Feed

Achieving rapid glycemic stability with intensive insulin therapy in hospitalized type 2 diabetes patients remains unpredictable. We hypothesized that underlying insulin resistance is a key determinant of this response. This exploratory study aimed to test this hypothesis and develop a predictive nomogram.

We retrospectively analyzed 393 hospitalized patients initiating insulin pump therapy with continuous glucose monitoring. Stability was defined as Time in Range >70% for 24 consecutive hours. Collected data included body mass index, hemoglobin A1c (HbA1c), fasting glucose, fasting insulin, lipids, and calculated insulin resistance indices. Cox regression identified predictors of time to stability.

Analysis confirmed insulin resistance's significant role. Multivariate Cox regression identified older age, higher fasting insulin, elevated HbA1c, and a higher triglyceride-glucose (TyG) index as independent risk factors for prolonged time to stability (all P<0.05), with the TyG index interpreted as a marker of lipotoxic metabolic inflexibility rather than merely a surrogate of insulin resistance. A nomogram integrating these factors demonstrated good predictive accuracy, with a C-index of 0.801 (95% CI: 0.77-0.83). Calibration and decision curve analysis supported its clinical utility.

Insulin resistance significantly influences time to glycemic stabilization. The developed nomogram, incorporating the TyG index, fasting insulin, age, and HbA1c, provides a practical tool for early risk stratification, potentially guiding more personalized inpatient diabetes management.