External validation of integrated genetic-epigenetic biomarkers for predicting incident coronary heart disease

Area of Science:

• Cardiovascular disease research
• Genetics and epigenetics
• Machine learning in healthcare

Background:

• Traditional risk scores like the Framingham Risk Score (FRS) and Pooled Cohort Equations (PCE) have limitations in accurately predicting incident coronary heart disease (CHD).
• Improved risk stratification is crucial for effective prevention strategies against cardiovascular events.

Purpose of the Study:

• To develop and validate a novel integrated genetic-epigenetic tool for enhanced CHD risk prediction.
• To compare the performance of the new tool against established risk assessment methods.

Main Methods:

• Utilized machine learning techniques on datasets from the Framingham Heart Study (FHS) and Intermountain Healthcare (IM).
• Developed and validated an integrated model incorporating genetic and epigenetic markers.
• Assessed prediction accuracy using sensitivity and specificity metrics.

Main Results:

• The novel genetic-epigenetic model demonstrated superior sensitivity and specificity compared to FRS and PCE across both validation cohorts.
• Achieved a sensitivity/specificity of 79%/75% in the FHS test set and 75%/72% in the IM set.
• Outperformed FRS (15%/93% FHS, 31%/89% IM) and PCE (41%/74% FHS, 69%/55% IM) in predicting 3-year incident CHD.

Conclusions:

• The developed integrated genetic-epigenetic tool offers a more accurate method for CHD risk stratification.
• Clinical implementation of this tool could significantly enhance the identification of patients at high risk for heart attack.
• This approach shows high generalizability and improved predictive performance over current standards.