Predictive model for recurrent myocardial infarction in patients with type 2 diabetes mellitus

Abstract

Background: Recurrent myocardial infarction (MI) is one of the commonest and serious cardiovascular complications among diabetic patients. This study was planned to predict the recurrent MI development in diabetic patients using parameters of energy and adipokine metabolism. Methods: In total, 74 patients with type DM and acute MI were enrolled in the study. Measurements of serum adropin, irisin, fatty acid binding protein 4 (FABP4) and C1q/TNF related protein 3 levels were performed using enzyme-linked immunosorbent assay. The study employed generalized linear mixed models (GLMMs) to predict recurrent MI development. Results: The accuracy of predicting the absence of recurrent MI during a year in diabetic patients was 98.4%, and the accuracy of predicting the probability of recurrent MI during a year in diabetic patients was 92.3%. The global model accuracy was 97.3%. Conclusions:GLMM has shown that the levels of irisin on day 14, insulin on day 1 and the combined effect of blood glucose on days 1 and 14 were negative prognostic factors. The total impact of FABP4 on day 1, HOMA-IR on day 1 and adropin on day 14 had a positive prognostic value.