Abstract 1187: Imaging Measures for Prediction of Cardiovascular Events in Men and Women: The Multi-Ethnic Study of Atherosclerosis (MESA)
Background: Imaging-derived subclinical disease measures, including coronary artery calcium (CAC), carotid intima-media thickness (IMT) and left ventricular mass (LVM) have been shown to be independently associated with an increased risk of future cardiovascular disease (CVD). We compared these imaging subclinical disease measures for their overall- and gender-specific ability to predict incident CVD.
Methods: The study sample consisted of a total of 4965 participants (48% males, mean age 62±10 years; 39% Caucasian, 13% Chinese, 26% African American, 22% Hispanic) who had available measures of CAC by computed tomography, IMT by ultrasonography and LV mass and volume by magnetic resonance imaging. All participants were free of clinical CVD at baseline and were prospectively followed for a median of 4.8 years.
Results: There were 257 CVD events during follow up, including 159 coronary heart disease (CHD) events, 76 heart failure (HF) events and 57 strokes. After adjustment for traditional risk factors and other imaging measures, incident HF was best predicted by body size adjusted LVM (Hazard Ratio [HR]: 4.39, 95% CI 2.44–7.90, for top quartile vs. bottom two quartiles of LVM). Incident stroke was best predicted by increased LVM-volume ratio (HR: 1.42 per 1 standard deviation [SD] increment, 95% CI 1.16–1.72). Incident CHD was most strongly associated with log-transformed CAC (HR: 2.44 per 1 SD, 95% CI 1.98–3.02). Log-transformed CAC was also the best imaging measure of incident total CVD events overall (HR: 1.69 per 1 SD, 95% CI 1.45–1.97) and for men (HR: 1.96 per 1 SD, 95% CI 1.61–2.37) and women (HR: 1.39 per 1 SD, 95% CI 1.08–1.79) separately. As compared to traditional risk factors alone, overall risk prediction (c-statistic) for future CHD, stroke, HF and all CVD was significantly improved by adding CAC, LVM-volume ratio, LVM and CAC respectively (all p<0.05). None of the imaging measures showed a significant interaction with gender for any of the CVD outcomes.
Conclusions: In a large multi-ethic population free of clinical CVD at baseline, LVM predicted HF, LVM-volume ratio predicted stroke, and CAC was most strongly associated with CHD and all CVD. There was no evidence that imaging measures differed in their association with incident CVD by gender.