Executive Function and Symptom Expression Among High Functioning Children with ASD

Background: It has been hypothesized that some symptoms of autism may be related to prefrontal dysfunction. Yet ASD symptom severity has not been consistently correlated with executive function (EF) ability (e.g., Landa & Goldberg, 2005; Ozonoff & McEvoy, 1994; Ozonoff et al., 2004). A relation between social communication skills and tasks related to executive function was found in preschoolers with autism and controls (McEvoy et al., 1993). However, Stahl & Pry (2002) presented evidence for a relation between shifting and joint attention only in typical controls, but not a group with ASD. 

Objectives: To test whether young, high-functioning children with ASD exhibit impairments in an EF task requiring set-shifting and flexibility, the Dimensional Change Card Sort Task (DCCS; Carlson, 2005; Frye et al., 1995). And, to test whether performance on this measure relates with symptom expression.

Methods: Subjects were 21 6 and 7-year-olds with ASD and 21 age and IQ-matched controls. Diagnosis was confirmed with the ADOS, ADI-R and DSM-IV-TR and additional assessment of symptom severity included the Social Skills Rating System (SSRS) and Repetitive Behavior Scale-Revised (RBS-R). All children in both groups had cognitive ability in the average to above average range. The DCCS was administered to assess EF. Children were first presented with a sorting dimension, and then the sorting dimension was switched. Finally, both dimensions were combined and an extra-dimensional symbol indicated the dimension by which to sort.

Results: Groups did not differ in the number of items correct during the post switch phase, t(17) = -1.37, p = .19, or the proportion of passes during the extra-dimensional phase, χ² (1, N = 38) = 2.62, p = .11. Yet, the number of correct sorts during the extra-dimensional phase made by each group approached significance, t(36) = -1.926, p = .06. With groups combined, the SSRS standard score related with the number of correctly sorted cards in the extra-dimensional phase, r(39) = .505, p = .001. Among children with ASD only, the score for extra-dimensional sorting related to the SSRS standard score, r(19) = .558, p = .009, and the Responsibility subscale, r(19) = .543, p = .011. A significant inverse relation between the total number correct during the extra-dimensional phase and the ADOS Social Affect score, r(19) = -.48, p = .03 was also detected. Children with greater symptom scores in the Social Affect ADOS domain had lower accuracy in extra-dimensional sorting during the DCCS. The DCCS did not relate with the RBS-R.

Conclusions: Groups did not significantly differ on the DCCS and children with ASD were not universally impaired on this task. Yet, this measure of basic EF directly related with observed ADOS Social Affect symptoms and overall social function measured by parent report on the SSRS. This suggests that by early school age, EF ability is related to social communication functioning in non-cognitively impaired children with ASD. These findings have implications for understanding the neural substrates of social impairment in ASD.