Abstract 16019: What Does Carotid Intima-Media Thickness and Plaque Add to Prediction of Stroke and Cardiovascular Disease Risk in Older Adults? The Cardiovascular Health Study

Abstract

Background: Framingham Risk Score (FRS) and other traditional stroke and cardiovascular disease (CVD) risk factors are limited since the majority of incident stroke and CVD events occur in low- and intermediate-risk groups. We evaluated whether addition of carotid ultrasound intima-media thickness (CIMT) measurements and presence of plaque help predict incident stroke and CVD in older adults.

Methods: Ten-year risk prediction Cox proportional hazards models for stroke and CVD were calculated using CHS-specific coefficients for FRS factors. CIMT was calculated as average of standardized maximum right and left internal and common CIMT. CVD was defined as coronary heart disease + heart failure + stroke. Categories of CIMT and CIMT + plaque were added to baseline FRS prediction models and net reclassification improvement (NRI) and area under receiver-operating characteristic curve (AUC) were calculated for each model.

Results: In 4,384 CHS participants (61% women, 14% black, baseline age 72 ± 5 yrs) without CVD at baseline, higher CIMT category and presence of plaque were both associated with higher incidence rates for stroke (and CVD) per 1,000 person-years (see Table). Addition of CIMT improved ability of FRS model to discriminate cases from non-cases of incident stroke (AUC: 0.711 vs. 0.699, p=0.01) and CVD (AUC: 0.679 vs. 0.669, p<0.001), with no additional improvement by adding plaque. Addition of CIMT to FRS model modestly improved NRI for prediction of both stroke (NRI=4.23%, p=0.037) and overall CVD (NRI=2.2%, p=0.007). For both outcomes, NRI was driven by reclassifying those without incident disease into a lower risk category. Addition of plaque to CIMT did not improve NRI significantly for either model.

Conclusion: In older adults, addition of CIMT modestly improves 10-year risk prediction for stroke and CVD beyond basic FRS model, mainly by down-classifying risk in those not experiencing stroke or CVD; addition of plaque to CIMT adds no incremental benefit.