Insulin Replacement Across the Menstrual Cycle in Women with Type 1 Diabetes: An In-Silico Assessment of the Need for Ad-Hoc Technology.

Background Women with type 1 diabetes (T1D) of fertile age may experience fluctuations in insulin needs across the menstrual cycle. When present, these fluctuations complicate glucose management and oftentimes worsen glycemic control. In this work, an in-silico analysis was conducted to assess whether current technology is sufficient to handle changes in insulin needs due to the menstrual cycle in women with T1D. Methods Euglycemic clamp studies were performed in 16 women with T1D in the follicular phase (FP) and luteal phase (LP) of the menstrual cycle. Inter-phase insulin sensitivity (IS) variability observed in the data was modeled and introduced in the University of Virginia/Padova T1D Simulator. Open-loop (OL) and closed-loop (CL) insulin delivery were tested in two in-silico studies, without (nominal study) and with (informed study) a-priori knowledge on cycle-related IS variability informing insulin therapy. Glycemic metrics were computed on the obtained glucose traces. Results In the pool of studied women, glucose infusion rate area under the curve significantly decreased from FP to LP (P = 0.0107), indicating an average decrease of IS in LP. When introduced in the simulator, this pattern led to increased time spent >180 mg/dL and >250 mg/dL during LP vs. FP in the nominal studies, irrespective of the insulin delivery strategy. In the informed studies, glycemic metrics stabilized across the cycle. Conclusion This work suggests that current insulin delivery technology may benefit from informing the dosing algorithm with knowledge on menstrual cycle-related IS changes. Clinical validation of these results is warranted.