Introduction

Over the past few decades, the prevalence of individuals with overweight and obesity has increased worldwide. Particularly, the incidence of obesity has doubled in over 70 countries since 1980, increasing the risk of several non-communicable diseases [1]. Although the exact causes of diabetes are not fully understood, several factors increase the risk of developing different types of diabetes mellitus (DM) [2]. For type 2 diabetes, these factors include being overweight or obese (having a body mass index [BMI] of ≥30). Obesity accounts for 80–85% of the risk of developing type 2 diabetes [3]. Both conditions have severe consequences and cause significant healthcare burdens and productivity losses [4]. Notably, obesity and DM have become global public health concerns.

Helicobacter pylori (HP) infection can cause gastric and duodenal ulcers, chronic gastritis, gastric cancer, gastric lymphoma, nervous system diseases, and rheumatic diseases [5]. This infection is more common in individuals with obesity, suggesting that obesity may be a risk factor [6,7,8,9]. Additionally, HP increases the risk of developing diabetes [10, 11]. Eradicating the infection may improve insulin sensitivity, thereby reducing the risk of diabetes in affected individuals [12,13,14,15]. However, it remains unclear whether eradication directly causes weight loss or improves insulin sensitivity. Treating obesity and DM often involves a combination of lifestyle changes, medication, and, in some cases, insulin therapy [16]. Bariatric surgery is the most effective treatment for severe obesity [17]. Bariatric surgery is generally recommended for people with a BMI of ≥40 or ≥35 with obesity-related health conditions, such as type 2 diabetes. Sleeve gastrectomy is widely used to treat obesity, especially in individuals with obesity-related conditions such as diabetes mellitus (DM). It is currently the most commonly performed bariatric procedure worldwide [18,19,20].

This study aimed to confirm the effect of sleeve gastrectomy according to the state of the gastric microorganisms in 40 patients with obesity, DM, and H.P infection. We compared changes in clinical features among patients with H.P and DM, with DM only, with H. pylori infection only, and without either DM or H.P infection. We also investigated the microbial changes in the stomach based on H.P infection in patients with DM, predicted their alterations after sleeve gastrectomy, and examined their relationship with clinical changes.

Materials and methods

Patients and sample preparation

In Korea, the indications for bariatric & metabolic surgery covered by the National Health Insurance are: (1) a BMI of 35 kg/m2 or higher, or (2) a BMI of 30 kg/m2 or higher accompanied by obesity-related complications such as hypertension, DM, hyperlipidemia, or non-alcoholic fatty liver disease. This study included 40 patients who met these criteria and voluntarily agreed to participate, among those who consecutively underwent laparoscopic bariatric & metabolic surgery at Seoul National University Bundang Hospital between April 2019 and August 2021. Patients with a history of H. pylori eradication therapy were excluded. The primary endpoint of this study is to determine whether H. pylori infection affects the clinical course of sleeve gastrectomy. The secondary endpoint is to investigate the differences in gastric microbiome among patients with obesity based on the presence of DM and H. pylori infection.

The collection of clinical data and patient-derived tissues was approved by the Institutional Review Board of SNUBH (IRB No: B-1812-513-302). During sleeve gastrectomy, after the stomach specimen was removed from the patient, the gastric mucosa sized 1 × 1 cm on the midbody greater curvature side was obtained via aseptic techniques. The samples were frozen at −80 °C immediately after collection.

Weight loss outcomes were reported using the percent of total weight loss (%TWL = [(initial weight) − (postoperative weight)] / ([(initial weight)] × 100) and percent excess weight loss (%EWL = [(initial weight) − (postoperative weight)] / [(initial weight) − (ideal weight]). %EWL was based on a patient ideal weight that results in a BMI of 25 kg/m2 [21]. The Hepatic Steatosis Index (HSI) was calculated using the following formula: HSI = 8 × ALT/AST ratio + BMI (+2, if DM; +2, if female) [22].

