Slide Slam J2 Sandbox Series
Exploring Predictors of Naming Ability in Primary Progressive Aphasia
Katlyn Nickels1, Kindle Rising2, Alyssa Sachs3, Pélagie Beeson4; 1University of Arizona
INTRODUCTION: Anomia, or word-finding difficulty, is an early and universal feature of primary progressive aphasia, particularly in the logopenic (lvPPA) and semantic variants (sPPA), which disrupt phonological and semantic systems, respectively. Given that cognitive models of naming posit interactive processing of phonological and semantic networks to facilitate lexical retrieval, it follows that differential impairment of these underlying processes should be evident and predictive of naming performance. To examine this hypothesis, we examined the predictive value of phonological and semantic abilities, as well as a metric of speech production abilities, on naming accuracy in 22 individuals with primary progressive aphasia (PPA). To date, we have a sizeable cohort of individuals with the logopenic variant (n=18) , but the semantic cohort is not yet adequate (n=4). Preliminary findings raise some questions worthy of discussion. METHODS: The individuals with PPA reflected a range of aphasia severity (62.7-97.2 WAB AQ). Composite scores were computed for the following skills and compared to a group of age and education matched group of control subjects: a) Phonological awareness/manipulation skills using Arizona Phonological Battery, b) Semantic processing assessed using Pyramids and Palm Trees Test and subtests 47 and 48 of the Psycholinguistic Assessments of Language Processing in Aphasia, and c) Spoken repetition of words/nonwords as a proxy for speech production ability Multiple linear regressions were implemented to examine the predictive value of these three composite scores on naming ability as measured by the Boston Naming Test (BNT), Cambridge Naming Test, and the Naming composite score from the Western Aphasia Battery (WAB). RESULTS: The cohort with PPA was significantly impaired on phonological skills (51.0%), semantic processing (85.9%), and accuracy of speech production (86.0%). Linear regression models were significant for the prediction of naming performance on all three measures. Semantic skills emerged as a significant predictor for performance on the BNT (β =.534, p=.008), the Cambridge Naming Test (β =.786, p=.000), and the Naming composite of the Western Aphasia Battery (β =.740, p=.000). Conversely, phonological skills and speech production did not significantly predict performance on any of the naming measures, despite marked phonological impairment across the cohort. Separate regression models using only those with the lvPPA were not significant. CONCLUSIONS: Despite the disproportionate representation of individuals with lvPPA in our cohort and the overall impairment of phonological processing skills, our findings to date indicate that underlying semantic skills were a strong predictor of naming ability in individuals with PPA. This is in contrast to findings in stroke aphasia that demonstrate a strong predictive association between underlying phonological and semantic skills and naming. These preliminary results motivate continued investigation to better understand the relationship between these central language processes and naming ability in PPA.