminal, enterocytic and hepatic concentrations generate interaction liability, documented in humans for a few pairs and mechanistic for most, along four mechanistic axes: inhibition, and transcriptional induction, of CYP3A4, with CYP2D6/CYP2C9 inhibition that is quasi-irreversible through a metabolite-intermediate complex; transporter modulation (P-glycoprotein, OCT1/OCT2, and MATE1); pharmacodynamic additivity (hypoglycemia, hypotension, and QT prolongation); and microbiome- and gut-barrier-mediated effects, the last of these being a candidate axis rather than a demonstrated one. Product-quality variability is treated separately, as a modifier of exposure. The clinical anchor is increased cyclosporine exposure in renal-transplant recipients (AUC +34.5%; trough 29.3% above control). These elements are integrated into a three-tier risk-stratification framework that combines perpetrator potency, victim-drug vulnerability, and patient vulnerability, with each tier being linked to a defined pharmacy action. Conclusions: In patients on multiple medications, and particularly when berberine is co-administered with drugs of narrow therapeutic index, it should be managed as an active pharmacological perpetrator rather than as an inert supplement. Unstandardized product quality and an unsettled European regulatory framework, under which national limits differ by more than an order of magnitude, further widen the uncertainty around the dose actually delivered. Berberine use should therefore be elicited routinely at medication reconciliation and stratified by mechanism, by victim-drug vulnerability, and by patient risk, with particular attention to metabolic self-medication in the GLP-1 era.

Overview

CoChat Summary

TL;DR

Berberine is an active pharmacological perpetrator of supplement–drug interactions despite very low systemic bioavailability, acting through CYP inhibition/induction, transporter modulation, pharmacodynamic additivity, and a candidate microbiome axis. The review proposes a four-axis mechanistic taxonomy plus product-quality exposure modifier and a three-tier risk-stratification framework, anchored clinically by increased cyclosporine exposure in renal-transplant recipients.

Key findings

  • Berberine oral bioavailability is very low (0.68% in rats; below a few percent in humans), with plasma concentrations of a few ng/mL (roughly 1–10 nM after 400–500 mg), but nominal upper-small-intestinal luminal concentration near 5 mM; in vitro potencies cluster at 1–45 µM, so luminal concentrations exceed that band by two to three orders of magnitude while plasma falls below it by two to four.
  • The clinical anchor is cyclosporine: 12 days of berberine 0.2 g three times daily increased cyclosporine AUC by 34.5% and reduced apparent oral clearance by 40.4%; in a parallel controlled trial (n = 104), trough concentrations rose 88.9% from baseline in the berberine group versus 64.5% in the berberine-free group, with between-group differences at study end of 29.3% for trough concentration and 27.8% for concentration/dose ratio.
  • Proposed mechanisms include quasi-irreversible CYP2D6 inhibition via a metabolite-intermediate complex, CYP3A4/CYP2C9 inhibition, PXR-mediated CYP3A4 induction, transporter modulation (P-gp, OCT1/OCT2/MATE1; OCT IC50s 18.8, 1.02, and 10.7 µM), pharmacodynamic additivity (hypoglycemia, hypotension, QT prolongation), and a candidate microbiome/gut-barrier axis; a 16-volunteer probe-cocktail study found goldenseal extract reduced metformin exposure while increasing midazolam AUC (geometric mean ratio 1.43; 90% CI 1.35–1.53).
  • The risk-stratification framework combines perpetrator potency, victim-drug vulnerability (especially narrow therapeutic index), and patient vulnerability (age, renal/hepatic function, cardiac comorbidities, pregnancy); evidence grading gives only cyclosporine Grade A, with most pairs potential or mechanistic and Axis 4 having nothing above Grade D.

Methods

  • Structured narrative review prepared per SANRA quality criteria; no meta-analysis was performed. PubMed/MEDLINE, Scopus, Web of Science Core Collection, and Embase were searched without date restriction up to May 2026, with targeted searches for cyclosporine, metformin, warfarin, and midazolam, plus manual reference screening.
  • Included in vitro enzyme/transport studies, preclinical models, clinical pharmacokinetic/pharmacodynamic studies, case reports and series, systematic reviews, and regulatory documents; excluded works without interaction or mechanism data, unverifiable conference abstracts except relevant safety signals, and purely commercial literature. Data were extracted descriptively and synthesized thematically across four mechanistic axes and the exposure modifier.
  • Evidence was graded A–E: Grade A controlled clinical study using the victim drug itself; Grade B human clinical study with probe substrates, enzyme-activity, or pharmacodynamic endpoints; Grade C in vivo preclinical; Grade D in vitro or mechanistic/PBPK; Grade E no direct data. Case reports were graded separately.

Limitations

  • The paper states that as a narrative review it lacks an aggregated quantitative assessment and is more exposed to selection subjectivity than a registered systematic review; much mechanistic evidence comes from in vitro and preclinical studies in which berberine concentrations may exceed human systemic exposure; commercial product heterogeneity limits generalizability of effect-size estimates.
  • The paper states that prospective clinical interaction studies are almost absent: across the field it identified one controlled study in patients (renal-transplant recipients receiving cyclosporine), one probe-cocktail study in healthy volunteers, and one human enzyme-phenotyping study; everything else is preclinical, in vitro, or observational. It also notes likely publication bias overstating frequency and size, case-report attribution problems, under-reporting to pharmacovigilance, and a translational gap in which local concentrations are calculated but not measured in humans and PBPK models are not validated against clinical data.
  • Evident from the text and acknowledged by the authors: the patient-vulnerability dimension of the risk framework is expert interpretation because no study stratified a berberine interaction by age, renal/hepatic function, or cardiac comorbidity; the High risk row is anchored on Grade A evidence for cyclosporine alone, with other drugs placed there by mechanism and consequences rather than direct data; Axis 4 is a candidate mechanism, and no study has measured a co-administered drug’s disposition before and after berberine while characterizing the microbiota in the same subjects.

