Developmental Trajectory of Theory of Mind Abilities in Children with Autism Spectrum Disorder

Background:  Although theory of mind (ToM) impairments in children with Autism Spectrum Disorder (ASD) have been widely researched, the developmental trajectory of these skills is unclear. Research has indicated that ToM ability in children with ASD may follow a similar course of development, but at a slower rate compared to neurotypical (NT) children (Steel et al., 2003). However, there is evidence that children with ASD may have a unique trajectory where skills progress in a different sequence (Peterson et al., 2012) or even plateau (Ozonoff & McEvoy, 1994). These conflicting findings may be due to variance in children’s linguistic abilities and task complexity.

Objectives:  Using the Developmental Trajectory approach (Thomas et al., 2009), we sought to understand (1) whether differences in performance on a range of ToM skills (first-order to higher-order) between children with and without ASD are due to a delay in onset and/or a difference in the rate of development, and (2) whether these mentalizing abilities are better explained by age or language ability.

Methods: This study compared the ToM ability of 19 children with ASD to 26 NT children (7-12 years of age). Children completed a battery of ToM measures, including: (1) the Unexpected Contents Task (first-order false belief; Perner et al., 1989), (2) Birthday Puppy Story (second-order false belief; Sullivan et al., 1994), (3) Strange Stories (higher-order ToM; Happé, 1994), and (4) the Frith-Happé animations (higher-order ToM; White et al., 2011). Children were matched on age, nonverbal-reasoning, and receptive language (Table 1).

Results:  Across all ToM measures, children with ASD scored lower than NT children, indicating greater challenges with ToM (Table 1). Using the Developmental Trajectory Approach, we contrasted comparisons between groups’ ToM trajectories, plotted according to chronological age (CA) with those plotted according to verbal mental age (VMA). Looking at the relations between VMA and ToM, children with ASD had a delay in onset of ToM development compared to NT children across all ToM measures. VMA was predictive of performance on all ToM tasks in children with ASD, but only predicted the rate of growth on the Strange Stories Test for NT children. Overall, CA predicted performance on the Unexpected Contents Task and the Frith-Happé animations. However, children with ASD showed growth on these ToM measures whereas NT children’s performance did not change with their CA. CA was not related to performance on the Strange Stories Test or the Birthday Puppy Story in either group.

Conclusions:  In line with past research (Steel et al., 2003), children with ASD showed significant improvement in ToM ability with increasing VMA, and in some instances CA. CA appeared to play a greater role in performance on tasks that required less language comprehension (e.g., Unexpected Contents) in children with ASD. However, the rate of growth in ToM ability differed between groups (i.e., ToM ability in NT children was less affected by CA and VMA). Taken together, these findings provide evidence that the development of ToM in children with ASD may be both delayed and deviant compared to NT children.