The Interplay of Language on Executive Functions for Children with ASD

Background: Impairment in use and development of language is a defining feature of autism spectrum disorders (ASD). Difficulties in executive functioning (EF) have also been consistently observed in individuals with ASD. A strong relationship exists between these two skill sets; in many cases, language operates as a self-regulatory function during problem solving and is thought to play a mediating role in EF. Use of inner or self-directed speech in particular may be a key component of EF. Findings on the association between language and EF, however, have been inconsistent. These inconsistencies merit additional examination of the link between language and EF. 

Objectives: The purpose of this study is to elucidate the specific EF profile of children with ASD by exploring the relationship between verbal and nonverbal skills on EF, using direct assessments and indirect parent and teacher reports.

Methods: 87 participants with ASD aged 8-12 years were administered assessments assessing EF (NEPSY-2, DKEFS), cognition (DAS, WISC), and language (CELF). Parents and teachers completed the BRIEF questionnaire (EF), the Vineland (adaptive functioning), and the CCC-2 (language).

Results: Significant correlations were found between measures of EF and both verbal and nonverbal measures, meriting further investigation into these relationships. As such, simple linear regressions were calculated between language and cognitive ability and measures of EF in four domains: working memory (WM), organization (O), shift (S), and inhibition (I). Direct assessment measures used were WISC Letter-Number Sequencing (WM), NEPSY Animal Sort task (O), DKEFS Number-Letter Switching (S), and DKEFS Color-Word Interference (I). Language skill, represented by DAS VIQ, was a significant predictor of WM (F(2,55)= 20.27, p < .05; R²= .424) and O (F(2,41)= 11.008, p < .01; R²= .349). Non-verbal cognitive ability, represented by WISC PRI, was a significant predictor of WM (F(2,55)= 20.27, p < .02; R²= .424) and S (F(2,50)= 3.881, p < .01; R²= .134).

Indirect reports of EF were likewise examined. Teacher report of adaptive language was a significant predictor of teacher report of WM (F(3,40)= 4.749, p < .05; R²= .263). No significant parent predictors were found. Further, a stepwise regression revealed that a significant predictor of direct WM assessment was language skill (F(1,29)= 20.027, p < .05; R²= .408) and parent report of language development (F(2,28)= 14.096, p < .05; R²= .502).

Conclusions: Language and nonverbal cognitive skills in individuals with ASD predicted different domains of EF. Verbal assessment predicted WM and O, whereas nonverbal assessment predicted WM and S. Indirect assessment of language skill substantiated the relationship between language skill and WM. The association of language and WM found in this study comports with the importance of self-talk in maintaining information in the phonological loop. Organization requires consolidating different parts into a whole, and language would intuitively be helpful in defining and categorizing a number of stimuli as part of this process. The present results indicate a profile of EF in individuals with autism that is mediated by language ability.  This association has important implications for both the diagnosis of ASD and for the development of treatment interventions.