Comparison of the Framingham risk function-based coronary chart with risk function from an Italian population study.

AIMS The aim is to compare the coronary risk chart published by the European Task Force for Prevention of Coronary Heart Disease and produced using a Framingham risk function, with a risk function derived from an Italian population study. METHODS AND RESULTS Coronary risk function in this study was the result of longitudinal experience in an Italian middle-aged population of 1656 male subjects followed-up for 25 years. To comply with the Framingham equation the same risk factors (age, systolic blood pressure, total serum cholesterol and smoking habits), end-points (any possible coronary event including angina pectoris), and length of follow-up (10 years) were used, and the model (log-linear accelerated time failure model, accommodating the Weibull distribution) was similar. Comparisons were made computing the coronary risk for each cell of the coronary risk chart for men aged 40, 50 and 60 years. The Italian risk function produced highly significant coefficients for all four risk factors. Forty-four out of a total of 120 cells had a coronary risk of 20% or more in 10 years following the coronary risk chart, whereas this was reduced to four while using the Italian risk function (P<0.001). The Italian risk function largely underestimated the corresponding levels produced by the coronary risk chart and vice versa. CONCLUSION The Framingham risk function-based coronary risk chart overestimates absolute coronary risk in countries characterized by a lower incidence of coronary events and should be used with caution.