Gut microbiota-based prediction for the transition from normal glucose tolerance (NGT) to impaired glucose tolerance (IGT) in a remote island cohort study.

AIM: The present cohort study explored whether specific gut microbiota (GM) profile would predict the development of impaired glucose tolerance (IGT) in individuals with normal glucose tolerance (NGT). METHODS: A total of 114 study subjects with NGT in Kumejima island, Japan participated in the present study and underwent 75 g oral glucose tolerance tests at baseline and one year later. We compared the profile of GM at baseline between individuals who consistently maintained NGT (NRN, n = 108) and those who transitioned from NGT to IGT (NTI, n = 6). RESULTS: Within-individual bacterial richness and evenness as well as inter-individual bacterial composition showed no significant differences between NRN and NTI. Of note, however, partial least squares discriminant analyses revealed distinct compositions of GM between groups, with no overlap in their 95 % confidence interval ellipses. Multi-factor analyses at the genus level demonstrated that the proportions of CF231, Corynebacterium, Succinivibrio, and Geobacillus were significantly elevated in NTI compared to NRN (p < 0.005, FDR < 0.1, respectively) after adjusting for age, sex, HbA1c level, and BMI. CONCLUSIONS: Our data suggest that increased proportion of specific GM is linked to the future deterioration of glucose tolerance, thereby serving as a promising predictive marker for type 2 diabetes mellitus.

Perspectives on genetic studies of type 2 diabetes from the genome-wide association studies era to precision medicine.

Genome-wide association studies (GWAS) have facilitated a substantial and rapid increase in the number of confirmed genetic susceptibility variants for complex diseases. Approximately 700 variants predisposing individuals to the risk for type 2 diabetes have been identified through GWAS until 2023. From 2018 to 2022, hundreds of type 2 diabetes susceptibility loci with smaller effect sizes were identified through large-scale GWAS with sample sizes of 200,000 to >1 million. The clinical translation of genetic information for type 2 diabetes includes the development of novel therapeutics and risk predictions. Although drug discovery based on loci identified in GWAS remains challenging owing to the difficulty of functional annotation, global efforts have been made to identify novel biological mechanisms and therapeutic targets by applying multi-omics approaches or searching for disease-associated coding variants in isolated founder populations. Polygenic risk scores (PRSs), comprising up to millions of associated variants, can identify individuals with higher disease risk than those in the general population. In populations of European descent, PRSs constructed from base GWAS data with a sample size of approximately 450,000 have predicted the onset of diseases well. However, European GWAS-derived PRSs have limited predictive performance in non-European populations. The predictive accuracy of a PRS largely depends on the sample size of the base GWAS data. The results of GWAS meta-analyses for multi-ethnic groups as base GWAS data and cross-population polygenic prediction methodology have been applied to establish a universal PRS applicable to small isolated ethnic populations.