Reduced visual associative learning is linked to Alzheimer’s disease pathology

Abstract

AbstractBackgroundReduced learning is an early sign of memory impairment and relevant in the context of Alzheimer’s disease. The Visual Association Test (VAT) is a brief visual associative memory task comprising two sets (A and B) of six interactive pictures (e.g., a monkey holding an umbrella). We hypothesized that the VAT may be able to elicit learning deficits on set B when administered directly following set A, particularly in patients with Alzheimer’s disease‐related episodic memory deficits. We related learning on the VAT to medial temporal lobe atrophy (MTA) and amyloid‐β positivity.MethodWe selected 3,599 patients (63.9±8.9 years, 41% female) who completed two VAT sets. Patients had subjective cognitive decline (n = 1,369, 38%), mild cognitive impairment (n = 756, 21%), Alzheimer’s disease dementia (n = 856, 24%), or other types of dementia (n = 618, 17%). Average MTA scores of both hemispheres from magnetic resonance imaging were available for all; amyloid‐β status was obtained from cerebrospinal fluid or positron emission tomography for n = 2,769 (77%). We modeled learning curves on both VAT sets using linear mixed models, including an interaction with amyloid‐β status. Using logistic regressions, we analyzed the relationship between learning of associations on both sets (either all associations on both sets, or all associations on the first, but not the second set) and MTA, as well as amyloid‐β status.ResultMost patients learned all associations on both sets (n = 2,442, 68%). Patients who learned all associations on the first, but not the second set (n = 640, 18%) were more likely to have higher MTA scores and to be amyloid‐β positive than those who learned all associations (Table 1). Figure 1 shows learning curves on both sets of the VAT, stratified by amyloid‐β status. Less steep learning curves seem to suggest that amyloid‐β positive patients benefited less from completing a first set than amyloid‐β negative patients.ConclusionWe observed reduced learning on the second set, with those who did not learn all associations on both sets having more medial temporal lobe atrophy and more often being amyloid‐β positive. The VAT may be a useful, simple tool to assess early episodic memory deficits in the presence of Alzheimer’s disease pathology.