Understanding conversation breakdown in Primary Progressive Aphasia

Abstract

Primary Progressive Aphasia (PPA) is a clinical syndrome caused by neurodegenerative disease in which language impairment is the initial and most prominent symptom (Mesulam, 1982). Currently, three variants of PPA are recognized (Gorno-Tempini et al., 2011); the non fluent/agrammatic variant (nfvPPA), the logopenic variant (lvPPA) and the semantic variant (svPPA). Difficulty in conversing is the primary complaint of people with PPA and their carers (Knibb, Woollams, Hodges, & Patterson, 2009). Although much is known about cognitive and linguistic impairments in PPA and their neural and pathological bases, little attention has been focused on the nature of conversation breakdown in PPA. Therefore this thesis describes two studies that have investigated trouble and repair behaviours during conversations of people with primary progressive aphasia. In both studies ten minute video recorded conversations between people with PPA and their most frequent conversation partner were examined. In the first study participants were three people with PPA; one with lvPPA and two with nfvPPA. Participants in the second study were seven people with svPPA. Participants in both studies were matched to healthy controls and all participants engaged in conversation with their most frequent communication partner. Conversations were investigated for productivity measures such as words, words per minute and words per turn (Herbert, Hickin, Howard, Osborne, & Best, 2008). Then the conversations were investigated for trouble and repair behaviours using a method described by Watson and colleagues in their study of conversation in people with Dementia of the Alzheimers Type (DAT) (Watson, Chenery, & Carter, 1999). In the first study individuals with nfvPPA and lvPPA had reduced mean length of turn but maintained their share of turn taking. They demonstrated a variety of trouble indicating behaviours rather than the non-interactive repairs that have been observed in studies of conversation in DAT. The partners bore the greater burden of highlighting trouble and need for repair and used interactive trouble indicating behaviours. The individuals with PPA used repair behaviors and demonstrated success in restoring conversation and social interaction. In the second study individuals with svPPA were found to be active participants in conversation generally taking an equal proportion of turns. As a group, the people with svPPA had good output and this was consistent with Wilson et al’s (2010) conclusion that apart from difficulty accessing lexical content, the connected speech production of this group is preserved. There was a wide variety of trouble and repair behaviours observed across the seven dyads, but overall the patterns of trouble and repair observed were unlike the patterns observed in people with nfvPPA and lvPPA, and also unlike those seen in DAT and there was a wide variety of trouble and repair behaviours observed across the seven dyads. With more significant cognitive impairment, there was a trend for greater non-interactive trouble indicating behaviours as has been observed in DAT. Finally, severity of naming impairment and the contribution of the communication partner appeared to be the most critical factors for the success of conversational repair. Examination of trouble and repair in conversation in PPA may enable the clinician to develop and deliver effective behavioural interventions to this clinical group that will improve communication and thereby promote positive social relationships and the maintenance of independence in daily life. The heterogeneity of individual conversational behaviours in these two studies demands individually tailored intervention. The studies also suggest important broad differences between variants in the types of intervention likely to be required. Understanding the nature of conversation breakdown in PPA may improve the effectiveness of behavioural interventions and thereby promote quality of life for individuals with PPA and families.