Title: Association and Predictive Ability of LDL and HDL Cholesterol for Coronary Heart Disease Using Binary Logistic Regression
Authors: Umi Habibah, Nur Chamidah, Marisa Rifada, Dita Amelia
Volume: 10
Issue: 3
Pages: 57-62
Publication Date: 2026/03/28
Abstract:
Coronary heart disease (CHD) is a major cause of mortality worldwide and is closely related to lipid profile abnormalities. This study aimed to examine the effects of Low-Density Lipoprotein (LDL) and High-Density Lipoprotein (HDL) cholesterol levels on CHD incidence using binary logistic regression. Primary data were collected from medical records and interviews at RSUA, involving 38 patients. CHD status was treated as a binary response variable, while LDL and HDL served as predictors. Descriptive results showed that 65.8% of patients were diagnosed with CHD, with noticeable variation in lipid profiles. Logistic regression analysis indicated that HDL had a significant negative association with CHD (p < 0.05), whereas LDL was not statistically significant. Model evaluation demonstrated acceptable performance, with an accuracy of 73.7%, sensitivity of 84%, specificity of 53.8%, and an AUC value of 0.840, indicating good discriminative ability. Overall, HDL showed a stronger contribution to CHD prediction than LDL.