The Relationship Between Discourse Processing and Autistic Symptomatology in Adolescents with ASD

Background:  Discourse processing is integral to appropriate communication and social interactions, which are known to be impaired in individuals with an autism spectrum disorder (ASD) (APA, 2000). A recent study using the Pittsburgh Inference Test (PIT), a measure of discourse processing, reported impaired physical- and emotion-related inference making abilities in adolescents and adults with ASD (Bodner, Engelhardt, Minshew, & Williams, under review).  The relationship between discourse processing and more general ASD symptomatology and temperament, however, remains unclear.

Objectives:  The objective of the current study was to replicate and extend previously reported findings of impaired functioning on the PIT in adolescents with ASD.  The current study investigated the relationship between inference making abilities (as assessed by the PIT) and ASD symptomatology and temperament.

Methods:  A sample of 33 adolescents with ASD and 16 typically developing (TD) adolescents between the ages of 11 and 15 years (mean = 13.2 years), group-matched for age [t(47)=.158, p=.88] and IQ [FSIQ, VIQ, & PIQ, ts(47)<.956, ps>.34] participated.  Autism diagnosis was determined with the ADOS and/or ADI, and clinical impression. All participants attained Full Scale IQ’s > 70.  The PIT and parent questionnaires were administered, including the Social Responsiveness Scale (SRS), Social Communication Questionnaire Lifetime (SCQ-L), and Early Adolescent Temperament Questionnaire (EAT-Q).  The PIT consists of short stories and questions about typical life situations to which the participant is expected to give a verbal response. The PIT is comprised of questions designed to elicit responses that described physical relationships only or responses that describe physical, mental or emotional inferences (ToM). The responses for each story categorized as a physical or ToM response, and further categorized by type: emotion-ToM or other-ToM response.  

Results:  Hierarchical regressions confirmed previously reported findings that individuals with ASD performed worse on the PIT in comparison to TD participants, particularly in ascertaining physical and emotional states (pr²>.10, ts(45)>2.20, ps<.03), while other types of inference making abilities remained intact (pr²=.002, t(45)=.283, p=.78).  Furthermore, adolescents with ASD displayed improvements in performance with age for other-ToM responses (pr²=.13, t(42)=2.55, p=.02), but no improvement with age in physical or emotion-ToM responses (pr²<.02, ts(42)<.805, ps>.45).  Significant negative correlations were evident between all PIT variables (except physical responses) and SCQ-L (r range = -.683 to -.445, ps < .02).  In addition, significant negative correlations were evident between PIT variables (except physical and emotion totals) and SRS Social Cognition subdomain (r range = -.565 to -.400, ps < .03).  Finally, PIT variables (except physical and emotion totals) were positively correlated with EAT-Q Shyness subdomain (r range = .498 to .589, ps< .02).

Conclusions:  The current study replicated and extended previous findings that verbally able individuals with ASD display impairments in physical and emotion related inference making abilities.  These findings further highlight the utility of the PIT in teasing apart underlying difficulties in discourse processing.  In addition, poorer performance on the PIT appears to be related to more significant impairments in Social Cognition and ASD lifetime symptoms.  Furthermore, higher levels of shyness are related to better performance on the PIT.