Microbial genomic DNA extraction and 16S sequencing from gastric tissues

Participants were classified into four groups: non-DM without H. pylori infection (ND), non-DM without H. pylori infection (ND-HP), DM without H. pylori infection (DM), and DM without H. pylori infection (DM-HP) groups. Fresh gastric mucosa samples were collected from each group, snap-frozen in liquid nitrogen, and stored at −20 °C. Total microbial metagenomic DNA from each sample was extracted using the QIAamp DNA microbiome kit according to the manufacturer’s instructions (Qiagen, Hilden, Germany). The quality of the extracted genomic DNA was determined using a bioanalyzer (Agilent 2100, Agilent Technologies, Inc., Santa Clara, CA, USA), and the DNA was stored at −20 °C. The Illumina platform targeted an area containing the V3–V4 hypervariable region of the bacterial 16S rRNA gene. PCR amplification of the target region was initiated immediately after mDNA extraction. The 16S V3–V4 amplicon was amplified using the KAPA HiFi Hot Start Ready Mix (2×) (Roche, Penzberg, Germany). A pair of V3–V4 target-specific universal primers recommended by Illumina were used. The primer sequences were as follows: 16S 341F forward primer: 5′-TCGTCGGCAGCGTCAGATGTGTATAAGAGACAGCCTACGGGNGGCWGCAG-3′ and 16S 806R reverse primer: 5′-GTCTCGTGGGCTCGGAGATGTGTATAAGAGACAGGACTACHVGGGTATCTAATCC-3′. After PCR amplification, all PCR products were cleaned using AMPure XP beads (Beckman Coulter, Brea, CA, USA). Additional PCR amplification was performed to add multiplexing indices and Illumina sequencing adapters using a Nextera XT Index Kit (Illumina, San Diego, CA, USA). The final PCR products were purified using AMPure XP beads. After amplicon library construction, 16S metagenome sequencing was performed using the paired-end 2 × 300 bp Illumina MiSeq protocol (Illumina MiSeq).

Bioinformatic analyses

Raw sequencing data generated using a MiSeq sequencer (Illumina) were processed using the plugins provided in the Quantitative Insights into Microbial Ecology version 2 (QIIME 2) pipeline. Quality-controlled amplicon sequences were corrected, PhiX-filtered, and dereplication was verified using the Divisive Amplicon Denoising Algorithm 2 plugin in QIIME 2 (Table S1). The bacterial amplicon sequence variants were identified and aligned, and a phylogenetic tree was generated using the align-to-tree-mafft-fasttree plugin. Various α-diversity indices (observed features, Chao1 index, Shannon’s index, Simpson’s index, and Pielou’s evenness) and β-diversity indices (Bray–Curtis and unweighted unique fraction metric [UniFrac]) were determined with a rarefied depth of 1980 reads per sample using the diversity plugin in QIIME 2. Principal coordinate analysis was used to investigate the similarities between the bacterial communities based on metadata using the Bray–Curtis and unweighted UniFrac methods. Bacterial taxonomic assignments were performed using the SILVA 138v 99% 16S ribosomal RNA database, specifically for the V3–V4 hypervariable region of the 16S sequences. All classifications were implemented using the feature classifier sklearn plugin.

Statistical analyses

Comparisons among the four groups were conducted using Fisher’s exact and Friedman tests for categorical and continuous variables, respectively. The Wilcoxon signed-rank sum test was used to compare changes in continuous variables before and after surgery.

The Kruskal–Wallis (Wilcoxon signed-rank sum) test was performed to determine significant differences in the α-diversity indices of taxonomic groups. Permutational multivariate analysis of variance was applied to the Bray–Curtis and unweighted UniFrac distance matrices using the “Adonis” function (999 permutations) of the vegan package in R statistical software to estimate significant differences in the β-diversity metrics of bacterial communities.

Correlation analyses

Correlations between clinical parameters and the microbiome were visualized using a correlation matrix prepared using Spearman’s method. Following normalization of the Kyoto Encyclopedia of Genes and Genome (KEGG) pathways identified through pathway analysis (see subsection “Metabolic pathway analysis”), correlations between clinical parameters and KEGG pathways were analyzed using Pearson’s correlation method.

