Author + information
- Chintan S. Desai,
- Colby R. Ayers,
- Matthew Budoff,
- Raimund Erbel,
- Maryam Kavousi,
- Amit Khera,
- Nils Lehmann,
- Kiang Liu,
- Stefan Mohlenkamp,
- Hongyan Ning,
- Jacqueline Witteman and
- Philip Greenland
Coronary artery calcium (CAC) testing is considered most useful for coronary heart disease (CHD) risk estimation primarily in “intermediate risk” individuals. We sought to determine prognostic significance and added predictive value of CAC in “low risk” women, in whom global risk factor scores are suboptimal.
We included women without diabetes and 10-year Framingham risk score (FRS) for CHD < 10% from the Multi-Ethnic Study of Atherosclerosis (MESA), Dallas Heart Study (DHS), Heinz Nixdorf Recall (HNR) study, and the Rotterdam Study. CAC score was classified into the following strata: 0, 1-10, 11-100, 101-300, > 300. The primary outcome was CHD, defined as nonfatal myocardial infarction and CHD death. We used Cox proportional hazards models to calculate multivariable-adjusted hazards ratios (HR) for CHD and calculated category-free net reclassification improvement (NRI) for addition of CAC to the baseline model containing study-specific coefficients for age, systolic blood pressure, total and HDL cholesterol, smoking, and antihypertensive medication. We used fixed-effects meta-analysis to combine effect measures.
In 6970 low-risk women (mean age 58 years), the prevalence of CAC was: MESA, 32% (N = 880); DHS, 42% (N = 496); HNR, 50% (N = 908); Rotterdam, 73% (N = 949). Over 8 years of follow up, the CHD event rate in women with no CAC ranged from 0.3% (DHS) to 1.2% (Rotterdam); the event rate in women with CAC > 0 ranged from 1.0% (DHS) to 3.3% (Rotterdam). Multivariable-adjusted HRs for CHD in women with any CAC compared to women without CAC ranged from 1.64 (0.65-4.13, HNR) to 2.08 (0.72-6.02, Rotterdam). The HR for CHD from meta-analysis was 1.85 (1.12-3.04). The HR from meta-analysis for CHD in women with CAC > 300 compared to women with no CAC was 6.48 (3.07-13.67). The NRI with addition of prevalent CAC to the baseline model ranged from 0.27 (Rotterdam) to 0.61 (DHS), and was 0.37 (P < 0.001) for the meta-analysis. When CAC strata were added to the baseline model instead of CAC presence, the NRI from meta-analysis was 0.50 (P < 0.001).
In low-risk women from four large cohorts, CAC presence was associated with incident CHD and improved risk classification for CHD.
Poster Sessions, Expo North
Sunday, March 10, 2013, 9:45 a.m.-10:30 a.m.
Session Title: Imaging: CT/Multimodality VI
Abstract Category: 20. Imaging: CT/Multimodality
Presentation Number: 1230-366
- 2013 American College of Cardiology Foundation