Short term memory in aphasia: effects of modality and relationship with Western Aphasia Battery-R performance

Abstract

ABSTRACT Background Producing and comprehending language requires maintenance of linguistic units in short term memory (STM). While there is evidence of verbal STM impairments in persons with aphasia (PWA), there are unresolved questions about other aspects of STM in aphasia, particularly the role of input and output modalities on STM performance. This includes the extent to which visual STM abilities are impaired in PWA and if the modality of eliciting responses (verbal versus pointing) influences STM performance. A better understanding of the relationship between STM performance and clinical language assessments of PWA will inform test interpretation. Aims This study investigated three questions about short term memory (STM) performance in post-stroke aphasia: 1) whether visual STM is impaired in PWA given that a common neural network subserves encoding-recall of both verbal and visual STM, 2) whether the response modality (verbal vs pointing) impacts STM score, 3) the relationship between STM and language performance, specifically on the Western Aphasia Battery-Revised (WAB-R). Methods & Procedures Verbal and visuo-spatial performance on a STM battery was compared between 45 PWA and a representative neurotypical sample (N=40). For PWA, individual performance across STM subtasks was examined to identify dissociations and correlated with language performance on the WAB-R. Outcomes & Results PWA scored lower than neurotypical controls on verbal but not visual STM tasks. At the individual level, about one-half of PWA were impaired on the three verbal STM tasks, and one-third had dissociations among the three verbal STM tasks. There was no effect of response modality (verbal versus pointing) on verbal STM performance. There was a strong association between verbal STM and individual subtests of the WAB-R. In general, the relative phonological and semantic demands of the task (e.g. repetition) influenced the association strength with digit recall versus picture recall scores. Conclusions Verbal STM deficits are: 1) more prevalent than visual-spatial STM deficits in PWA, 2) strongly associated with WAB-R performance in a pattern that is consistent with the phonological and semantic demands of the tasks. The clinical implications of this study are that it is crucial to assess STM performance in PWA and examine its potential impact on language assessments. Further, a verbal or pointing response modality can be used for STM measurement without negatively impacting the score (unless there are clear motoric contra-indications). This study provides performance data for a set of aphasia-friendly STM tasks that can serve as future reference.