Multidimensional Assessment of Heart Disease Risk: Associations Between Changes in Physiological Parameters and Morbidity

Abstract

Heart disease is one of the main reasons for death worldwide, with coronary heart disease, hypertensive heart disease, and cardiac arrhythmias being prevalent. Modern lifestyle changes (e.g., unhealthy diets and irregular work schedules) have increased the frequency of heart disease and impacted people of all ages, so it is important to accurately analyze its risk factors. However, there is still a research gap in assessing multiple factors' effects on heart disease through different models. In this paper, LDA (Linear Discriminant Analysis) is used as the core research method to model and analyze multiple health indicators (e.g., blood pressure, sex, age, heart rate, etc.). Compared with different models, like Logistic Regression, and QDA, in accuracy, sensitivity, and specificity, to get the best model. The study shows that LDA has high accuracy in classifying and identifying high-risk groups. Although it also cannot classify every single variable, after removing irrelevant variables, it also shows the best accuracy of the dataset. Therefore, the conclusion states that the LDA model performs well in heart disease risk assessment and can provide strong support for early analysis and accurate assessment.