Abstract P325: Association of Risk Factors and Extra-coronary Atherosclerosis with Progression of Coronary Artery Calcification in the Framingham Heart Study
Background: Progression of coronary artery calcification (CAC) is associated with future incidence of adverse cardiovascular disease. However, the interplay of CAC progression and atherosclerosis in different vascular beds has been less studied. We tested for the association of risk factors and baseline abdominal aortic calcification (AAC) with CAC progression in a large community cohort.
Methods: We determined risk factors, baseline AAC/CAC, and CAC progression in a sample of 1994 asymptomatic white participants from the Framingham Heart Study (FHS) who had undergone serial CT scanning within an average of 6.1 years. The primary outcomes were: (a) incident calcification (CAC > 0) in participants free of CAC at baseline; and (b) absolute progression of CAC (CAC follow-up > CAC baseline) in participants with detectable baseline CAC. To test the effect of risk factors and baseline AAC/CAC on CAC progression, we employed multivariable stepwise logistic and linear regression models of CAC progression with/without AAC, adjusted for traditional CVD risk factors.
Results: There were 1118 participants free of CAC at the baseline scan, of which 18.8% developed CAC in the follow-up scan. Of the 776 participants with detectable CAC at baseline, 84.9% developed progression of baseline CAC. In both subsets, AAC was a highly significant predictor of CAC incidence or progression, independent of other risk factors. Of note, the stepwise model including baseline AAC as a candidate for entry resulted in the same set of non-AAC predictors of CAC incidence or progression as the model without AAC. Table 1 displays the final logistic regression results for the cohort free of CAC. The AUC improved to 0.738 compared to an AUC of 0.722 for the model without AAC (p=0.002).
Conclusion: AAC is a strong independent predictor of CAC progression determined by incident CAC and CAC progression. Addition of AAC to the model that predicts incident CAC improves discrimination in a cohort free of baseline CAC.
Author Disclosures: O.K. Onuma: None. K.M. Pencina: None. J.M. Massaro: None. U. Hoffman: None. C.J. O'Donnell: None.
- © 2014 by American Heart Association, Inc.