Integrated biomarker profiling of the metabolome associated with type 2 diabetes mellitus among Tibetan in China

Abstract

Abstract Introduction Metabolomic signatures of type 2 diabetes mellitus (T2DM) in Tibetan Chinese population, a group with high diabetes burden, remain largely unclear. Identifying the serum metabolite profile of Tibetan T2DM (T-T2DM) individuals may provide novel insights into early T2DM diagnosis and intervention. Methods Hence, we conducted untargeted metabolomics analysis of plasma samples from a retrospective cohort study with 100 healthy controls and 100 T-T2DM patients by using liquid chromatography–mass spectrometry. Results The T-T2DM group had significant metabolic alterations that are distinct from known diabetes risk indicators, such as body mass index, fasting plasma glucose, and glycosylated hemoglobin levels. The optimal metabolite panels for predicting T-T2DM were selected using a tenfold cross-validation random forest classification model. Compared with the clinical features, the metabolite prediction model provided a better predictive value. We also analyzed the correlation of metabolites with clinical indices and found 10 metabolites that were independently predictive of T-T2DM. Conclusion By using the metabolites identified in this study, we may provide stable and accurate biomarkers for early T-T2DM warning and diagnosis. Our study also provides a rich and open-access data resource for optimizing T-T2DM management.

Keywords

• Tibetan
• Type 2 diabetes mellitus
• Serum metabolomics
• Machine learning
• Biomarker