The trial was conducted at Peking Union Medical College Hospital in Beijing between 2020 and 2021. After a two-week run-in, participants were randomly assigned in a 1:1 ratio to berberine plus standard coronary disease therapy or standard therapy alone. The berberine dose was increased in three stages: 100 mg taken orally three times daily for approximately four weeks, then 200 mg three times daily for approximately four weeks, and finally 300 mg three times daily for approximately four weeks. Investigators monitored adherence with pill counts at three follow-up visits, and no participant fell below the prespecified 80% adherence threshold. One participant in the berberine group voluntarily withdrew before the final visit, leaving 11 participants with final clinical measurements in that group and 11 in the standard-therapy group.

Investigators assessed endothelial function using brachial flow-mediated dilation, with image analysts blinded to treatment assignment. Flow-mediated dilation increased numerically in the berberine group, from 6.7% to 9.7%, and the reported comparison did not meet the threshold for statistical significance at p = 0.07. The authors described the vascular result as a directional improvement rather than a demonstrated effect. The strongest lipid signal involved LDL cholesterol, which declined from 1.74 to 1.55 mmol/L with berberine plus standard therapy and from 2.33 to 2.01 mmol/L with standard therapy alone. A significant group-by-time interaction indicated a greater relative reduction in the berberine group, with a between-group difference at follow-up of p = 0.028. Total cholesterol showed only a marginal group-by-time interaction, so the LDL result was not accompanied by equally clear changes across the full lipid profile.

Triglycerides fell in the berberine group, while apolipoprotein B showed a similar direction of change that did not reach statistical significance. Several inflammatory mediators also decreased between baseline and week 12 in participants receiving berberine: interleukin-1α, interleukin-17, monocyte chemoattractant protein-1, macrophage inflammatory protein-1β and vascular endothelial growth factor. These were within-group observations, so the study does not establish that every decline was caused by berberine. High-sensitivity C-reactive protein declined in both groups, and the paper did not establish a significant between-group difference in that measure. Self-reported constipation assessment scale scores improved in the berberine group, whereas Bristol stool form scores showed no significant difference.

Shotgun metagenomic sequencing showed that berberine was associated with a broad restructuring of the fecal microbiome. The researchers sequenced 68 stool samples and assembled a catalog of 3,094,494 non-redundant microbial genes. Community diversity increased in the berberine group, while microbial community structure differed between baseline and follow-up within that group and between the groups at week 12. Forty-three species changed in relative abundance after treatment, and 17 were classified as berberine-responsive taxa after correction for multiple comparisons. Several organisms declined, including Megasphaera micronuciformis, Megasphaera elsdenii, a Clostridiales bacterium and an Azospirillum species, while Ruminococcus, Roseburia and Anaerostipes hadrus increased. The abundance of M. micronuciformis was negatively correlated with total cholesterol, lipoprotein(a) and C-reactive protein, although these correlations do not demonstrate that the bacterium caused the lipid differences.

The microbial response also involved organisms associated with fermentation and short-chain fatty acid production. Roseburia and Ruminococcus are capable of converting dietary polysaccharides or oligosaccharides into short-chain fatty acids, while Flavonifractor plautii and Eubacterium ventriosum, also associated with short-chain fatty acid production, increased in the combined berberine and standard-therapy group but declined in the standard-therapy group. Targeted fecal metabolomics detected a significant rise in butyric acid, from a median of 11.5 at baseline to 21.3 at week 12, with p = 0.04, and a smaller increase in isobutyric acid, from a median of 1.1 to 1.4, with p = 0.037. Acetic acid showed an upward trend during the middle of treatment before declining to a level only marginally above baseline. Gene profiles were consistent with this shift: pathways for acetate formation became progressively enriched, and the abundance of the but gene, which encodes an enzyme involved in terminal butyrate production, increased significantly after berberine treatment.

Berberine was particularly associated with changes in microbial aromatic amino acid catabolism rather than a simple increase in the amino acid precursors phenylalanine and tyrosine. Fecal L-dopa rose to 3.67 in the berberine group at follow-up, compared with 1.05 in the standard-therapy group, in a significant group-by-time interaction. L-arginine, 3-hydroxyphenylacetic acid and phenylpropionic acid followed a similar berberine-enriched pattern, with the reported follow-up L-arginine values being 22.6 and 9.8 in the berberine and standard-therapy groups, respectively. By contrast, tyrosine and L-tryptophan increased over time regardless of berberine exposure. The investigators interpreted those shared changes as a trajectory associated with standard therapy rather than a berberine-specific response.

A gene-level analysis offered a possible explanation for the metabolite shifts. The abundance of the microbial dopamine β-monooxygenase gene correlated with fecal L-dopa, the porA gene correlated with phenylacetic acid and the acdA gene correlated with phenylpropionic acid, with reported correlation coefficients of 0.6 to 0.7. Seventeen KEGG modules changed appreciably after berberine treatment, including pathways associated with tyrosine, phenylacetate, phenylalanine and tryptophan metabolism. The intervention increased several downstream products without a comparably clear rise in their upstream substrates. This pattern is consistent with the authors’ interpretation that berberine remodels microbial metabolic capacity, but the trial did not directly test whether those changes mediated the clinical outcomes.

Some fecal metabolites were statistically associated with circulating lipids, including L-dopa, hydroxyphenyl lactic acid, L-phenylalanine and imidazole propionic acid. These are associations within a small cohort, not evidence that any metabolite caused a lipid change. The authors noted that previous research had linked phenylpropionic acid to intestinal epithelial barrier function and aryl hydrocarbon receptor signaling, while L-dopa has a context-dependent relationship with cardiovascular biology. They also proposed that the observed fecal L-dopa enrichment might reflect a local gut–vascular axis, but acknowledged that dedicated tracer studies would be needed to determine whether the compound reaches the circulation or acts directly on host tissues. Berberine’s low oral bioavailability makes interactions with the gut microbiota plausible, yet direct effects on host tissues remain possible as well.

The trial is best understood as exploratory rather than definitive. Its two groups differed substantially in age and sex distribution, and potential residual confounding from smoking, alcohol consumption, diet and physical activity could not be fully excluded despite clinical annotation. Missing observations for some clinical parameters also prevented robust between-group comparisons of certain outcomes, while the limited cohort restricted analysis of downstream metabolites. The parallel design helped account for shared coronary disease secondary-prevention therapies, but the findings do not establish that berberine prevents heart attacks, improves survival or should replace established secondary prevention. The results instead identify measurable changes in LDL cholesterol, vascular function, inflammation and the gut ecosystem in a specific group of stable coronary artery disease patients after percutaneous coronary intervention, while leaving their clinical meaning for further investigation.

Subject of Research: Pilot randomized trial of berberine, vascular function, lipids, gut microbiota and fecal metabolites in stable coronary artery disease after percutaneous coronary intervention.

Article Title: Berberine improves lipid levels, vascular function, and gut microbial-dependent metabolites in stable coronary artery disease patients

Article References: Liu, H., Feng, S., Tang, M., Sun, J., Tian, R., & Zhang, S. (2026). Berberine improves lipid levels, vascular function, and gut microbial-dependent metabolites in stable coronary artery disease patients. Frontiers in Nutrition, 13, Article 1945710. https://doi.org/10.3389/fnut.2026.1945710

Image Credits: AI Generated

DOI: 10.3389/fnut.2026.1945710

Keywords: Berberine, Coronary artery disease, Gut microbiome, Short-chain fatty acids, LDL cholesterol, Endothelial function, Metabolomics, Metagenomics, Inflammation, Clinical trial, improves, lipid