Abstract
Sleep-related alterations in glucose metabolism and autonomic regulation play a critical role in the early development of type 2 diabetes mellitus (T2DM), especially among cancer survivors. Motivated by this interplay, we develop a physiologically informed mathematical model linking nocturnal glucose levels with resting heart rate dynamics. The model comprises of a coupled system of ordinary differential equations, and is fitted to wearable sensor data using particle swarm optimization. Sensitivity analysis is carried out to obtain the key parameters driving metabolic dysregulation, enabling the formulation of a personalized risk score. Successful validation using synthetic patient data and cancer survivor patient data suggests that this framework provides a non-invasive, interpretable tool for early T2DM risk stratification based on nocturnal physiological patterns.