Analyzing Heart Disease Risk Factors Using a Logistic Regression Approach

Heart disease remains one of the leading causes of disability and premature mortality worldwide, with its prevalence steadily increasing. This study analyzes a dataset comprising 13 variables related to heart disease to identify key predictors and develop robust predictive models. We address data quality issues, including outliers and influential points, and employ logistic regression and random forest techniques to predict the likelihood of heart disease and determine the most influential factors. The optimal predictive model was a reduced model utilizing seven key attributes: sex (male), chest pain types (cp1, cp2, cp3), exercise-induced angina (exang), stress test depression (oldpeak), ST segment slope (slope1, slope2), major vessels (ca1, ca2, ca3), and thallium test results (thal0, thal1, thal2). Among these, oldpeak, thal2, and exang1 were identified as the most impactful predictors of heart disease. This study highlights critical variables for heart disease prediction and demonstrates the utility of statistical and machine learning approaches in medical research.