Atherosclerosis and Dementia
A Cross-Sectional Study With Pathological Analysis of the Carotid Arteries
Background and Purpose—Previous ultrasound-based studies have shown an association between carotid artery atherosclerosis and dementia. Our aim was to investigate this association using postmortem examination.
Methods—Postmortem morphometric measurements of carotid stenosis and intima-media thickness were performed in individuals with dementia (n=112) and control subjects (n=577). Multivariate logistic regression models were applied.
Results—High-grade left internal carotid stenosis (≥70%) was associated with increased odds for dementia (OR, 2.30; 95% CI, 1.14–4.74; P=0.02). Intima-media thickness was not associated with dementia.
Conclusions—The likelihood of dementia is increased with high-grade left internal carotid artery atherosclerosis after adjusting for demographic and cardiovascular risk factors.
Previous studies using ultrasound as an assessment method have shown an association between carotid artery atherosclerosis (CAA), including intima-media thickness (IMT), and a higher risk of dementia.1–3 Although ultrasound is a well-established method in routine clinical practice, it is associated with interexaminer variability in IMT and plaque measurements.4 Moreover, specific segments of the carotid artery are difficult to evaluate, particularly the bulb and the internal carotid artery.5 Morphometric evaluation of CAA in autopsy specimens allows evaluation of the limits of the plaque along the entire arterial length and the precise definition of the boundaries of the intima and media layers. We aimed to investigate the association of dementia and CAA measured by morphometric methods in a large autopsy series.
Specimens were collected from 2005 to 2008 at the Human Brain Bank of the Brazilian Aging Brain Study Group, including cases >50 years of age from a general autopsy service in Sao Paulo, Brazil. Demographics, cardiovascular risk factors, and cognitive status were obtained with the next of kin after signing a statement of informed consent approved by the local ethics and research committee. See http://stroke.ahajournals.org for complete methodology in the supplemental section.
The percentage of carotid stenosis was calculated by subtracting the lumen area from the intima area, dividing the difference by the intima area, and multiplying the result by 100 (Figure). IMT was calculated by dividing the intima-media area (defined as the area internal to the external elastic lamina up to the lumen) by the media perimeter (delineated by the external elastic lamina).
Neuropathological examination was performed in a subgroup of participants. Microvascular changes were analyzed semiquantitatively using hematoxylin and eosin staining in 14 routinely sampled areas plus additional areas detected at the macroscopic examination. The presence of lacunae, microinfarcts, and infarcts was registered. Clinical and neuropathological evaluations were carried out double-blinded.
We conducted univariate and multivariate logistic regression analyses to assess the correlation between dementia (binary outcome) and carotid artery obstruction (defined as of ≥70% obstruction) and quartiles of IMT. Results were adjusted to demographic, cardiovascular risk factors, and presence of stroke.
A total of 689 cases were analyzed. Demographics and cardiovascular risk factors of the sample are shown in the Table. Stroke reported by the family was more prevalent among demented individuals (25.0% versus 8.8%; P<0.001). Of the 689 participants, 233 (33.4%) had neuropathological examination. Lacunar infarctions were present in 32 participants (13.7%) and was similarly distributed between individuals with and without dementia (14 of 71 versus 18 of 162, P=0.08, respectively). High-grade stenosis was detected in 3% to 9.4% of participants according to the artery segment (Supplemental Table I). The mean IMT was 1.12±0.30 mm for the common carotid artery and 0.86±0.34 mm for the internal carotid artery.
In the multivariate model adjusting for potential confounding factors, we observed that high-grade left internal carotid stenosis is associated with increased odds for dementia (OR, 2.55; 95% CI, 1.26–5.15; P=0.009). This association remains significant even after additional adjustment for the presence of stroke (P=0.02). There was a nonsignificant trend for an association between right internal carotid artery stenosis and dementia (OR, 1.96; 95% CI, 0.94–4.08; P=0.07). Stenosis located at the common carotid arteries was not associated with dementia (left: OR, 1.76; 95% CI, 0.67–4.61; P=0.25 and right: OR, 0.24; 95% CI, 0.02–2.52; P=0.23; Supplemental Table I). Results are similar when selecting only the moderate and severe cases of dementia versus no dementia. Individuals with and without dementia were similar for IMT (common: OR, 0.94; 95% CI, 0.79–1.13; P=0.52 and internal: OR, 0.87; 95% CI, 0.73–1.05; P=0.14).
Although widely used, the interpretation of ultrasonographic evaluation may be dependent on sonographer expertise and can be difficult to perform due to unfavorable patient biotype. In the Rotterdam Study, common carotid artery IMT was measurable in 96% of individuals, carotid bulb in 64%, and internal carotid artery in 31%.5 Our study is the first to describe the association between dementia and CAA using morphometric evaluation in autopsy specimens.
We found differences that were specific for side and segment of the carotid artery segment. Severe left-sided atherosclerotic lesions are expected to have a greater impact on cognition than CAA on the right side.1 Differences in hemodynamics among distinct carotid artery segments may explain why the association between CAA and dementia is significant only for severe stenosis located in the internal carotid artery. In our study, IMT showed no association with dementia. Previous studies have reported conflicting results regarding this association.1–3
Cognitive impairment may be associated with high-grade stenosis of the internal carotid artery even without evidence of infarction on MRI or a history of stroke.1 Indeed, in our study, the analysis of 233 patients showed that prevalence of lacunar infarctions is similar between individuals with and without dementia. Cerebral small vessel diseases may mediate the link between atherosclerosis and dementia through direct vascular brain damage or by accelerating clinical expression of dementia. Silent embolization and chronic cerebral hypoperfusion are also possible mechanisms implicated in CAA-related cognitive impairment. Because we did not assess language and aphasia may impact the dementia diagnosis, our results need to be interpreted in light of this limitation. Further studies, including animal models, are needed to clarify the precise etiology of the dementia and to determine the impact of vascular brain lesions on cognitive function.
Sources of Funding
Supported by LIM22-USP, FAPESP, CAPES, CNPq, and IIEPAE-Einstein.
The online-only Data Supplement is available at http://stroke.ahajournals.org/lookup/suppl/doi:10.1161/STROKEAHA.111.628156/-/DC1.
- Received June 6, 2011.
- Revision received June 26, 2011.
- Accepted June 29, 2011.
- © 2011 American Heart Association, Inc.
- Romero JR,
- Beiser A,
- Seshadri S,
- Benjamin EJ,
- Polak JF,
- Vasan RS,
- et al
- Schulte-Altedorneburg G,
- Droste DW,
- Kollár J,
- Beyna T,
- Felszeghy S,
- Módis L,
- et al
- del Sol AI,
- Moons KG,
- Hollander M,
- Hofman A,
- Koudstaal PJ,
- Grobbee DE,
- et al