Multi-omics analysis of bariatric surgery's impact on type 2 diabetes and prediabetes

Abstract

Bariatric surgery is a promising intervention for managing obesity-related metabolic disorders; however, its effects on cardiovascular disease (CVD) in individuals with type 2 diabetes (T2D) remain poorly understood. To address this gap, we utilized a multi-omics approach to investigate how bariatric surgery influences Emirati individuals with T2D and prediabetes, focusing on the early detection of CVD markers. We longitudinally profiled 19 patients over a 9-month period, collecting omics data, including whole-genome sequencing (WGS), protein immunoassays, untargeted metabolomics, and 16S rRNA sequencing to determine gut microbiome composition. Through this study, we showed that bariatric surgery reduced CVD risk and inflammatory biomarkers while inducing changes in the gut microbiome and metabolomic profiles. We identified four inflammatory biomarkers (FGF-basic, TNFSF13, IL-8, and IL-1Ra) that were significantly altered following surgery (p < 0.05). Additionally, we identified 98 metabolites that showed significant changes after surgery that are involved in folate biosynthesis, glycerophospholipid, and retinol metabolism. Eighteen microbial genera were found to differentiate between the pre- and post-surgery states. Our analysis revealed four microbial genera (Enterobacter, Enterococcus, Gemella, and Erysipelotrichaceae UCG-003) associated with two T2D SNPs (rs11830243 and rs6978118) and three CVD SNPs (rs9490306, 62207434, and rs34606058). These genera formed the network's central hub, connecting host genetic variants, metabolic pathways, and clinical data, highlighting their role in host-microbiome interactions. The study quantifies the impact of phenotypic factors on CVD progression among UAE nationals, contributing to a deeper understanding of cardiometabolic health within this population.

Keywords: Bariatric surgery; Cardiovascular disease; Omics; Precision medicine; Prognosis; Risk predictor; Type 2 diabetes.

Fig. 1

Patient demographics and changes in inflammatory biomarkers. (a) Bubble plot illustrating patients cluster by sex and surgery type based on BMI and age. (b, c) The percent mean increases in EBWL and TWL after surgery. (d) BMI decreases over time after surgery in patients of both genders. (e) Significant changes in inflammatory biomarkers measured before surgery and at 3-, 6-, and 9-months post-surgery. Colored dots represent the patients, while the purple dotted line is a visual guide that indicates the transitions between time points. Results were analyzed using a one-way ANOVA with Bonferroni post-hoc test for time differences. Data are expressed as mean ± standard deviation. * p < 0.05.

Fig. 2

Enrichment of metabolites and metabolic pathways. (a) Bar plot showing the fold changes of differential metabolites before and after bariatric surgery. Each bar represents an individual metabolite, with purple indicating increased pre-surgery levels and blue indicating decreased post-surgery levels. The fold changes were calculated as the ratio of metabolite levels in post-surgery samples to pre-surgery samples and log-transformed (Log₂ FC). The horizontal axis represents the magnitude of change, while the vertical arrangement shows the metabolites sorted by fold change. The complete list is provided in Supplementary Data Fig. S4. (b) A bubble diagram illustrating an overview of KEGG metabolic pathway enrichment. Each bubble corresponds to a specific pathway. The statistics of the enriched pathways are summarized in Supplementary Table S1.

Fig. 3

Microbial composition and diversity analysis between patients’ gut microbiome before and after surgery at the genus level. (a) The plot shows the relative abundance of the bacterial genus in each patient pre-surgery and 3-, 6-, and 9-months post-surgery. (b) The average order level abundance across all patients pre- and post-surgery. (c) Alpha diversity plot comparing pre- and post-groups using Chao1, Shannon, and Simpson indices, analyzed using the Wilcoxon test. The overall alpha diversity metrics showed no significant difference in genera richness or evenness (Chao1 p = 0.87, Shannon p = 0.97 and Simpson p = 0.87). (d) The beta diversity is presented as a PCoA using the Bray–Curtis (dissimilarity) distance comparing the two groups. The density plots on the axis show the resulting distribution of the samples on each axis. A PERMANOVA test was performed to check for significance between the two groups (p = 0.0001).

Fig. 4

Differential abundance analysis to determine the differences in the abundances of individual taxa at the genus level between pre- and post-surgery groups. (a) The fold-change plot shows the significant bacteria genera distinguishing between pre- and post-surgery groups. The bacteria are in descending order of their fold-change, with the most enriched in post-surgery to the most enriched before surgery. Streptococcus is the most significant bacteria genus enriched post-surgery. (b) The accompanying table details relative abundances, absolute differences, p-values, and FDR-adjusted q-values of the significant bacteria genus. (c) The functional analysis shows the 72 significantly enriched pathways that distinguish between pre- and post-surgery groups.

Fig. 5

Multi-omics interaction network analysis in bariatric surgery patients. Circular chord diagram illustrating key associations between host genetic variants, bacterial genera, metabolic pathways, and clinical measures. The left side of the diagram represents bacterial genera, while the right and bottom sections correspond to metabolic pathways. The upper section displays SNPs and clinical data.