Abstract 17432: Predictive Value of Framingham Risk and Coronary Calcium in High and Low Risk Populations
OBJECTIVE: Evaluate predictors of mortality in a high-risk population at a tertiary health care center in comparison to the low risk MESA (Multiethnic Study of Atherosclerosis) population using coronary artery calcium score (CACS) and the Framingham risk score (FRS).
DESIGN AND METHODS: Retrospective analysis and comparison of the MESA and VA Loma Linda cardiac CT databases was performed. The VA database included 907 volunteers and patients who underwent imaging for various clinical indications including chest pain, heart failure, abnormal/equivocal stress tests. The mean age and percent male of the VA and MESA databases were 59 and 62 yrs; 90% and 47% respectively. CACS measured in Agatston Units was used as a categorical variable (0, 1-100, 101-300 and >300) and as a continuous log variable. Cox proportional hazard modeling with appropriate adjustments was used to test association with all cause mortality (ACM). Receiver operator curves were constructed to calculate the area under the curve (AUC) for ACM as predicted by CACS and FRS. Similar analyses were performed using the multicenter MESA database which included 6814 asymptomatic individuals selected from the various communities across the US.
RESULTS: Only Age [HR: 1.63 (1.22-2.18) vs 2.27 (1.82-2.82) per SD increase], smoking history [HR: 2.73 (1.77-4.20) vs 1.84 (1.38-2.44)] and logCAC [HR: 2.14 (1.59-2.88) vs 1.21 (1.01-1.43) per SD increase] emerged as significant predictors of ACM in the VA database. Counter-intuitively, low LDL emerged as a significant predictor of ACM in the MESA database. Addition of CAC to FRS significantly improved the AUC for ACM in the VA (FRS only: 0.64 vs FRS+CAC: 0.70; p value < 0.0001) and the MESA (FRS only: 0.68 vs FRS+CAC: 0.70; p value 0.001) databases.
CONCLUSIONS: Age, smoking history and logCAC are significant predictors of mortality in both VA and MESA databases. CAC also improves risk stratification above and beyond the FRS in both high and low risk populations.
- © 2012 by American Heart Association, Inc.