Enriched Metabolic Phenotyping Refines Phenotypic Resolution of Type 2 Diabetes

Area of Science:

• Endocrinology and Metabolism
• Computational Biology
• Precision Medicine

Background:

• Type 2 diabetes (T2D) exhibits metabolic heterogeneity, with existing clustering based on limited clinical variables identifying major subgroups like severe insulin-deficient diabetes (SIDD) and severe insulin-resistant diabetes (SIRD).
• The potential for deeper metabolic phenotyping to refine these classifications and reveal novel T2D subgroups remains underexplored.

Purpose of the Study:

• To investigate whether comprehensive metabolic phenotyping can enhance the resolution of T2D subgroups beyond established low-dimensional frameworks.
• To identify novel, physiologically coherent T2D phenotypes through exploratory cluster analysis.

Main Methods:

• An exploratory cross-sectional cluster analysis was conducted on 94 adults with newly diagnosed T2D using deep metabolic phenotyping data.
• Principal Component Analysis (PCA) informed the selection of 13 key variables across six physiological domains.
• K-means clustering was applied to standardized variables, with robustness assessed using various variable panels and alternative clustering methods.

Main Results:

• Five distinct phenotypic clusters emerged, including one with preserved insulin sensitivity and compensatory hyperinsulinaemia (IS-HI).
• Two clusters corresponded to canonical SIDD-like and SIRD-like phenotypes.
• Two novel clusters were identified: a myogenic-anaemia (MA) phenotype and a hepato-lipotoxic (HL) phenotype, highlighting distinct pathophysiological patterns.

Conclusions:

• Increased phenotypic depth in metabolic profiling can potentially improve mechanistic understanding of T2D heterogeneity beyond routine clinical data.
• These findings are hypothesis-generating and require external validation and longitudinal outcome data before informing clinical decision-making or treatment strategies.