Metabolic pathway analyses

The Phylogenetic Investigation of Communities by Reconstruction of Unobserved States 2 (PICRUSt2) pipeline was used to infer metagenomic functional content based on microbial community profiles obtained from representative sequences and sequence features. The predicted functional genes were categorized using KEGG ontology and pathway analysis. Linear discriminant analysis effect size (LEfSe) analysis was used to evaluate differential functional abundance based on the predicted KEGG pathways. The LEfSe method was used to identify the taxonomical biomarkers contributing to the groups with high stringency (linear discriminant analysis [LDA] score ≥3.0 and p value < 0.1). Significant biomarkers were extracted at the genus and species levels, and features marked as “unclassified” were filtered out.

Co-occurrence network analyses

Network inference was performed using the CoNet app with Cytoscape 3.4.0, using an approach similar to that of Faust et al. [23] with the following modifications: initial filtering was performed by removing operational taxonomic units (OTUs) that had a relative abundance <0.05% and occurred in <50% of the samples. The co-presence and mutual exclusion networks were analyzed using the “NetworkAnalyzer” tool of Cytoscape (https://cytoscape.org/) and exported for further analysis, with node and edge tables for the entire association network. The betweenness centrality (BNC) of a node reflects the amount of control it exerts over the interactions with other nodes in the network [24]. This measure favors nodes that join communities (dense subnetworks) rather than nodes that lie inside a community. The BNC measures how central a bacterial genus is in the network based on the number of shortest paths between other nodes that pass through it. A higher BNC value indicates a more central role in the network. The clustering coefficient (CC) of a node is the number of triangles (3-loops) that pass through this node relative to the maximum number of 3-loops that possibly pass through the node. The CC measures the degree to which a node is connected to other nodes that are connected. A high CC value indicates that a node is connected to other closely connected nodes, whereas a low CC value indicates that the node is not closely connected to other nodes in the network.

Results

Clinical results before and after sleeve gastrectomy

We collected gastric microbiome profiling and clinical follow-up data for 12 months after sleeve gastrectomy from 15 individuals without diabetes and H. pylori infection (ND), 6 individuals without diabetes and with H. pylori infection (ND-HP), 11 individuals with diabetes (DM), and 8 individuals with diabetes and H. pylori infection (DM-HP). The majority of participants were in their 30s to 50s. The ND group was significantly younger compared to other groups. Preoperative BMI and blood pressure showed no significant differences, and C-peptide levels were also similar among 4 groups. In the comparison between the 2 groups with DM, the median duration of DM was higher in the DM-Hp group, but this difference was not significant (5.0 [4.5;18.5] in the DM group vs. 9.0 [4.5;11.9] in the DM-Hp group, p = 0.868). The proportion of patients using insulin was higher in the DM group, but this difference was also not statistically significant (72.7% vs. 37.5%, p = 0.181) (Figs. 1, S1 and Tables 1, S2).

Fig. 1: Comparison of clinical changes by group before (pre-) and after (12M) sleeve gastrectomy.
figure 1

Each group (ND, ND-HP, DM, and DM-HP) was compared using Wilcoxon and t tests for several clinical parameters, including A aspartate transferase (AST), B hemoglobin A1c (HbA1c), C high-density lipoprotein (HDL), D serum uric acid, E body mass index (BMI), F blood glucose and G hepatic steatosis index (HSI). ND non-DM without H. pylori infection, ND-HP non-DM with H. pylori infection, DM diabetes mellitus, DM-HP DM with H. pylori infection.

