Integrated metagenomic and metabolomic profiling across glycated hemoglobin-defined glycemic states identifies microbe-metabolite networks in glycemic dysregulation and coronary artery disease

Le résumé fourni par la source

Dysglycemia and hyperglycemia often coexist with coronary artery disease (CAD), but how gut microbiota and circulating metabolites differ across glycated hemoglobin (HbA1c)-defined glycemic categories and relate to cardiovascular risk remains incompletely understood. We aimed to characterize cross-sectional microbial and metabolic profiles across HbA1c-defined glycemic categories and to compare the patterns identified at the species and species-level genome bin (SGB) levels. In 600 participants from the ACS-GUT CARDIOME cohort, glycemic status was categorized as Normoglycemic, Dysglycemic, or Hyperglycemic according to HbA1c criteria. Fecal metagenomic sequencing and plasma metabolomics/lipidomics were performed. Species-level, SGB-level, polar-metabolite, and lipid features were associated with glycemic status using multivariable models adjusted for CAD status, demographic and anthropometric characteristics, lifestyle factors, and medication exposures. We used Light Gradient Boosting Machine (LightGBM) models to explore whether microbial and metabolic profiles could discriminate HbA1c-defined glycemic states. In individuals with elevated glycemia, we compared those with and without CAD to identify CAD-related multi-omics patterns. We additionally used SGB co-abundance networks, cross-comparison concordance, and matched association analyses to connect the glycemic and CAD findings. Across the HbA1c-defined categories, selected commensal taxa were less abundant in the elevated-HbA1c groups, accompanied by differences in sugars, amino-acid-related metabolites, and lipids. Species- and SGB-level analyses showed concordant alpha-diversity patterns, whereas significant community-level separation was detected only at the SGB level. Selected SGBs assigned to the same species also showed divergent glycemic associations, indicating that SGB-level profiling provided additional resolution for selected microbial associations beyond species-level summaries. SGB-level microbial models yielded higher area under the receiver operating characteristic curve (AUC) point estimates than species-level models across all three comparisons, with the largest difference observed for Hyperglycemic versus Normoglycemic classification. Combined microbial and metabolic models showed moderate discrimination for comparisons involving the Hyperglycemic group but limited discrimination between Dysglycemic and Normoglycemic participants. Among participants with elevated glycemia, CAD was associated with additional microbial and metabolic differences. An integrative analysis identified concordant microbial and metabolic candidates across glycemic and CAD comparisons and a denser SGB co-abundance network in CAD among matched participants with elevated glycemia. HbA1c-defined glycemic categories are associated with coordinated differences in gut microbial composition, SGB-level features, and circulating metabolites, with additional CAD-related multi-omics variation. These findings provide hypothesis-generating candidates for future longitudinal, mechanistic, and intervention studies of glycemic dysregulation and cardiometabolic risk.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.