Large-Scale Proteomics Uncovers Pre-Disease Inflammation-Lipid Subtypes to Refine Risk Stratification and Prediction...

Area of Science:

• Endocrinology and Metabolism
• Proteomics
• Biostatistics

Background:

• Type 2 diabetes (T2D) presents significant heterogeneity before clinical onset, complicating early prevention efforts.
• Identifying distinct pre-disease subgroups is crucial for developing targeted interventions and improving patient outcomes.

Purpose of the Study:

• To identify pre-disease subgroups in individuals at risk for T2D based on proteomic profiles.
• To develop subgroup-specific prediction models for early T2D detection and precision prevention.

Main Methods:

• Analysis of proteomic data from 41,030 UK Biobank participants without diabetes.
• Identification of protein markers associated with incident T2D using Cox regression and LASSO.
• Clustering participants into risk subgroups using finite Gaussian mixture models.
• Development of subgroup-specific prediction models via Cox regression.

Main Results:

• Three pre-disease subgroups were identified: metabolically healthy group (MHG), mild inflammation group (MIG), and dyslipidemia with inflammation group (DLIG).
• DLIG exhibited the highest T2D risk (HR=2.58), followed by MIG (HR=1.71) compared to MHG.
• Proteomic models demonstrated superior predictive performance (AUC 0.820-0.889) over traditional clinical models (AUC 0.709-0.784).

Conclusions:

• Proteomic profiling effectively identified three distinct pre-disease T2D subtypes characterized by inflammation and lipid metabolism.
• These subtypes offer improved risk stratification and enable precision prevention strategies for T2D.
• Subgroup-specific prediction models enhance early detection and personalized intervention for T2D.