Full size image
Table 1 Baseline characteristics for this study.
Full size table

The weight loss at 12 months postoperatively (TWL and EWL) did not show significant differences among 4 groups, but the ND group had the highest median TWL (31.4%) and the DM-Hp group had the lowest (22.7%). Similarly, the DM-Hp group had the lowest EWL (64.1%). Indicators related to blood glucose, lipids, and hepatic steatosis improved in all groups postoperatively (Tables 2 and S2). Especially, significant differences were noted from preoperative to postoperative values in the DM group. The DM group showed statistically significant reductions in fasting glucose, HbA1c, ALT, and the hepatic steatosis index at 12 months postoperatively, and a significant increase in HDL cholesterol. However, the DM-Hp group did not show significant differences in any of these indicators.

Table 2 Postoperative changes of body weight and laboratory findings.
Full size table

These results suggest that clinical improvement after sleeve gastrectomy is effective in patients with obesity and diabetes, but may be hindered in patients with H. pylori infection.

Composition and diversity of gastric microbiome in patients before sleeve gastrectomy

The relative abundance of the gastric microbiome was compared among the four groups. Firmicutes accounted for 35.6%, 17.5%, 33%, and 18% in the ND, ND-HP, DM, and DM-HP groups, respectively. Actinobacteriota were dominant in the H. pylori-negative groups, accounting for 44.2% and 46.9% in the ND and DM groups, respectively. Actinobacteriota represented 25.5% and 28.7% in the ND-HP and DM-HP groups, respectively. Notably, Campilobacterota were dominant in the ND-HP and DM-HP groups, accounting for 46.4% and 34.8%, respectively (Fig. S3). Cutibacterium was the most dominant genus in the ND and DM groups, accounting for 38.11% and 42.8%, respectively. As expected, in the ND-HP and DM-HP groups, H. pylori was the most dominant species, accounting for 45.97% and 33.96%, respectively. These results indicated that Firmicutes were prevalent in all groups; however, their relative abundance varied. Actinobacteriota were more common in the H. pylori-negative groups, whereas Campilobacterota were more common in the H. pylori-positive groups. The genus Cutibacterium was predominant in the H. pylori-negative group, whereas Helicobacter was predominant in the H. pylori-positive group (Fig. 2A).

Fig. 2: Relative abundance and diversity of gastric microbiome in the ND, ND-HP, DM, and DM-HP groups.
figure 2

A Relative abundance in the ND, ND-HP, DM, and DM-HP groups. B Observed operational taxonomic units (OTUs) (number of distinct features): calculates the number of distinct OTUs. Chao1 index: estimates diversity from abundant data. Pielou’s evenness: measures relative evenness of species richness. Simpson’s index: measures the relative abundance of the different species making up the sample richness. Shannon’s index: accounts for the abundance and evenness of the taxa. C β-diversity plot between the four groups (upper panel). The table shows significant differences through the permutational multivariate analysis of variance test (bottom panel).

Full size image

The α-diversity decreased in the order of ND > DM > ND-HP > DM-HP (Fig. 2B and Tables S3 and S4). The α-diversity of the DM-HP group was significantly reduced compared to that of the ND group, based on Shannon’s index, Simpson’s index, and Pielou’s evenness (p < 0.01). Moreover, the ND-HP group showed a more significant reduction in diversity than the ND group (p < 0.01). Pielou’s evenness was significantly lower in the ND-HP group than in the DM group (p = 0.02). Significant group differences were observed in α- and β-diversity when comparing the ND and DM-HP and DM and ND-HP groups (p ≤ 0.05) based on H. pylori infection (Fig. 2C and Table S5). Therefore, β-diversity differs depending on the presence of diabetes or H. pylori infection.

