Abstract: Objective To establish a risk assessment model of cardio-cerebrovascular events in type 2 diabetes mellitus (T2DM) patients. Methods The study was conducted retrospectively to analysis 2 175 T2DM patients from diabetes mellitus management database. A total of 26 risk factor variables were chosen. The independent risk factors for the incidence of cardiovascular and cerebrovascular diseases of T2DM were evaluated by univariate and multivariate Cox regression analysis, and to construct a predictive incidence risk score model of cardio-cerebral vascular disease for T2DM. Results The median follow-up duration was 5.1 years in the model group. A total of 145 cases were diagnosed as the occurrence of cardio-cerebrovascular events at the end of follow-up. The cardio-cerebrovascular risk model for T2DM patients was 0.059×age (years)+0.936×smoking (yes=1)+0.006×glycosylated hemoglobin (%)+0.380×duration of diabetes (years)+0.048×body mass index (kg/m2)+0.009×systolic blood pressure (mmHg)+0.807×atrial fibrillation (yes=1)+0.175× non-high-density lipoprotein cholesterol (mmol/L)-0.034×log glomerular filtration rate(mL/min). The predictive probability of cardio-cerebrovascular disease in T2DM patients for five years was P ∧ = 1 - 0.928exp(∑i p= 1βi χi - 12.736 ). The goodness-of-fit result for the risk assessment model in the validation group was HL χ2=1.49, P=0.81, and the area under ROC curve was 0.798 and 95%CI was 0.759-0.818. Conclusion A simple risk assessment model on the occurrence of cardiocerebrovascular events with T2DM patients is established, which will not only contribute to monitor the risk of cardiocerebrovascular incidents of T2DM patients, but help to provide basis for clinical treatment.

Key words: arteriosclerosis, cardiovascular diseases, diabetes mellitus, type 2, risk assessment model