Machine Learning-Based Prediction Model for Predicting the Effect of the Serum γKlotho Level on Susceptibility to...

Area of Science:

• Cardiology
• Biomarkers
• Machine Learning in Healthcare

Background:

• Coronary heart disease (CHD) remains a leading cause of mortality worldwide.
• Identifying novel biomarkers and improving predictive models for CHD risk is crucial for early intervention.
• Serum γKlotho has emerged as a potential factor influencing cardiovascular health.

Purpose of the Study:

• To investigate the association between serum γKlotho levels and the risk of developing coronary heart disease (CHD).
• To develop and validate a machine learning (ML) model for predicting CHD risk.
• To evaluate the clinical utility of serum γKlotho as a biomarker for CHD.

Main Methods:

• Analysis of 1435 subjects, randomly assigned to training (70%) and validation (30%) groups.
• Utilized univariate and least absolute shrinkage and selection operator (LASSO) regression to identify independent risk factors for CHD.
• Developed and evaluated nine ML models, selecting the best performing model (Random Forest - RF) for validation using decision curve analysis (DCA).

Main Results:

• Key factors independently associated with CHD risk include age, serum γKlotho levels, LDL-C, sex, diabetes, hypertension, and smoking status.
• The Random Forest (RF) model demonstrated superior performance compared to eight other ML models.
• Validation confirmed the promising clinical applicability of the developed RF model for CHD risk prediction.

Conclusions:

• Serum γKlotho is a novel biomarker positively correlated with coronary heart disease (CHD) risk.
• The Random Forest (RF) model provides a robust and accurate method for predicting CHD risk.
• The RF model is well-suited for clinical application in assessing and managing CHD risk.