Separating Risk Prediction: Myocardial Infarction vs. Ischemic Stroke in 6.2M Screenings

Area of Science:

• Cardiology
• Epidemiology
• Public Health

Background:

• Traditional cardiovascular disease risk models combine myocardial infarction (MI) and ischemic stroke (IS) risk.
• This combined approach may not accurately reflect the distinct causes and risk factors for MI and IS.
• Developing separate models is crucial for precise risk stratification.

Purpose of the Study:

• To develop and validate separate 5-year risk prediction models for MI and IS.
• To identify distinct predictors and their differential impact on MI versus IS risk.
• To improve early detection and identification of at-risk populations for atherosclerotic cardiovascular disease.

Main Methods:

• Analysis of 6,242,404 individuals aged over 40 from a 2009 cardiovascular health screening.
• Utilized Cox proportional hazards models to construct separate 5-year risk prediction models for MI and IS.
• Assessed model performance using discrimination (c-indices) and calibration.

Main Results:

• The MI model achieved a c-index of 0.709, and the IS model achieved 0.770.
• Key predictors included age, sex, BMI, smoking, alcohol, physical activity, diabetes, hypertension, dyslipidemia, CKD, and family history.
• Hypertension and age had a greater impact on IS risk, while smoking, BMI, dyslipidemia, and CKD were more influential for MI risk.

Conclusions:

• Separate risk prediction models for MI and IS are necessary for accurate assessment.
• Tailored risk stratification can enhance the early detection of heterogeneous at-risk populations.
• These findings support personalized approaches to managing atherosclerotic cardiovascular disease risk.