Comparing the short-term memory binding test and RAVLT as predictors of hippocampal atrophy

Cecchini, MA and Foss, MP and Tumas, V and Patrocinio, F and Correia, RDC and Novaretti, N and Brozinga, TR and Bahia, VS and de Souza, LC and Guimarães, HC and Caramell, P and Lima-Silva, TB and Cassimiro, L and Brucki, SMD and Nitrini, R and Sala, Sergio Della and Parra, MA and Yassuda, MS (2019) Comparing the short-term memory binding test and RAVLT as predictors of hippocampal atrophy. In: Alzheimer's Association Satellite Symposium, 2019-04-10 - 2019-04-12, Renaissance Sao Paulo.

[img]
Preview
Text (Cecchini-AASS-2019-Comparing-the-Short-term-Memory-Binding-test-and-RAVLT-as-predictors)
Cecchini_AASS_2019_Comparing_the_Short_term_Memory_Binding_test_and_RAVLT_as_predictors.pdf
Accepted Author Manuscript

Download (254kB)| Preview

    Abstract

    BACKGROUND: The Rey Auditory Verbal Learning Test (RAVLT) assesses the long-term verbal episodic memory, while the short-term memory binding (STMB) tests assess conjunctive memory binding. In the STMB, participants should remember the integration of shapes (or objects) and colors, forming a unique representation in memory. OBJECTIVE: The objective of this study was to compare the STMB and the RAVLT as predictors of hippocampal atrophy. METHODS: All participants underwent neuropsychological assessment and MRI data were collected. Participants were 17 patients with mild cognitive impairment (MCI) and 12 patients with Alzheimer’s disease (AD). All participants performed the RAVLT test and two different paradigms of the STMB test: change detection and free recall. In the change detection task, patients need to recognize if there was a difference in shapes and colors (unbound) or shape-color integrations (bound) between two consecutive screens. In the free recall task, patients were asked to recall aloud objects and colors individually (unbound condition) or object-color integrations (bound condition) that they had just seen in a screen. The groups were compared using ANCOVA analyses, and regression analyses were used to evaluate which cognitive paradigm better predicted the hippocampal atrophy. RESULTS: 10 patients showed no hippocampal atrophy (all were MCI patients) and 19 had atrophy (7 MCI and 12 AD). The group with atrophy was older and showed worse performance in the cognitive tasks. The regression model using the atrophy (positive or negative) as the dependent variable, and the STMB tasks, RAVLT delayed score, age and education as the predictors, indicated that only the STMB change detection bound task was retained in the model (R = 0.517, R2 = 0.268, p = 0.004), explaining 26,8% of the variance. Present findings suggest that the change detection STMB task may be a better marker of neurodegeneration than the other tests.