A physical activity-intensity driven glycemic model for type 1 diabetes.

Area of Science:

• Endocrinology and Metabolism
• Computational Physiology
• Diabetes Technology

Background:

• Glycemic response to physical activity in type 1 diabetes (T1D) is complex, influenced by activity intensity, duration, insulin levels, and fitness.
• Accurate modeling of these factors is crucial for managing T1D during exercise and developing effective automated insulin delivery systems.

Purpose of the Study:

• To develop and validate a physiologically plausible model for predicting glucose excursions during physical activity in individuals with T1D.
• To improve the simulation accuracy of glycemic responses by incorporating key metabolic factors beyond simple glucose utilization.

Main Methods:

• Proposed several physiological models incorporating terms for endogenous glucose production (EGP), glucose utilization, and glucose transfer.
• Evaluated model fits using clinical data from individuals with T1D during physical activity.
• Assessed model performance based on reduction in model error, specifically mean absolute percentage error (MAPE).

Main Results:

• The proposed model, including terms for EGP, glucose transfer, and insulin-independent glucose utilization, significantly improved simulation accuracy.
• Achieved a greater reduction in model error compared to traditional models (MAPE: 16.11 ± 4.82% vs. 19.49 ± 5.87%, p=0.002).
• Demonstrated accurate representation of the interplay between plasma insulin, activity intensity, glucose production, and utilization.

Conclusions:

• A physiologically informed model incorporating multiple glucose metabolism contributors outperforms models relying solely on glucose utilization.
• The model accurately predicts glucose responses (increasing, decreasing, or stable) under various physical activity conditions.
• Enables in silico evaluation of automated insulin dosing algorithms for mitigating exercise-induced glycemic excursions in T1D.