Alteration of gastric microbiota in patients with obesity based on the presence of DM and H. pylori

Figure 3 (Fig. S2 and Table S6) shows the results of the network analysis, involving BNC and CC. The plot indicates that in the ND group, Lachnospiraceae-NK4A136-group (0.056) and Lactobacillus (0.051) had high BNC values, whereas Porphyromonas (1), Prevotella (0.86), and Lachnoclostridium (0.809) had high CC values (Fig. 3A). In the ND-HP group, Helicobacter (0.011) and Streptococcus (0.011) showed high BNC values. Staphylococcus (0.90) had a high CC value, whereas Lactobacillus (0.80) had a low CC value (Fig. 3B). In the DM group, Cutibacterium (0.052) had the highest BNC value, and Lachnoanaerobaculum (0.73) had the highest CC value (Fig. 3C). In the DM-HP group, Helicobacter (0.051) had the highest BNC value, and Romboutsia (0.001) had the lowest BNC value. Acinetobacter (0.73) had the highest CC value, and Leptotrichia (0.461) had the lowest (Fig. 3D). In the ND-HP group, the CC value was much higher than 0.40 (horizontal red dotted line: arbitrary boundary line) compared to that of other groups. Therefore, the connectivity between nodes is higher in the ND-HP group than in other groups. These results suggest that the presence of Helicobacter is associated with a higher CC and greater co-occurrence among bacterial genera in gastric microbiota.

Fig. 3: Co-occurrence analysis between the gastric microbiome before sleeve gastrectomy in each group.
figure 3

The y-axis indicates betweenness centrality (BNC), and the x-axis indicates the clustering coefficient (CC). The genera located at the bottom right of the plots have the lowest rates of co-occurrence with other genera (low CC) and play a central role in the topology of the network (high BNC). The red dotted line is an arbitrary line for comparing the positions above the microbiota on the coordinates of the BNC and CC. The red dotted line on the vertical axis represents a random cutoff based on a BNC of 0.10, and the red dotted line on the horizontal axis represents a random cutoff based on a CC of 0.4. A ND is non-DM without H. pylori infection group, B ND-Hp is non-DM with H. pylori infection group, C DM is diabetes mellitus group, (C) DM-Hp is DM with Indicates H. pylori infection group..

Full size image

We identified pathways with an LDA score of >3.0 based on the microbiome in each group (Fig. 4A and Table S7). In the ND-HP group, several pathways related to cell motility (LDA = 3.8), bacterial chemotaxis (LDA = 3.76), lipopolysaccharide biosynthesis (LDA = 3.79), and biotin metabolism (LDA = 3.54) showed high LDA scores, indicating that these pathways were more abundant in this group than in the other groups. In the ND group, all features corresponded to pathways in the metabolism category, with a particularly high LDA score for the xenobiotic biodegradation and metabolism pathways. In the DM-HP group, the taurine and hypotaurine metabolism pathways (LDA = 3.2) had high LDA scores, whereas most of the features in the DM group corresponded to pathways in the metabolism category.

Fig. 4: Comparison of functions using PICRUSt2 from the microbiome in each group.
figure 4

A Results of LEfSe analysis: the y-axis represents a gastric microbial-related pathway, and the x-axis indicates the linear discriminant analysis (LDA) score. B Heatmap of the relative abundance of the microbiome according to features based on the LDA in each group.

Full size image

The microbiome abundance patterns in the ND-HP and DM-HP and ND and DM pairs were similar based on the heatmap results (Fig. 4B and Table S8), indicating that the pathways corresponding to these pairs were associated with certain characteristics depending on the presence or absence of Helicobacter. Therefore, when considering the characteristics of the strains representing each group, these pathways can be used to differentiate between the groups based on the presence or absence of Helicobacter.

Correlation of clinical parameters and the gastric microbiome before sleeve gastrectomy

We performed a correlation analysis between the gastric microbiome and clinical indicators to confirm the preoperative gastric environment and clinical relationships in each group. Figure 5 shows the correlation analysis between clinical parameters and the predominant top 20 microbial communities in the gastric tissue before sleeve gastrectomy. Notably, the ND-HP group showed a higher correlation with the clinical parameters of the microbiome than the other groups (Fig. 5 and Table S9). The ND group showed a lower correlation between microbes and clinical parameters than the other groups. However, Streptococcus showed a correlation with HbA1c (R = 0.57) and low-density lipoprotein (LDL) (R = 0.584). Bosea was positively correlated with TG (R = 0.547), HDL (R = 0.522), and blood glucose levels (R = 0.587). Fusobacterium (R = 0.535), Tannerella (R = 0.653), and Bosea (R = 0.587) showed strong positive correlations with blood glucose levels, whereas Neisseria (R = −0.635) and Actinomyces (R = −0.571) showed a strong negative correlation. In addition, Bacteria; unclassification positively correlated with HIS (R = 0.621).

