Prediction Model of Cardiovascular Risk Factors through Modified Fasting Therapy in 3449 Patients with Hyperglycemia

Abstract

Abstract Objective: This study aimed to develop a matrix prediction model based on weight loss and improvement in CVD risk factors to evaluate the cardiovascular benefits of modifiedfasting therapyin overweight/obese patients with hyperglycemia. Methods: CVD-related clinical predictors were extracted from a group of 3449 hospitalizedindividuals after modifiedfasting therapy using logistic regression. Matrix prediction models were formulated, and a corresponding scoring system was developed in a separate cohort. A separate validation was conducted on a separate cohort of 715 inpatients. Results: Three Cox prediction models based on the improvement of CVD risk factors associated with weight loss were assessed. The AUC of model 3 was higher than that of prediction model 1 and model 2 (AUC = 0.91 > 0.73 > 0.79). Two matrix prediction models were employed to assess the sensitivity of predicting weight loss outcomes. Effective weight loss indicators included diastolic blood pressure (DBP) > 80 mmHg, fasting C-peptide (FCP) ≥ 260 pmol/L, and total cholesterol (TC) ≥ 5.2 mmol/L. Conclusions: The prediction model showed that the weight loss of fasting therapy had a higher clinical benefit on CVD risk factors in overweight/obese patients with hyperglycemia by DBP > 80 mmHg, FCP ≥ 260 pmol/L, and TC ≥ 5.2 mmol/L.