Modifications to the CVLT-II Recognition Discriminability Indices to Enhance the Characterization of Recognition Memory Impairment in Healthy Aging and Neurodegenerative Disease
Rationale: Although studies using the California Verbal Learning Test (CVLT) have shown that recognition memory declines with age, age-related differences on more nuanced aspects of recognition have not been explored in detail. CVLT research also has shown that Alzheimer’s disease (AD) is associated with impaired recall and recognition that reflect an encoding deficit stemming from neuropathology in medial temporal cortices. Huntington’s disease (HD) is also associated with impaired recall but, in contrast to AD, in early stages shows less compromised recognition, reflecting primarily a retrieval deficit due to frontal-system dysfunction. Although original CVLT research demonstrated that recognition deficits are less severe in HD than in AD, more recent CVLT-II evidence has yielded mixed findings, highlighting the need for more refined measures to elucidate the extent of recognition memory impairment in AD and HD and thereby enhance characterizations of the profiles of memory loss associated with these neurodegenerative conditions.
Method and Results: Study 1 examined whether AD and HD differences on the CVLT-II Total Recognition Discriminability (RD) index (which assesses the ability to distinguish List A [original word list] targets from all distractors on the Yes/No Recognition trial) varied across nonparametric (used in CVLT) and parametric (used in CVLT-II) calculations of Total RD. Comparisons of group differences on nonparametric versus parametric Total RD indices could only be done in the context of raw scores (the CVLT-II does not by default include a nonparametric Total RD index, and no standardized nonparametric Total RD scores were therefore available), and the magnitudes of group differences were not shown to significantly differ across nonparametric and parametric methods. However, in contrast to what was observed in prior CVLT studies, analyses indicated that relative to the AD group, the HD group exhibited comparable standardized parametric Total RD scores (despite higher raw nonparametric and parametric Total RD scores [after accounting for demographic factors when appropriate]). A potential explanation for the discrepancy in performance of individuals with HD on the CVLT versus CVLT-II recognition trials was that an increased proportion of prototypical distractors (i.e., those that are semantically related to targets) on the CVLT-II relative to the CVLT disproportionately amplified the difficulty of the recognition trial for individuals with frontal-system dysfunction (e.g., HD). Study 2 investigated the utility of more refined RD indices generated to parse the degree of semantic association between targets and distractors (on Source and Novel RD indices that assess the ability to distinguish List A targets from List B [interference word list] and novel distractors, respectively) in characterizing nuanced aspects of yes/no recognition memory in healthy older and young adults. Although older adults performed worse than young adults on all RD indices, age group differences were smaller on more refined RD indices that excluded semantically-related distractors in RD calculations. Building on findings from Studies 1 and 2, Study 3 examined performance in individuals with AD and HD on Total RD and List A vs. Novel/Unrelated RD indices. The latter is a new index that isolates the ability to distinguish List A targets from distractors that are novel and semantically unrelated to targets. The List A vs. Novel/Unrelated RD index is therefore less subject than the Total RD index to source memory difficulties and semantic confusion, both of which are often seen in individuals with frontal-system dysfunction (e.g., HD), and may yield more purified assessments of yes/no recognition memory in HD. Analyses indicated that the List A vs. Novel/Unrelated RD index yielded more robust AD versus HD differences than the Total RD index, providing greater differentiation between individuals whose memory disorder is primarily at the encoding/storage level (e.g., AD) versus at the retrieval level (e.g., early HD).
Relevance: Given the expanding older population and that memory loss is a hallmark feature of cognitive impairment in both healthy aging and neurodegenerative disease, more refined memory indices are needed to elucidate the nature of memory changes that may be associated with normal aging and neurodegenerative processes. Collectively, Studies 1, 2, and 3 utilized innovative psychometric approaches with the CVLT-II to explore age-related differences on more nuanced aspects of yes/no recognition memory, as well as elucidate the extent of yes/no recognition memory impairment in AD and HD.