Abstract
Background: Word class naming deficits are commonly seen in aphasia resulting from stroke (StrAph) and primary progressive aphasia (PPA), with differential production of nouns (objects) and verbs (actions) found based on StrAph type or PPA variant for some individuals. However, studies to date have not compared word class naming (or comprehension) ability in the two aphasic disorders. In addition there are no available measures for testing word class deficits, which control for important psycholinguistic variables across language domains. This study examined noun and verb production and comprehension in individuals with StrAph and PPA using a new test, the Northwestern Naming Battery (NNB; Thompson & Weintraub, experimental version), developed explicitly for this purpose. In addition we tested verb type effects, based on verb argument structure characteristics, which also is addressed by the NNB.
Method: A total of 52 participants with StrAph (33 agrammatic, Broca's (StrAg); 19 anomic (StrAn)) and 28 PPA(10 agrammatic (PPA-G); 14 logopenic (PPA-L); 4 semantic (PPA-S)) were included in the study. Nouns and verbs were tested in the Confrontation Naming and Auditory Comprehension subtests of the NNB, with scores used to compute noun to verb ratios as well as performance by verb type. Performance patterns within and across StrAph and PPA groups were then examined. The external validity of the NNB was also tested by comparing (a) NNB Noun Naming scores to the Boston Naming Test (BNT; Kaplan, Goodglass, & Weintraub, 2001) and Western Aphasia Battery (WAB-R; Kertesz, 2007) Noun Naming subtest scores, (b) NNB Verb Naming scores to the Boston Diagnostic Aphasia Examination (BDAE; Goodglass, Kaplan, & Barresi, 2001) Action Naming score (for StrAph participants only), and (c) NNB Comprehension subtest scores to WAB-R Auditory Comprehension subtest scores.
Outcomes & Results: Both agrammatic (StrAg and PPA-G) groups showed significantly greater difficulty producing verbs compared to nouns, but no comprehension impairment for either word class, whereas three of the four PPA-S participants showed poorer noun compared to verb production, as well as comprehension. However, neither the StrAn nor the PPA-L participants showed significant differences between the two word classes in production or comprehension. In addition, similar to the agrammatic participants, the StrAn participants showed a significant transitivity effect, producing intransitive (one-argument) verbs with greater accuracy than transitive (two- and three-argument) verbs. However, no transitivity effects were found for the PPA-L or PPA-S participants. There were significant correlations between NNB scores and all external validation measures.
Conclusions: These data indicate that the NNB is sensitive to word class deficits in stroke and neurodegenerative aphasia. This is important both clinically for treatment planning and theoretically to inform both psycholinguistic and neural models of language processing.