OBJECTIVES: The aim was to validate the Framingham coronary heart disease (CHD) risk function with the formula by Wilson et al. (1998) in Czech men. METHODS: The validation was performed within the 20-year primary prevention study of atherosclerotic risk factors (STULONG) including 1417 middle-aged men from the Czech Republic (Prague). A total of 646 men examined in 1979-1988, and followed-up for ten years, were included into the validation study. The calibration and discrimination ability of the Framingham risk function in the Czech population were explored. RESULTS: The estimated 10-year risk of CHD by the Framingham risk function was 12.8% in 646 men, significantly higher than the observed risk (16.4 %), p = 0.013. The trend in the 10-year incidence of CHD was significantly increasing with quintiles of the estimated risk, p < 0.001. After the recalibration of the Framingham risk function, there was an insignificant difference between the estimated (18.2%) and observed (16.4%) risks of CHD, p = 0.320. The Framingham risk function classified men into those with and without CHD in the 10-year period with accuracy over 60%. CONCLUSIONS: Unlike some validation studies from Western Europe, the Framingham risk function significantly underestimated the 10-year CHD risk in the Czech Republic. In agreement with these studies, the incidence of CHD was significantly increasing across quintiles of the estimated risk.

EuroMISE Centre Charles University and Academy of Sciences Institute of Computer Science AS CR Prague Czech Republic