Machine Learning-Based Analysis of Lifestyle Risk Factors for Atherosclerotic Cardiovascular Disease: Retrospective...

Area of Science:

• Cardiology
• Data Science
• Public Health

Background:

• Atherosclerotic cardiovascular disease (ASCVD) risk is influenced by lifestyle and chronic conditions.
• Individual ASCVD risk varies significantly.

Purpose of the Study:

• To evaluate the predictive accuracy of machine learning (ML) models for ASCVD risk.
• To incorporate lifestyle risk behaviors into ML models for enhanced prediction.
• To utilize the Korean nationwide health database for this analysis.

Main Methods:

• Utilized data from the Korea National Health and Nutrition Examination Survey (KNHANES).
• Applied five ML algorithms: logistic regression (LR), support vector machine, random forest, extreme gradient boosting, and light gradient boosting.
• Employed propensity score matching (PSM) for demographic confounder adjustment and Shapley additive explanations (SHAP) for variable importance.

Main Results:

• The extreme gradient boosting model achieved the highest area under the receiver operating characteristics curve (AUC).
• The light gradient boosting model showed superior performance in accuracy, recall, and F1-score.
• Key predictors identified included smoking, age, BMI, omega-3 intake, and LDL cholesterol.

Conclusions:

• ML models integrating lifestyle factors enhance ASCVD risk prediction compared to traditional methods.
• Personalized prevention strategies targeting lifestyle modifications can effectively mitigate ASCVD risk.