Model-written summary. Check it against the paper before relying on it.

Integrity signals

Registry

Crossref · 10.3390/ph19081313

Cites

120 of 120 references verified0 citation errors · 0 not checkable

Statistics

Nothing to recompute.

Figures

Figure 1 Site-of-interaction map and the bioavailability paradox. Despite low systemic bioavailability, berberine reaches concentrations orders of magnitude higher at the sites of absorption (intestinal lumen, enterocyte) and elimination (liver, kidney), where the majority of interactions occur [37,43,48]. Only the boxes numbered Axis 1 to Axis 4 are mechanistic axes of the taxonomy. The box labeled Exposure modifier-Product quality is deliberately left unnumbered: product identity, declared versus actual alkaloid content, and formulation are not a biological mechanism of interaction but a cross-cutting modifier that sets the perpetrator dose on which all four mechanistic axes operate (Section 2.2.5). It is drawn at the intestinal-lumen level because that is where the delivered dose first becomes relevant, and its position immediately below Axis 4 carries no hierarchical meaning. Panel ( A ) maps the successive compartments through which orally administered berberine passes-intestinal lumen, enterocyte, portal vein and liver, systemic circulation, and renal proximal tubule-and aligns each compartment with the mechanistic axis that operates at that site; the grey downward triangles indicate the direction of this transit and carry no quantitative meaning. Panel ( B ) compares, on a logarithmic scale, the concentrations attained in the intestinal lumen, at the gut wall and liver, and in systemic plasma; it is schematic and not drawn to scale. Original figure by the authors.
Figure 1 Site-of-interaction map and the bioavailability paradox. Despite low systemic bioavailability, berberine reaches concentrations orders of magnitude higher at the sites of absorption (intestinal lumen, enterocyte) and elimination (liver, kidney), where the majority of interactions occur [37,43,48]. Only the boxes numbered Axis 1 to Axis 4 are mechanistic axes of the taxonomy. The box labeled Exposure modifier-Product quality is deliberately left unnumbered: product identity, declared versus actual alkaloid content, and formulation are not a biological mechanism of interaction but a cross-cutting modifier that sets the perpetrator dose on which all four mechanistic axes operate (Section 2.2.5). It is drawn at the intestinal-lumen level because that is where the delivered dose first becomes relevant, and its position immediately below Axis 4 carries no hierarchical meaning. Panel ( A ) maps the successive compartments through which orally administered berberine passes-intestinal lumen, enterocyte, portal vein and liver, systemic circulation, and renal proximal tubule-and aligns each compartment with the mechanistic axis that operates at that site; the grey downward triangles indicate the direction of this transit and carry no quantitative meaning. Panel ( B ) compares, on a logarithmic scale, the concentrations attained in the intestinal lumen, at the gut wall and liver, and in systemic plasma; it is schematic and not drawn to scale. Original figure by the authors.CC BY · From the PDF

Verification

Claims the paper makes

78 Valid · 8 Not valid · 35 Not checked · 38 Own results and interpretations

51 more claims · 33 with their cited sources checked

Create a free account to see all 159 claims and to verify the 3 references still open.

See all claims — free account
  • It may state two separate claims as one. 95
  • Probably not a substantial claim. 22

Create a free account to see these claims.

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

Each contradiction between two passages listed here was confirmed by an independent check.

1 of 1

Citations

Cites

Show all 120 references

Cited by

See all 2

Metadata

Authors
Dumitru, C. N., Marcu, T., Dumitru, A. O., Tudor, S. S., Ferțu, I. D., Elisei, A.-M., & Goroftei, L.
Affiliations
  • "Dunarea de Jos" University of Galati (RO)
  • Research Centre in the Medical-Pharmaceutical Field, Faculty of Medicine and Pharmacy, “Dunărea de Jos” University of Galați, 35 Al. I. Cuza Street, 800010 Galați, Romania
Corresponding author
Dumitru, Alina Oana
Venue
Pharmaceuticals
Publisher
MDPI AG
Published
2026-08-20
Volume
19
Issue
8
Pages
1313
ISSN
1424-8247 (ISSN-L)
OpenAlex
W7203828452
Cited by
2
References
115
Citations by year
2026: 2
FWCI
2.07
Type
journal article
Language
en
Open access
Locations
  • Pharmaceuticals (publishedVersion, cc-by)
  • PubMed Central (publishedVersion, cc-by)
  • PubMed (publishedVersion)
  • DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion)
Topics
  • Health Sciences › Medicine › Pharmacology (medical) › Berberine and alkaloids research
  • Health Sciences › Medicine › Complementary and alternative medicine › Complementary and Alternative Medicine Studies
  • Life Sciences › Pharmacology, Toxicology and Pharmaceutics › Pharmacology › Pharmacogenetics and Drug Metabolism
Keywords
berberine, drug-supplement interactions, drug interactions, CYP3A4 inhibition, CYP2D6 inhibition, CYP2C9 inhibition, P-glycoprotein, MATE1, cyclosporine, kidney transplant recipients, QT prolongation, hypoglycemia, hypotension, gut microbiota, intestinal barrier, risk stratification, medication reconciliation, older adults, herbal supplements, regulatory limits, narrow therapeutic index

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

The claim verdicts this report counts are experimental: their accuracy is not yet measured.

A language model drew the claims from the paper on 2026-10-08, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).

A second model labelled each claim on 2026-10-08 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).

Cited sources are checked with a language model’s help (gpt-6-sol).

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review

Data sources

Bibliographic data: Semantic Scholar, OpenAlex, Crossref.

Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).

Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.