Novel Detection and Progression Markers for Diabetes Based on Continuous Glucose Monitoring Data Dynamics

Area of Science:

• Diabetes research
• Biomedical engineering
• Data analysis in healthcare

Background:

• Static continuous glucose monitoring (CGM) measures do not fully capture blood glucose dynamics.
• Subtle declines in glycemic control in early type 1 diabetes are often missed by traditional metrics.

Purpose of the Study:

• To develop a diabetes diagnostic tool utilizing CGM dynamic markers.
• To assess CGM entropy rate (ER) and Poincaré plot (PP) ellipse area (S) for diabetes classification and risk assessment.

Main Methods:

• Calculated ER and S from 5754 daily CGM profiles of individuals with type 1 diabetes, type 2 diabetes, or healthy controls.
• Determined optimal cut-points using the Youden index for health vs. diabetes, type 1 vs. type 2 diabetes, and type 1 immunological risk.
• Evaluated marker discriminative power using area under the receiver operating characteristics curves (AUC).

Main Results:

• ER and S demonstrated high discrimination for health vs. diabetes (AUCs 0.98 and 0.99).
• Markers effectively differentiated type 1 from type 2 diabetes (AUCs 0.81 and 0.76).
• The markers showed potential in identifying low vs. high type 1 immunological risk (AUCs 0.72 and 0.66).

Conclusions:

• CGM dynamic markers (ER and S) show significant potential as diagnostic tools.
• These markers can serve as an alternative to fasting plasma glucose or glucose tolerance testing.
• CGM dynamics analysis aids in identifying individuals at higher immunological risk for type 1 diabetes progression.