Abstract 13552: Carotid Atherosclerosis and Dementia Risk in Older Adults. The Three-City Study
Background: While carotid atherosclerosis has been associated with cognitive decline or prevalent dementia, there is limited evidence relating carotid atherosclerosis with incident dementia on a prospective basis.
Objectives: To study the respective association of baseline carotid plaques and carotid intima media thickness (IMT) with incident all-cause dementia and dementia subtypes. Furthermore, the added value of carotid atherosclerosis for individual dementia risk prediction is quantified.
Methods: Within the population-based Three-City study, 6,025 dementia-free men and women aged 65-85 underwent a standardized high-resolution -B-mode carotid ultrasound examination at baseline allowing to detect carotid plaques in the common carotid arteries (CCAs), the bifurcations and the origin of the internal carotid arteries and to quantify IMT per se in the CCAs. Incident all causes and dementia subtypes (Alzheimer disease (AD) and vascular/mixed dementia (VaD)) were validated by an independent expert committee. Hazards Ratio (HR) of the number of sites with carotid plaques (0, 1 and ≥2) and of 1 SD increase in CCA-IMT were estimated using Cox regression models. The added value of carotid atherosclerosis for dementia risk prediction was measured by the Harell‘s C index and the continuous net reclassification improvement index (NRI).
Results: After a mean follow-up of 5.4 years, 421 subjects had incident dementia, including 272 AD and 83 VaD. In multivariate analyses, carotid plaques were independently related to VaD only (HR for≥2 sites with plaques=1.93, 95% confidence interval (CI) 1.13-3.28; p for trend=0.013). CCA-IMT was not associated with dementia of any type (standardized HR for VaD=1.07, 95% CI: 0.85-1.33). Further adjustment for intercurrent stroke as time dependent variables or controlling for competing risk by death marginally modified the results. Adding carotid plaques to established dementia risk factors improved the Harell’s c index from 0.716 to 0.739 (p=0.07) and yielded a significant continuous NRI of 43% (95% CI= 20.2-66.2; p < 0.001) regarding VaD risk.
Conclusion: In elderly men and women, carotid plaques on 2 sites or more are independent predictors of incident VaD and may improve the individual risk prediction of VaD.
- © 2013 by American Heart Association, Inc.