Insistence on Sameness Behaviours Selectively Associate with Reported Childhood Pretend Play in Adults with ASD

Background: Research with children with ASD shows that restricted and repetitive behaviours (RRB) reliably  group into two subtypes: repetitive sensory and motor (RSM) and insistence on sameness (IS) and that high levels of RRBs in general are associated with lack of play (Honey et al., 2007).  The present study tests whether RSM and IS subtypes are found in adults with ASD, and whether these subtypes are associated with childhood pretend play.

Objectives: 1) Assess the RRB subgroups found in adults using the Adult Repetitive Behaviours Questionnaire–2 (RBQ-2A; Barrett et al., 2015), the first published self-report measure of RRBs for adults.  Its component structure has not yet been assessed in adults with ASD.
2) Test whether there are any difference in RRBs between participants who played pretend as a child and those who did not play pretend.

Methods: The RBQ-2A was administered to UK participants (N = 787) in an online survey, along with open-ended questions about their childhood pretend play (quantified for the present analysis).  Analyses were only conducted on complete datasets from participants who reported either a clinical diagnosis of ASD or no diagnosis of any kind (i.e. neurotypical [NT]). Participants in the final sample (N = 583) were aged 18-66 years (M = 36.51, SD= 11.67), 219 were male and 317 participants reported a clinical diagnosis of ASD.  The RBQ-2A data were analysed using principal components analysis (PCA) with direct oblimin rotation to allow for correlation between components. The Autism-Spectrum Quotient (AQ; Baron-Cohen et al., 2001) was also completed.   

Results: The final rotated PCA solution explained 45.36% of the total variance and resulted in the two predicted components, RSM and IS.  Participants with ASD scored significantly higher in terms of total AQ score, mean total RBQ-2A score and mean RBQ-2A subscale scores.  Fifty percent of the ASD group and 76% of the NT group reported playing pretend games as children.  For both groups, participants who did not play pretend as a child scored higher on the imagination subscale of the AQ, supporting the role of pretend play as a predictor for imagination.  In the ASD group, participants who did not play pretend as a child scored significantly higher on the IS subscale than those who did play pretend (p < .001), whereas no significant differences depending on pretend play were found for the NT group on either of the RBQ-2A subscales.

Conclusions: The component structure of the RBQ-2A was assessed for the first time in a sample that includes adults with ASD.  As in previous research with children, we identified two reliable subscales comprising RSM and IS, with expected group differences across the total score and its subscales.  The second objective was to test whether childhood pretend play has an effect on later RRBs and we found an ASD-specific effect for IS only.  This is the first evidence that childhood pretend play may be associated with RRBs in adulthood.  Further attention should be given to the importance of the relationship between imagination and RRBs in ASD.