Do Repetitive Behaviours and Restricted Interests Predict Later Cognitive Ability in Children with Autism Spectrum Disorder?

Background: Restricted and repetitive behaviors and interests(RRBIs) are one of the two domains of symptoms that constitute the diagnostic criteria of Autism Spectrum Disorder(ASD)(APA, 2013) and are therefore an important pillar for ASD diagnosis. The functional impact of RRBIs is source of debate in the literature. When studied as a single construct, RRBIs have shown to be negatively correlated to cognitive and adaptive functioning (e.g. Troyb et al. 2016). However, some argued that RRBI symptoms should be examined individually, rather than as a single homogenous construct, as they represent a vast array of heterogenous behaviours with possibly distinct underpinnings differentially affecting functioning across the lifespan (Lidstone et al. 2014). Support for this hypothesis comes from studies showing that when RRBIs are broken down into specific behaviours, some have been found to have an “adaptive” function and be associated to higher cognitive abilities(i.e. stereotyped language, Kim & Lord, 2010; restricted interests, Bishop et al., 2006; rituals, Militerni et al., 2002; insistence on sameness, Mooney et al., 2009). Nonetheless, there is still a paucity of research investigating individual RRBIs and their functional impact across development.

Objectives: Using data from a large longitudinal inception cohort of children with ASD, we explored whether the presence or absence of specific RRBIs, as reported by parents around the time of their child's diagnosis, could help predict non-verbal cognitive ability later in childhood.

Methods: Data were drawn from Pathways in ASD, a longitudinal study of 421 preschoolers diagnosed with ASD followed yearly across five Canadian provinces. Analyses included 206 children for whom the Autism Diagnostic Interview – Revised(ADI-R) was completed when they were 2-4 years (M=39.6 months; SD=8.6) and the Wechsler Intelligence Scales for Children – Fourth edition(WISC-IV) was completed when they were 8-9 years (M=104.7months; SD=2.5). Selected RRBI predictors of non-verbal cognitive ability (measured using the WISC-IV Perceptual Reasoning Index (PRI) standard score) included: stereotyped language (Q. 33), restricted interests (Q.68), rituals (Q.70), and insistence on sameness (Q.75), grouped as “absence” (score=0) or “presence” (scores of 1-3) of symptom as reported on the ADI-R. Student tests and linear regression models were performed to answer research question. Model validation was examined using adjusted R squared and homoscedasticity test of residuals variance.

Results: We found that the group of children for whom parents reported rituals (R²=0.02 F=4.5, p=0.04) or insistence on sameness (R2=0.02 F=4.2, p=0.04) around the time of diagnosis had a tendency to have a slightly more elevated mean PRI standard score later in childhood as compared to the group of children for whom parents did not report these behaviors. No significant results emerged for the other two RRBIs (p>.05)

Conclusions: Our results support the notion that RRBIs should be considered as distinct behaviours that can differentially affect functioning across development. Further research is warranted using observational data and/or more dimensional measures capturing frequency of the behavior (rather than presence or absence) to better understand the differential impact of RRBIs on different spheres of functioning and across the lifespan to ultimately better inform RRBIs’ targets in interventions.