Discourse Profiles of Mothers of Children with ASD and Female FMR1 Premutation Carriers

Background: Subclinical pragmatic language differences comprise a core feature of the broad autism phenotype (BAP), believed to reflect genetic liability to ASD. Similar pragmatic language profiles among FMR1 premutation carriers (PM; Losh et al., 2012) suggest a potential role of the FMR1 gene in ASD-related pragmatic language (Lee et al., 2017). An important next step is to investigate linguistic contributions to these broader pragmatic differences. Discourse markers (e.g., filled pauses, backchannels) are words and nonverbal behaviors that manage conversations, promote turn-taking, and create rapport; even subtle deviations in their use can significantly influence the course of conversation. This study examined potential overlap in patterns of discourse marker use in mothers of individuals with ASD (M-ASD) and PM carriers, compared to controls, to investigate similarities in previously documented pragmatic differences in these groups.

Objectives: Characterize and compare patterns of conversational discourse marker use in mothers of individuals with ASD, female PM carriers, and controls, and evaluate links to global pragmatic language skills.

Methods: Eighty-seven M-ASD, 63 PM carriers, and 43 age- and IQ-matched female controls participated in semi-structured conversations with a trained examiner. Transcripts and videos of these conversations were coded for discourse marker use including speaker markers (i.e., lexical fillers, revisions, filled pauses) and listener markers (i.e., backchanneling, such as nodding). Global pragmatic language ratings were coded from video by independent, blind raters using the Pragmatic Rating Scale (PRS; Landa et al.,1992). Pragmatic language total- and factor scores (i.e., dominant or reticent) were included in analyses.

Results: PM carriers used significantly more lexical fillers and backchannels than the M-ASD group and controls (ps < .05), controlling for overall word count. Groups did not differ significantly in filled pauses or revisions (ps > .10). M-ASD and PM carriers demonstrated higher PRS total and factor scores (i.e., more difficulty) than controls (ps < .05). In both M-ASD and PM carriers, but not controls, higher PRS total scores related to fewer backchannels (r = -.36, p < .01; r = -.31, p< .05); greater reticence factor scores correlated with fewer revisions (r = -.31, p< .01; r = -.42, p < .05) and filled pauses (rs = -.31,-.36, ps < .01). In the PM group, higher PRS totals and dominant factor scores related to more frequent revisions (r = .33, p < .01; r = .23, p< .05). Increased lexical fillers related to unusual rate of speech (r = .33, p< .05) in the PM group.

Conclusions: This study suggests a complex influence of discourse markers on pragmatic language. Consistent with prior literature, overall pragmatic language profiles between M-ASD and PM carriers were similar (Losh et al., 2012); however, PM carriers employed some discourse markers at increased rates compared to the M-ASD group. In spite of differences in the frequency of their use, discourse markers influenced pragmatic language in similar ways for the PM and M-ASD groups, but not controls. This suggests an important, but subtle, role in conversational management through discourse marker use, potentially indicative of subclinical profiles related to ASD.