The Minimally-Invasive Oral Minimal Model: Inpatient-to-outpatient Transition of the Oral Glucose Minimal Model

Type 1 diabetes (T1D) is a chronic condition characterized by the absolute deficiency of insulin secretion that leads to chronic elevated blood glucose (BG) levels and metabolic disorders of carbohydrates, fats, and proteins. If not properly managed, uncontrolled BG levels damage the vascular system causing the onset of more severe clinical complications, including cardiovascular disease, retinopathy, and nephropathy. For this reason, individuals with T1D are required to keep BG concentration within a tight safe range (approximately between 70 and 180 mg/dL) throughout the day. For this purpose, they undergo frequent administration of exogenous insulin analogs via insulin pens and pumps. However, optimal dosing of this hormone is notoriously a challenging task, due to the presence of several disturbing factors affecting BG levels, such as physical exercise, psychological stress, and food intake. It has been demonstrated that the presence of fat and protein in a meal strongly affects gastric retention (GR, i.e., the fraction of food in the stomach after the meal), glucose rate of appearance (Ra, i.e., the velocity at which glucose is absorbed in the blood), as well as insulin sensitivity (SI, i.e., the effectiveness of insulin in lowering BG levels). In turn, all these factors impact the postprandial glucose excursion and can cause early hypo- (i.e., BG concentration < 70 mg/dL) and late prolonged hyper-glycemia (i.e., BG concentration > 180 mg/dL) if not properly considered in the calculation of the prandial insulin bolus. Despite this strong evidence, current T1D therapies still do not take into account meal components different from carbohydrates, making optimal postprandial glycemic control a challenging task in the presence of different macronutrients in the meal. This is mainly due to the lack of a method quantifying the effect of different macronutrients on postprandial glucose dynamics at individual level, possibly usable in real-life conditions. To bridge that gap, in this thesis project, we developed a tool, the Minimally-Invasive Oral Minimal Model (MI-OMM), aiming at quantifying physiological parameters like GR, Ra, and SI in individuals with T1D from minimally-invasive devices and thus usable in free-living conditions. This was achieved starting from the well-known (in the field of diabetes) Oral glucose Minimal Model (OMM) proposed by Dalla Man and colleagues in 2002 to estimate Ra and SI in individuals with T1D undergoing a glucose challenge and a subsequent blood sampling for measuring plasma glucose and insulin concentrations. Into this model, we subsequently integrated additional mathematical models to translate its domain of applicability from inpatient to outpatient settings. The resulting MI-OMM was shown to satisfactorily describe the data coming from individuals wearing a continuous glucose monitoring sensor and insulin pump device, providing also precise and physiologically meaningful parameter estimates. In addition, it provided an accurate description of GR, Ra, and SI that compared against the reference ones well. Thereafter, we applied the MI-OMM to an independent dataset collected from individuals with T1D in free-living conditions where meal composition was carefully recorded. Thus, this scenario allowed the investigation of possible fat and protein effects on the estimated GR, Ra, and SI. This was done by a priori classifying each meal into four classes based on the content of fat and protein and subsequently comparing the results of the high- and low-content classes in terms of GR, Ra, and SI. The results of this analysis were consistent with the literature, showing that a high presence of fat and protein in the meal can significantly slow both gastric emptying and glucose intestinal absorption, as well as increase insulin resistance.