Modeling coronary heart disease risk based on age, fatty food consumption and anxiety factors using penalized spline nonparametric logistic regression

Coronary Heart Disease (CHD) represents a significant public health challenge in Indonesia, contributing substantially to morbidity and mortality rates. We propose a new method for modeling CHD risk by considering the influencing factors, namely age, fatty food consumption and anxiety using Penalized Spline Nonparametric Logistic Regression (PSNLR) approach to analyze the CHD and non-CHD risks affected by these influencing factors. The proposed method gave results a percentage of model classification accuracy of 94.31 % and the value of area under the receiver operating characteristic curve (AUC) of 0.96. This means that the proposed method is a powerful method in modeling the CHD risk and it can effectively identify the nonlinear relationship between influencing factors and CHD incidence. This research supports the achievement of Sustainable Development Goals (SDGs) point 3, especially the target of reducing deaths from non-communicable diseases such as CHD through improving disease prevention and treatment. The use of more accurate models in identifying risk factors also contributes to efforts to improve overall public health and strengthen health care systems.•A cross-sectional survey on influencing factors of CHD risk was conducted at the Airlangga University Hospital.•For analyzing relationship between influencing factors and CHD risk, we used PSNLR method.