Do semantic clustering deficits underpin long-term memory binding impairments in prodromal AD?

Background: The Free and Cued Selective Remind Test (FCSRT) was developed to measure the ability to bind semantic categories and exemplars and use such bindings as cues to access associative memory representations. However, the variable informing about such ability (i.e., sensitivity to cuing) does not fare better in the early detection of memory impairments in AD than variables measuring free recall. We investigated whether this is due to limited construct validity of the FCSRT. We used as a gold standard a well-known test of semantic clustering functions (i.e., Hopkins Verbal Learning Test - HVLT). We assessed patients with Mild Cognitive Impairment (MCI) due to Alzheimer’s disease (AD) with both tests. Methods: Of 70 MCI patients who underwent two neuropsychological assessments a year apart, 12 converted to AD. We compared baseline performance of MCI converters and non-converters with that of healthy controls on the FCSRT, HVLT, and other traditional neuropsychological tests. We ran correlation analyses and tests of mean (ANOVA) to investigate if impaired semantic clustering functions, as measured by these tests, correlated in MCI patients and if MCI converters were more impaired at baseline than non-converters.Results: MCI patients were in advanced stages from baseline (i.e., multi-domain amnestic MCI). MCI converters and non-converters showed a similar level of impairment at baseline in cognitive and functional abilities compared to controls and to each other. Relative to controls both groups showed significant impairments on all the variable from the FCSRT and the HVLT. Such impairments were larger in MCI converter than in non-converter. Sensitivity to cuing correlated with immediate and delayed recall variables form the HVLT which inform on sematic clustering abilities, thus suggesting shared underpinnings. Conclusions: Sematic clustering abilities do underpin performance on the FCSRT. Taken together the results from this study and those reported in earlier studies, it seems plausible to suggest that MCI due AD will impair this function only in the advanced prodromal stages. Future studies should investigate why semantic cuing, as assessed by the FSCRT, remains relatively preserved in the early stages of AD and the implications that this may have for the early detection of dementia.