Fig. 5: Correlation between the top 20 microbiome and clinical parameters in gastric tissues before sleeve gastrectomy for each group.
figure 5

Correlation analysis results using Spearman’s analysis. Red indicates a positive correlation, and blue indicates a negative correlation. Asterisks (*) represent the p value of the comparative statistical test for each bacterial genus (*p < 0.05, **p < 0.01, ***p < 0.001).

Full size image

In the ND-HP group, LDL showed an overall positive correlation, and higher correlations were observed for Staphylococcus (R = 0.828) and Veillonella (R = 0.828). In addition, most clinical parameters showed negative correlations in Rothia, whereas TG showed a relatively strong negative correlation (R = −0.942). Bifidobacterium showed relatively strong positive correlations with ALT (R = 0.811), TG (R = 0.811), and HSI (R = 0.840) levels. The ND-HP group showed negative correlations with blood glucose, insulin, and homeostatic model assessment of insulin resistance (HOMA-IR), except for Helicobacter and Rothia, which were among the top 20 microbiome. In particular, insulin levels had a relatively high overall negative correlation.

In the DM group, negative correlations were observed between Cutibacterium and ALT (R = −0.636) and aspartate transferase (AST) (R = −0.665). Staphylococcus was negatively correlated with HDL levels (R = −0.675). Tannerella and Bacteroides showed a strong positive correlation with ALT (R = 0.769 and R = 0.837, respectively) and AST (R = 0.739 and R = 0.811, respectively). Actinomyces showed strong positive correlations with HbA1c (R = 0.668) and HDL levels (R = 0.539). Bacteria; unclassification showed positive correlations with HDL (R = 0.611).

The DM-HP group showed a negative correlation with Cutibacterium and HDL (R = −0.802), and Helicobacter showed a positive and negative correlation with HDL (R = 0.718) and TG (R = 0.833), respectively. Cutibacterium and Helicobacter showed an inverse correlation with TG and HDL levels.

Cutibacterium and f_Lachnospiraceae; unclassification showed a strong positive correlation with insulin (R = 0.785) and HOMA-IR (R = 0.880). d__Bacteria;_ unclassification showed a negative correlation with HbA1c (R = −0.714) levels. Staphylococcus showed a negative correlation with BMI (R = −0.761). Additionally, Lactobacillus and Rhotia positively correlated with AST (R = 0.761) and ALT (R = 0.913) levels.

Discussion

This study investigated the differences in clinical changes before and after bariatric surgery in the microbiome environment in the gastric mucosa of patients with obesity, DM, and H. pylori infection. Notably, except for gastric studies on H. pylori, studies on commensal microbiome and their roles in DM and obesity are limited. Gastric bacterial communities other than H. pylori were first profiled in the early 2000s [25]. The most abundant 6 bacteria isolated from healthy human stomachs were Veillonella, Lactobacillus, Clostridium [26], Cutibacterium, Streptococcus, and Staphylococcus [27]. Some studies profiled all taxa at different taxonomic levels, whereas others focused only on microbiome capable of distinguishing certain gastric diseases from controls with significant differences in the definition of “healthy control.” Nevertheless, five major phyla—Proteobacteria, Firmicutes, Bacteroidetes, Actinobacteria, and Fusobacteria—have consistently been observed in healthy populations [28,29,30,31,32]. Especially at the genus level, Streptococcus, Prevotella, Veillonella, and Rothia were most frequently observed. Similar communities were detected in the present study. Helicobacter was dominant among the patients infected with H. pylori (the ND-HP and DM-HP groups), followed by Cutibacterium and Streptococcus. These results suggest that these bacteria interact with each other to survive in harsh gastrointestinal environments. We confirmed their interactions in the H. pylori-infected group through co-occurrence analysis, and their functions were predicted using PICRUSt2. The Cutibacterium genus was the most dominant, followed by Streptococcus and Staphylococcus, in the groups with obesity and DM (the ND and DM groups).

