Thirty-Year Multivariate Risk Assessment Is a Stronger Predictor of Cardiovascular Disease than the 10-Year Model

Abstract

Identifying individuals at risk for developing coronary heart disease (CHD)1 is fundamental to the practice of primary prevention in modern cardiology because it enables clinicians to determine which patients are eligible to receive evidence-based therapies that can prevent or delay the progression of atherosclerotic disease. Several multivariable risk-prediction algorithms are available, and the most widely used in the US is the Framingham cardiovascular disease (CVD) risk profile, from which the Framingham Risk Score (FRS) is derived. The FRS estimates an individual’s risk of developing nonfatal myocardial infarction or CHD-related death (“hard events”) over a 10-year period, on the basis of what are now considered to be the traditional risk factors of age, total cholesterol concentrations, HDL cholesterol, systolic blood pressure, and cigarette smoking status; the equations are sex specific. Individuals are then stratified into low risk ( 20% risk over 10 years) based on their score (1), and this risk score then guides a clinician’s decision to institute antiplatelet or lipid-lowering therapy or advise intensive lifestyle modification. Several multivariate risk assessment algorithms in addition to the FRS have since been proposed and validated in diverse populations to estimate risk for a period of less than 10 years. The FRS has been critiqued for its relatively limited view of only a10-year timeframe, because an individual’s lifetime risk of CHD may be high, but a 10-year risk prediction model sometimes underestimates this risk and can delay efforts to modify that risk. A greater duration of exposure to untreated risk factors allows for continued progression of subclinical atherosclerosis. Examination of the Framingham cohort showed that despite younger individuals in lower risk groups having a very low 10-year CHD …