Glucose Controlin Type 2 Diabetes: In Silico Study of Sub-Optimal Adherence to Insulin Therapy

Abstract

Abstraet- Subjects with type 2 diabetes (T2D) require multiple daily injections (MDI) of insulin analogues to cover basal and prandial insulin needs for optimal blood glucose (BG) control. Intentionally or non-intentionally failing in treatment compliance has proved to degrade BG control in type 1 diabetes, while the effect of poor adherence to insulin therapy in T2D is understudied. Here, we aim at in silico quantifying the detrimental effects on BG control of delaying or skipping insulin bolus in T2D subjects. We employed the recently developed Padova T2D Simulator (T2DS), an accurate nonlinear computer simulator of early-stage T2D and, thanks to a newly developed cloning method, we extended it to advanced-stage insulin-dependent T2D subje cts. One hundred in silico advanced-stage T2D were generated and models of basai-bolus insulins were incorporated into the T2DS to enable MDI therapy. MDI therapy parameters, i.e., basal and bolus insulin doses and injection times, were optimized for each subject by applying titration algorithms that iteratively update basal/bolus insulin amount based on BG deviation from prefixed target ranges. Finally, we assessed the impact of sub-optimal adherence to insulin therapy by running two 6-month in silico trials, with subjects receiving 3 meal/day under MDI. In the first trial, optimal basal and prandial insulin bolus at each meal were administered to each subject; in the other case, subjects received optimal basal insulin and randomly delayed or skipped the prandial insulin in 3 lunches during workingdays and 1 dinner during weekends. Results showed that, even in T2D, sub-optimal adherence to MDI therapy worsened postprandial BG control, leading to a significant increase of hyperglycemia, while apparently not increasing the risk of hypoglycemia.