Social Cognition Contributes to Social Functioning in Adults with Autism

Background: Adults with Autism Spectrum Disorder (ASD) are characterized by poor social functioning that contributes to smaller social networks, occupational challenges, and reduced quality of life. Identifying mechanisms of this poor social functioning offers promise for treatment and intervention.

Objectives: We examined the contribution of social cognitive ability to social functioning above and beyond general intellectual capacity and demographic features. Although social functioning has been most commonly evaluated in adults with ASD using self- and informant-report questionnaires, here we used performance-based measures that offer a more direct assessment of real-world social and functional abilities.

Methods: Adults with ASD (n=103; Age: M=24.28; IQ: M=105.64) completed a broad assessment of social cognitive ability: 1) emotion recognition tasks (Penn Emotion Recognition Task (ER-40) and the Bell Lysaker Emotion Recognition Task (BLERT)); 2) social appraisal tasks (The Awareness of Social Inference Task (TASIT), Reading the Mind in the Eyes, Cartoon Theory of Mind (CToM), and the Hinting task); 3) social perception tasks (Relationships Across Domains (RAD) and the Benton Facial Recognition Test); 4) an attribution style task (AIHQ); and 4) the Trustworthiness task.

Participants also completed two performance-based social functioning measures: 1) the Social Skills Performance Assessment (SSPA), a naturalistic role-play task enacting two real-life situations (e.g., meeting a new neighbor, negotiating with a landlord) later coded for discrete social skills (e.g., eye gaze, affect) and an overall index of social ability; and 2) the UCSD Performance-Based Skills Assessment (UPSA) assessed financial skills (e.g., counting change) and communication skills (e.g., scheduling appointments).

Results: A hierarchical regression was conducted for each outcome, entering general cognitive and demographic characteristics (e.g., Age, Gender, IQ, and Race) in the first block, followed by all social cognitive tasks in the second block. These predictors accounted for 52% of the variance in UPSA scores (F(17, 81)=5.13, p<.001), with cognitive and demographic information accounting for 33% of the variance (p < .001) and social cognitive tasks accounting for 19% (p=.006). Age and IQ, but not gender and race, were positively related to social functioning scores (p’s<.014). For social cognitive tasks, only the CToM individually predicted higher social functioning (p=.04), and no other tasks were significant predictors (p’s>.10).

Predictors together accounted for 29% of the variance in SSPA scores (F(17,75)=1.80, p=.04), however each block did not uniquely predict a significant proportion of variance (p’s>.08). Higher TASIT and Trustworthiness scores significantly predicted higher social skills scores (p’s<.03), and neither the demographic variables (p’s>.14) nor the other social cognitive variables (p’s>.26) significantly predicted social skills ability.

Conclusions: Social cognition contributes to social functioning for adults with autism over and above other characteristics including IQ and demographic variables. However, our results suggest different abilities may uniquely contribute to social skill and functional ability: general cognition and social appraisal were predictive of functional ability, but only social appraisal and trustworthiness performance were related to better social skills. Taken together, this may suggest that treatment of functional skills, but not social skill, may benefit from focusing on general cognitive ability in addition to specific social cognitive components.