Comparison of the coronary calcium score with the estimated coronary risk

Abstract

The role of coronary calcium scoring in coronary risk estimation is not well established. Calcium scoring could provide additional information in a certain subgroup of patients where the calcium score does not match the conventional Framingham risk estimates. We explored the characteristics of such a subgroup. The study participants were 1653 asymptomatic persons who underwent routine health screening and calcium scoring using the 16-slice multidetector computed tomography. Risk stratification was performed in five categories both by 10-year Framingham coronary risk and the Agatston coronary calcium score. Risk stratifications by coronary calcium score and absolute risk showed a large discrepancy (difference > or =3 classes) in about 9% of participants. The proportion increased with age (P for trend <0.0001). An exploratory analysis revealed that age (partial R=0.109, P<0.0001) and the presence of the metabolic syndrome (partial R=0.025, P<0.001) were independent variables that accounted for the variance of the residual of regression between the log-transformed value of coronary calcium score and the absolute risk. Calcium scoring may be clinically more useful in older (> or =50 years) participants and/or in participants with the metabolic syndrome because of the relatively higher probability of obtaining additional information that the conventional Framingham risk estimation cannot provide.