Abstract
BACKGROUND: Type 2 diabetes (T2D) presents clinical challenges due to its difficult early diagnosis and treatment insensitivity. Further, the relationship between gut microbiota and serum composition in T2D has not been fully characterized. This study aimed to determine the relationship between gut microbiome and serum metabolome in patients with T2D. METHODS: We collected fecal and serum samples from 30 T2D patients and healthy controls (HCs). The fecal microbiome composition was analyzed using 16S rRNA sequencing, and serum metabolites were detected by UHPLC-MS/MS. Alpha and beta diversity indices (Chao1, Shannon, PCoA, etc.) were calculated to assess microbial diversity and community structure. Differential metabolites were integrated to identify potential biomarkers, and random forest modeling was used for predictive analysis to investigate and validate the importance of specific gut microbial genera. RESULTS: The feces and blood of T2D patients demonstrated different characteristics of 20 differential microbiomes in the gut and 30 metabolite in the blood from HCs. Further, a significant correlation was observed between the gut microbiota and serum metabolomic profiles, reflecting the influence of the microbiota on metabolic activity. In addition, the states of T2D and HC groups were clearly distinguishable based on differences in gut microbes and metabolites, with the random forest model achieving excellent diagnostic performance (AUC values of 0.9764 and 0.9823, respectively). CONCLUSIONS: Our study provides a comprehensive profile of changes in the microbiome and serum metabolomics, indicating their potential application as biomarkers for future diagnosis and treatment in T2D.