Impairments in Real World Executive Function Increase From Childhood to Adolescence in Autism Spectrum Disorders

Background:  Executive dysfunction is frequently associated with autism spectrum disorders (ASDs) and has been linked to reduced adaptive functioning (Gilotty et al., 2002) and greater autism symptomatology (Kenworthy et al., 2009).  Understanding the trajectory of EF development in ASDs can offer key insights into the dynamic nature of the disorder, inform treatment options, and help frame research findings of abnormal neuroanatomical growth (e.g., Courchesne, Campbell, & Solso, 2011; Wallace et al., 2010; Farley et al., 2009).  While several studies have investigated developmental trajectories of EF in ASD using lab-based tasks, no study to date has directly measured how these skills change over time in an ecologically-sensitive manner.  

Objectives: The current study seeks to extend prior work by evaluating age-related changes in everyday EF during childhood and adolescence in a large cohort of children with ASDs.  We hypothesize that children with ASD will show divergence in EF skills over time compared to normative samples as indicated by increasingly impaired scores on the Behavior Rating Inventory of Executive Function (BRIEF; Gioia et al., 2000) subscales.

Methods: 260 children with an ASD but without Intellectual Disability participated in the study. Subjects were divided into four groups based on age (5-7; 8-10; 11-13; and 14-18 year olds), consistent with the age subdivisions described in the original BRIEF factor analysis paper. The four age groups did not differ in IQ, sex or autism severity ratings. A mixed-model ANOVA was run to investigate differences in BRIEF subdomains across age groups.

Results: Analyses revealed a significant interaction effect of BRIEF subdomain by age group (p=.015).  After post-hoc false discovery rate adjustment, significant age effects (i.e., worsening scores with increasing age) for the Initiate (p<.001), Working Memory (p=.011), and Organization of Materials (p=.016) subscales, and a trend toward significance for the Shift subscale (p=.038), were found.

Conclusions:  Findings generally confirmed the hypothesis that older children with ASD would show greater EF problems compared to the normative sample than younger children with ASD.  Specifically, despite being equated for IQ, sex, and ASD symptoms, the older children had more impaired scores on most subdomains in the Metacognitive Index of the BRIEF.  Findings support a general worsening, or divergence from the normative sample, in metacognitive abilities in children with ASD as they age, and have important implications for the challenges faced by high functioning adolescents and young adults with ASD as they attempt to enter mainstream work and social environments.