LOGISTIC REGRESSION MODELING OF DISEASE RISK FACTORS: A CASE STUDY

Abstract

Cardiovascular disease remains a major global health challenge, requiring accurate and interpretable predictive tools for early diagnosis and effective prevention. This study develops a multivariable logistic regression model to identify significant demographic and clinical factors associated with heart disease. Using a structured clinical dataset, the analysis examined key predictors including age, chest pain type, resting blood pressure, cholesterol, exercise-induced angina, and ST depression. The results reveal that age, abnormal chest pain categories, exercise-induced angina, and ST depression are strong independent predictors of disease presence. The final model demonstrated excellent performance with high classification accuracy and an AUC exceeding 0.80, indicating strong discriminative ability. Model diagnostics confirmed adequate goodness-of-fit and stable coefficient estimates. These findings highlight the importance of integrating clinical, physiological, and stress-test variables to enhance cardiovascular risk assessment. The study reinforces the value of logistic regression as an interpretable and reliable approach for informing clinical decision-making and provides a foundation for future research incorporating larger samples, additional biomarkers, and external validation to improve predictive precision.