Construction and Implications of Nomogram for Predicting Sustained Glycemic Remission After Short-Term Intensive Insulin Therapy in Newly Diagnosed Type 2 Diabetes

Abstract

Background

Early short-term intensive insulin therapy may induce sustained glycemic remission in type 2 diabetes. However, patients' responses vary greatly. Exploring key factors to improve glycemic control and predict the probability of remission precisely is meaningful for future treatment.

Methods

Patients with newly diagnosed type 2 diabetes receiving 2–3 weeks of intensive insulin therapy were followed up for at least 1 year in three randomized clinical trials. Data of theirs were extracted with the same inclusion criteria and divided into training and validation sets. A nomogram was constructed in the training set and tested in the internal validation set.

Results

Among 302 patients with transient intensive insulin therapy, 162 (53.64%) patients achieved the 1-year glycemic remission. Severe hyperglycemia at baseline did not impede future remission, as the remission rate was 60.70% in those with HbA1c greater than 13%. Three parameters were identified as predictive factors in the nomogram: fasting glucose after short-term insulin treatment, mean glucose during insulin therapy, and postprandial glucose/fasting glucose ratio at baseline. The odds ratios were 0.06 (95% CI, 0.02–0.25), 0.41 (0.18–0.93) and 6.55 (1.39–30.89), respectively. Incorporating these factors, the nomogram achieved an accuracy of 81.50%.

Conclusion

Short-term intensive insulin therapy assists patients with newly diagnosed type 2 diabetes and significant hyperglycemia to achieve glycemic remission. A rigorous control of glucose during insulin treatment favors future remission. We construct a nomogram to predict sustained glycemic remission, which may help determine whether subsequent medication is needed and thus reduce both overtreatment and therapeutic inertia.