January 10, 2026Vascular Medicine0 citations
View Full PaperCarotid artery ultrasound grayscale median and incident dementia: The Multi-Ethnic Study of Atherosclerosis (MESA)
Key Result
Lower carotid ultrasound grayscale median (GSM) independently predicted incident all-cause dementia with a hazard ratio of 1.45 (95% CI 1.11–1.90, p = 0.021).
Key Points
- This study aims to assess whether carotid ultrasound grayscale median can predict the risk of developing all-cause dementia.
- Measured carotid ultrasound grayscale median at baseline among adults with no atherosclerotic cardiovascular disease.
- Identified incident dementia events through hospital and death records over a median follow-up of 13.7 years.
- Analyzed data using Cox proportional hazards models with natural cubic splines.
- Out of 1788 participants, 157 incidents of all-cause dementia were recorded over the follow-up period.
- Lower grayscale median significantly predicted incident dementia, with a hazard ratio of 1.45 for those in the lowest tertile compared to the highest.
- Findings remained significant even after adjusting for carotid intima-media thickness.
Structured PICO
Does lower carotid ultrasound grayscale median (GSM) predict incident all-cause dementia in adults free of atherosclerotic cardiovascular disease?
P
Population
1,788 adults free of atherosclerotic cardiovascular disease, mean age 63.1, 53% women.
I
Intervention
Lower carotid ultrasound grayscale median (GSM) (1st tertile)
C
Comparator
Higher carotid ultrasound grayscale median (GSM) (3rd tertile)
O
Outcome
Incident all-cause dementia identified by hospital and death recordshard clinical
Lower carotid ultrasound grayscale median (GSM) independently predicts incident all-cause dementia, suggesting it may serve as an early imaging biomarker for dementia risk.
Main Result
Absolute Event Rate: 0% vs 0%
Abstract
Introduction: Vascular contributions to cognitive impairment and dementia are potentially modifiable. Early detection of reversible arterial injury may improve risk stratification and provide treatment monitoring. We hypothesized that carotid ultrasound grayscale median (GSM), a novel imaging biomarker of early arterial injury, would predict incident all-cause dementia in the Multi-Ethnic Study of Atherosclerosis (MESA). Methods: The MESA enrolled adults free of atherosclerotic cardiovascular disease. Common carotid GSM was measured at baseline. Incident all-cause dementia events were identified by hospital and death records. Cox proportional hazards models with natural cubic splines investigated the association of baseline GSM and all-cause dementia. Results: The 1788 participants were a mean (SD) 63.1 (10.3) years old and 53% were women. Over a median 13.7 years, 157 all-cause dementia events occurred. In fully adjusted models, with additional adjustment for carotid intima–media thickness, lower (worse) GSM independently predicted incident all-cause dementia (hazard ratio, 1st to 3rd tertile, 1.45 95% CI, 1.11–1.90, p = 0.021). Conclusions: Lower GSM independently predicts all-cause dementia, beyond traditional arterial injury measures, suggesting it may serve as an early marker of dementia risk.