Profiles of Pragmatic Language in Individuals with ASD and Their Parents

Background: Pragmatic (i.e., social) language abilities are universally impaired in individuals with autism spectrum disorder (ASD; Tager-Flusberg, 2002). Previous research indicates that parents of individuals with ASD may demonstrate subtle differences in pragmatic language that reflect genetic risk and comprise a component feature of the broad autism phenotype (BAP; Landa, 1991, 1992; Losh et al., 2008, 2012). This study aimed to build on these findings by examining parallel pragmatic language features in parents and children during semi-naturalistic conversation, using the Pragmatic Rating Scale (PRS; Landa, 1992) and Pragmatic Rating Scale-School Age (PRS-SA; Landa, 2011). We employed factor analyses in each sample to identify whether similar pragmatic language violations were observed in individuals with ASD and their parents. Further, we examined a range of potential phenotypic correlates in each group.

Objectives: To define potentially overlapping profiles of pragmatic language in individuals with ASD and their parents (with and without the BAP), and to explore correlated skills.

Methods: The PRS-SA and PRS consist of similar, operationally-defined pragmatic language features rated by blind coders from videotaped semi-structured conversational tasks (ADOS, and Life History Interview, respectively). To identify factors comprising pragmatic language characteristics of each group, Principal Component Analyses were completed on PRS-SA/PRS data from 53 school-aged individuals with ASD (IQ>80, age ≤19) and 186 parents of individuals with ASD. PRS-SA/PRS data were also included from controls (22 children and 60 parents) to examine group differences and related correlates. Analyses controlled for differences in IQ and chronological age. Given prior research demonstrating links between social cognition and pragmatic language use (Losh & Piven, 2007), all participants completed a social cognition battery. BAP status was determined using the Modified Personality Assessment Scale (Tyrer, 1988).

Results: Findings indicated a two-factor model of pragmatic language in parents of individuals with ASD, characterized by Reserved and Dominant conversational features. A three-factor model emerged in individuals with ASD, which included nearly identical profiles as their parents plus an additional Disinhibited factor. Individuals with ASD scored significantly higher than controls and BAP+ parents scored significantly higher than BAP- and control parents across factors. In the ASD group, Reserved scores positively correlated with ADOS severity overall (r = .47) and social affect severity (r = .61), whereas Dominant (r = .54) and Disinhibited (r = .44) styles were positively associated with severity of restricted and repetitive behaviors. Finally, higher Dominant scores were negatively associated with social cognitive ability in parents and their children with ASD (rs > .3).

Conclusions: To our knowledge, this is the first study to employ factor analysis to identify overlapping pragmatic language profiles in individuals with ASD and their parents. Results suggest that qualitatively similar types of pragmatic language differences characterize ASD and the BAP, although these features are very subtly expressed in parents. Further, factors related to different aspects of ASD severity, and social cognition emerged as key correlates of pragmatic language across groups. Together, findings highlight pragmatic language as a valuable marker of genetic risk.