Episodic Memory Decline Predicts Cortical Amyloid Status in Community-Dwelling Older Adults

Article type: Research Article

Authors: Darby, David G.^{a; b; c; d; e; *} | Brodtmann, Amy^{b; f} | Pietrzak, Robert H.^g | Fredrickson, Julia^h | Woodward, Michael^e | Villemagne, Victor L.ⁱ | Fredrickson, Amy^a | Maruff, Paul^{a; c} | Rowe, Christopherⁱ

Affiliations: [a] CogState Ltd, Melbourne, Australia | [b] Florey Neuroscience Institutes, Carlton South, VIC, Australia | [c] Centre for Neuroscience, University of Melbourne, Parkville, VIC, Australia | [d] Mental Health Research Institute, Parkville, VIC, Australia | [e] Austin Health, University of Melbourne, West Heidelberg, VIC, Australia | [f] Boxhill and District Hospital, Boxhill, VIC, Australia | [g] National Center for Post-traumatic Stress Disorder, VA Connecticut Healthcare System and Department of Psychiatry, Yale University School of Medicine, New Haven, CT, USA | [h] Deakin University, Burwood, VIC, Australia | [i] Department of Nuclear Medicine and Centre for PET, Austin Health, Heidelberg, VIC, Australia

Correspondence: [*] Correspondence to: A/Prof. David G. Darby, c/- CogState Ltd, Level 2, 255 Bourke St, Melbourne VIC 3000, Australia. Tel.: +613 9664 1300; Fax: +613 9664 1301; E-mail: ddarby@cogstate.com.

Abstract: Intra-individual decline in memory and cognition is characteristic of prodromal Alzheimer's disease (AD) and may allow detection of very early AD pathology. Episodic memory task scores on a brief computerized cognitive battery (CogState) were prospectively evaluated at baseline, and 3-, 6-, 9-, 12-, and 24-months post-baseline. Linear mixed models were conducted to compute age-adjusted slopes. Subjects with slopes declining ≥90th percentile (“memory decliners”) and age- and gender-matched subjects without such decline (“non-decliners”) were studied with clinical, neuropsychological, and neuroimaging evaluations. Of 195 who completed 24-month evaluation (age 51 to 80 years), 15 memory decliners (mean age 62.7 years, SD 7.6) were identified, and matched with 33 non-decliners (mean age 63.3 years, SD 8.2). Amyloid-PET imaging was qualitatively abnormal with excess cortical amyloid accumulation in 7 memory decliners (46.7%) and 4 (12.1%) non-decliners (odds ratio 6.34), and quantitatively abnormal with standardized uptake value ratios >1.4 in 5 memory decliners (33.3%) and 2 (6.1%) non-decliners (odds ratio 8.3). One of the memory decliners and none of the non-decliners fulfilled criteria for mild cognitive impairment, but the groups did not differ with respect to subjective memory impairment, neuropsychological evidence of episodic memory impairment, or MRI imaging abnormalities. Intra-individual decline in episodic memory can be detected using a brief computerized cognitive performance test optimized to detect change in community-dwelling non-demented older persons and appears predictive of the presence of cerebral amyloid in about half of these persons. This approach may help detect early prodromal AD pathology in wider-scale community screening programs.

Keywords: Alzheimer's disease, cognition, cognitive decline, memory, pre-clinical diagnosis

DOI: 10.3233/JAD-2011-110818

Journal: Journal of Alzheimer's Disease, vol. 27, no. 3, pp. 627-637, 2011