Unique Social Cognitive Profiles in Autism and Fragile X Syndrome: Influences on Pragmatic Language

Background: Social cognition refers to the ability to infer thoughts and emotions in others and includes skills such as intentionality, perspective-taking, and false belief understanding. Social cognition is impaired in ASD, and has been implicated in the pragmatic impairments that define ASD. Social cognitive impairments have also been documented in fragile X syndrome with co-occuring ASD (FXS-ASD), and similarly appear related to this group’s pragmatic language deficits (Losh et al., 2012). Importantly, the majority of past work has tended to analyze summary scores or single instruments of social cognition, which do not capture different types of social cognitive skills and their discrete impact on pragmatic language.

Objectives: This study used a series of social cognitive tasks ranging in difficulty (appropriate for age ranges in typical development from ~18 months-5 years) to compare social cognitive profiles across groups and examine the impact of these differences on pragmatic language in ASD and FXS-ASD.

Methods: Participants included children with ASD (29 boys) and FXS-ASD (46 boys, 15 girls). To determine the specificity of social cognitive patterns to ASD (idiopathic and syndromic), children with FXS only (13 boys, 27 girls), Down syndrome (20 boys, 22 girls), and typical development (19 boys, 18 girls) were also included. Tasks of perspective-taking (Slaughter et al., 2007) and 1^st order false belief (Matthews et al., 2003; Lewis and Mitchell, 1994) were administered. Pragmatic language was assessed from language samples obtained through the Autism Diagnostic Observation Schedule (ADOS; Lord, 2001) and analyzed with a detailed hand-coding system for contingent (on-topic) and perseverative (repetitive) language, which prior research has identified as areas of overlap between ASD and FXS-ASD (Martin et al., in prep; Roberts et al., 2007; Tager-Flusberg & Anderson, 1991).

Results: Individuals with ASD and FXS-ASD demonstrated a marked pattern of social cognitive performance in which they passed the 1^st order false belief task but failed the perspective-taking task. This pattern was very rarely observed in any of the other groups, and differentiated ASD and FXS-ASD from all other groups with 98% specificity. Participants with ASD and FXS-ASD who showed this pattern were also more perseverative and noncontingent in conversation than those who did not (ps<.05). Less contingent language in ASD and FXS-ASD, as well as more perseveration in FXS-ASD, were observed in those who failed the perspective-taking task compared to those who passed (ps<.05). Boys with ASD and FXS-ASD who failed false belief were more perseverative and less contingent, respectively, than those who passed (ps<.05).

Conclusions: A distinct pattern of social cognitive impairment was observed in both idiopathic and syndromic ASD, where children passed more advanced false belief tasks while failing a more basic task of perspective-taking. This pattern was generally not observed in the three other groups. This unique pattern of social cognitive difficulty also predicted shared pragmatic language impairments in individuals with ASD and FXS-ASD. Together, these data may help to define the role of social cognition in pragmatic language in ASD.