Cerebral amyloid angiopathy pathology and cognitive domains in older persons

• Published in: Annals of neurology Vol. 69; no. 2; pp. 320 - 327
• Main Authors: Arvanitakis, Zoe, Leurgans, Sue E., Wang, Zhenxin, Wilson, Robert S., Bennett, David A., Schneider, Julie A.
• Format: Journal Article
• Language: English
• Published: Hoboken Wiley Subscription Services, Inc., A Wiley Company 01.02.2011; Wiley-Liss; Wiley Subscription Services, Inc
• Subjects: Aging - pathology; Amyloid - metabolism; Biological and medical sciences; Brain - metabolism; Brain - pathology; Brain - physiopathology; Cerebral Amyloid Angiopathy - metabolism; Cerebral Amyloid Angiopathy - pathology; Cerebral Amyloid Angiopathy - physiopathology; Cognition - physiology; Cognition Disorders - metabolism; Cognition Disorders - pathology; Cognition Disorders - physiopathology; Dementia - metabolism; Dementia - pathology; Dementia - physiopathology; Humans; Linear Models; Longitudinal Studies; Medical sciences; Memory; Neurology; Neuropsychological Tests; Pathology; Religious orders; Severity of Illness Index; Vascular diseases and vascular malformations of the nervous system; Anatomic pathology; Cerebral amyloid angiopathy; Nervous system diseases
• ISSN: 0364-5134; 1531-8249; 1531-8249
• DOI: 10.1002/ana.22112

Objective To examine the relation of cerebral amyloid angiopathy (CAA) to cognitive domains in older community‐dwelling persons with and without dementia. Methods Subjects were 404 persons in the Religious Orders Study, a cohort study of aging, who underwent annual clinical evaluations, including 19 neuropsychological tests from which 5 cognitive domain and global summary scores were derived, and brain autopsy at time‐of‐death (mean age‐at‐death 86). Using amyloid‐β immunostaining, CAA severity was graded in 5 regions (midfrontal, inferior temporal, angular, calcarine, and hippocampal cortices), as 0 = none, 1 = mild, 2 = moderate, 3 = severe, and 4 = very severe. Because severity was related across regions (all rs > 0.63), and almost all persons had some CAA, we averaged regional CAA scores and created class variable predictors for no‐to‐minimal (<0.5), mild‐to‐moderate (0.5‐2.5) and moderate‐to‐very severe CAA (>2.5). Results CAA was very common (84.9%; 94 had no‐to‐minimal, 233 mild‐to‐moderate, and 76 moderate‐to‐very severe disease) and was related to AD pathology (rs = 0.68). In linear regression analyses controlling for age, sex, education, AD pathology, infarcts, and Lewy bodies, moderate‐to‐very severe CAA was associated with lower perceptual speed (p = 0.012) and episodic memory (p = 0.047), but not semantic memory, working memory, visuospatial skills, or a composite of all cognitive measures. No associations of mild‐to‐moderate CAA with cognition were found. Dementia did not modify these findings. Interpretation CAA pathology is very common in older community‐dwelling persons and is associated with AD pathology. Moderate‐to‐very severe CAA, but not mild‐to‐moderate CAA, is associated with lower performance in specific cognitive domains, most notably perceptual speed, separately from the effect of AD pathology. ANN NEUROL 2011