Notably, in the functional prediction analysis, pathways such as flagellar assembly, lipopolysaccharide biosynthesis, bacterial chemotaxis, and biotin metabolism were identified in the ND-HP group. Flagella are protein structures that many bacteria use for movement, and H. pylori and other bacterial species are flagellated pathogen. Flagella are important in H. pylori pathogenesis and allow the bacteria to move through the thick mucus layer of the stomach lining and reach the underlying epithelial cells, where they can establish an infection. In addition, flagella induce inflammation and immune responses that contribute to gastric disease development, such as gastritis, peptic ulcers, and gastric cancer [33].

Lipopolysaccharide (LPS) is a large molecule in the outer membrane of many gram-negative bacteria. LPS maintains the structural integrity of bacterial cells and the interaction of bacteria with their environment. Bacterial chemotaxis is the movement of bacteria toward or away from chemical cues in the environment. This process is crucial for bacterial survival and colonization in various environments. Biotin is an important vitamin for carbohydrate, fat, and amino acid metabolism by bacteria.

Co-occurrence analysis of the ND-HP and DM-HP groups showed that the CC of the microbial groups was higher than that of the ND. This finding is possibly due to the formation of stronger interactions between different microbial groups owing to the disturbance by H. pylori, which dominates the gastric environment, or the effect of their virulence factors. H. pylori causes an imbalance in the gut microbiome by altering the population of other microbiomes residing in the gut [34,35,36,37]. In particular, H. pylori can release virulence factors impacting the intestinal epithelium, causing inflammation and further perturbation of the gut microbiome [38].

Clinical parameters improved in all groups following sleeve gastrectomy. However, individuals harboring H. pylori did not show significant differences in clinical improvement. This suggests that the efficacy of surgery may be hindered depending on the presence of H. pylori infection in patients with diabetes and obesity. Although the relationship between obesity and H. pylori infection remains controversial, the eradication rate of H. pylori infection is particularly low in patients with obesity compared to that in normal individuals. Two possible reasons for this have been suggested. First, physiological changes that occur in obesity, such as delayed gastric emptying, can decrease the rate of drug absorption, regardless of the drug properties [39]. Second, the distribution volume of drugs may be altered in patients with obesity because the increase in adipose tissue may affect drugs with lipophilic properties [40, 41]. However, most studies on bariatric surgery and H. pylori infection focused on the effect of H. pylori on marginal ulcers after bariatric surgery. Therefore, studies are needed on H. pylori infection, changes in gastric microbes, and clinical changes caused by obesity surgery in people with obesity.

This study had some limitations. First, our study lacked sufficient number of clinical groups and simultaneous monitoring of changes in the stool or gastric microbiome before and after bariatric surgery. Second, owing to the absence of normal controls, the standard clinical data can only be compared with that before surgery. Moreover, experimental verification, such as metabolomic analysis, is needed to identify the causative factors that affect the clinical effects. Nonetheless, this study is the first to predict the prognosis of bariatric surgery through the state of the gastrointestinal microbes of in people with obesity and microbes in an environment dominated by Helicobacter. Therefore, these results can be used to develop clinical guidelines for obesity surgery and prescriptions to reduce obesity.

In conclusion, sleeve gastrectomy had an improvement effect in patients with obesity and diabetes. However, in people with obesity coexisting with H. pylori, it may hinder clinical improvement along with changes in the composition of gastric microbiome. Future studies will collect gastrointestinal microbiome and clinical data from healthy participants with and without H. pylori infection to investigate the effective and active improvement of obesity compared to the gastric microbiome after bariatric surgery in in people with obesity.