Framingham Risk Score Calculator

The Framingham Risk Score estimates the absolute probability that a person free of cardiovascular disease will suffer an event within ten years. Underneath the point table sits a Cox proportional hazards model fitted to observed follow-up data; the integer points are a bedside approximation of that model, produced by scaling each regression coefficient against a reference unit of risk. This is why the points behave in ways that look odd at first glance. Cholesterol and smoking points shrink as age rises, because the relative hazard of those exposures attenuates in older people even though their absolute risk climbs, and the age term absorbs the difference. HDL cholesterol above 60 mg/dL earns minus one point, the only protective term in the score. Separate tables exist for men and women because both baseline incidence and the age gradient differ substantially by sex. Results are read as under 10 percent low risk, 10 to 20 percent intermediate, and above 20 percent high, the last threshold being the coronary risk equivalent adopted by the ATP III cholesterol guidelines. Accuracy depends on resembling the source population. The score was fitted to a largely white cohort in one Massachusetts town, and validation work has found it overestimates risk in several European and East Asian populations while underestimating it in South Asians. It omits family history of premature disease, chronic kidney disease, obesity, coronary artery calcium and inflammatory markers, and it is only defined for ages 20 to 79.

The Framingham Heart Study began in 1948, run by the United States Public Health Service and later the National Heart, Lung and Blood Institute, after cardiovascular disease had become the leading cause of death in the country with almost no understanding of why. Investigators enrolled 5,209 residents of Framingham, Massachusetts aged 30 to 62 and examined them every two years. The study is the origin of the phrase risk factor itself, introduced by William Kannel and colleagues in a 1961 paper in the Annals of Internal Medicine that linked blood pressure and cholesterol to coronary disease. An Offspring Cohort was added in 1971 and the multi-ethnic Omni Cohort in 1994. Turning those observations into a usable prediction tool took decades. In 1998 Peter Wilson and coauthors published prediction of coronary heart disease using risk factor categories in Circulation, which became the basis of the ATP III risk calculator issued in 2001. In 2008 Ralph D Agostino Sr and colleagues published the General Cardiovascular Risk Profile for Use in Primary Care, also in Circulation, broadening the endpoint from coronary events to global cardiovascular disease including stroke, peripheral artery disease and heart failure. Recognition of the cohort limitations drove successors: the American College of Cardiology and American Heart Association released the multi-ethnic Pooled Cohort Equations in 2013, and European practice moved to the SCORE and later SCORE2 systems calibrated to regional risk.