Predictors of Myocardial Infarction using Weibull Distribution: A Cohort Study
Background and objective - Myocardial Infarction (MI) may be the primary symptom of coronary artery diseases. The present study aimed at determining the predictors of myocardial infarction and the time of MI event using the Weibull distribution.
Methods - In a cohort study, 5187 people aged over 30 years old were followed up for 6.5 years. Weibull distribution and Cox proportional hazard models were used in analysis .Data were analyzed with STATA Ver. 10.
Results - A total of 83 new cases of MI (53 men and 30 women) were observed. The cumulative incidence of MI was 2.5 per 1000. There was a significant association between fasting plasma glucose (FBS), age, gender, LDL-cholesterol, systolic blood pressure (SBP), and smoking with MI event in Weibull model. After adjusting age, gender, LDL-cholesterol, SBP, TG, and smoking, having a FBS level more than 126 mg/dl decreased the time of MI event up to 2.87 folds compared to normal people.
Conclusions - FBS, age, male gender, LDL-cholesterol, SBP, and smoking were the important predictive factors for Myocardial Infarction. FBS had the highest prediction level for the MI occurrence.
Keywords - Predictor, Myocardial Infarction, Time ratio, Hazard ratio, Weibull distribution