Frontiers in Psychology (Apr 2015)
Investigating the possibility of a syntactic impairment in the semantic variant of PPA using a constrained production task: Preliminary findings
Abstract
In the semantic variant of primary progressive aphasia (svPPA), syntactic skills are generally thought to be preserved, while in the non-fluent variant (nfvPPA) syntactic impairment is a core diagnostic feature (Gorno-Tempini et al., 2011). There are, however, some indications in the literature that syntactic processing may not be entirely normal in svPPA. Most studies of syntactic production in svPPA have used unconstrained tasks and have found no syntactic impairment (e.g., Bird et al., 2000; Kave et al., 2007). In the two published studies that have found a syntactic impairment in svPPA, one used a constrained task (Benedet et al., 2006), and the other (Meteyard & Patterson, 2009) did not. However, the authors of the latter article suggested that the observed syntactic errors were subtle. They also suggested that a syntactic impairment in svPPA might not be observed in spontaneous language samples due to an overreliance on simpler structures. In the current study, we used a constrained sentence production task to compare the syntactic abilities of individuals with nfvPPA, svPPA and healthy controls longitudinally, to investigate the existence of a syntactic impairment in the different PPA variants. We predicted that by using a constrained task we would observe a syntactic impairment in both variants of PPA. We tested 18 participants with nfvPPA, 13 with svPPA and 23 control participants. They were tested up to three separate times, with approximately one year between sessions. Groups were matched on age and years of education. The patient groups were matched on Mini Mental State Examination score (Folstein, Folstein & McHugh, 1975) and estimated time post onset of initial symptoms, but the nfvPPA group scored higher than the svPPA group on the Boston Naming Test (Kaplan, Goodglass, & Weintraub, 2001). We used the sentence production task from Caplan and Hanna (1998) to elicit active, passive, dative and dative-passive sentences. A mixed ANOVA (Group X Sentence Type) was run for three different measures of sentence production: accuracy on overall production, accuracy on verb morphology and ‘overall grammaticality’ (a measure we devised to reflect a sentence’s grammaticality, irrespective of its semantic content), for each of the three time points. Post hoc testing was performed using the Tukey-Kramer test. For all main effects and interactions, a p-value of less than 0.05 was considered significant. We found that both the nfvPPA and svPPA groups showed a syntactic impairment in this constrained sentence production task. However, the groups demonstrated a different pattern and progression of impairment, with the syntactic impairment initially being generally more severe and pervasive for participants with nfvPPA compared to participants with svPPA. Interestingly, the svPPA group demonstrated a syntactic impairment in the analysis of accuracy of verb morphology that was not observed in the analysis using the less stringent ‘overall grammaticality’ measure. This difference may mirror the differences observed when using constrained versus unconstrained tasks. Overall, the findings from this study contribute important information regarding the nature and progression of the language production impairment in the non-fluent and semantic variants of primary progressive aphasia.
Keywords
