Abstract

This study aimed to identify the primary behavioral and health risk factors associated with the likelihood of myocardial infarction (heart attack) and to construct a probabilistic predictive model using binary logistic regression among patients attending cardiac centers in Aden Governorate. The study employed a cross-sectional design utilizing primary data collected from a random sample of 51 patients across cardiac centers in Aden hospitals between October 2025 and January 2026.The model included five independent explanatory variables: age, smoking, Qat chewing, hypertension, and diabetes mellitus. The results demonstrated that the estimated model possesses overall statistical significance (p = 0.005) and an excellent goodness-of-fit according to the Hosmer–Lemeshow test (p = 0.836). Wald test results indicated that hypertension, smoking, age, and diabetes mellitus are positively and significantly associated with an increased odd of occurrence. Hypertension emerged as the most influential predictor ({Exp}(B) = 16.057, p = 0.010), followed by smoking ({Exp}(B) = 5.346, p = 0.030), age ({Exp}(B) = 1.070, p = 0.048), and diabetes ({Exp}(B) = 1.063, p = 0.020), whereas Qat chewing demonstrated no statistically significant effect. The model achieved an overall classification accuracy of 78.4% with a high diagnostic sensitivity of 94.6%. The findings conclude that the binary logistic regression model exhibits excellent fit and high diagnostic precision, highlighting hypertension and smoking as the primary risk factors driving the increased odds of heart attacks in Aden.