Restricted and Repetitive Behaviors in Young Children with ASD: Concurrent Correlates and Predictors of Change Over a One-Year Period

Background: Restricted and repetitive behaviors (RRBs) are a core feature of autism according to diagnostic criteria. Research has shown that RRBs–specifically, those with a repetitive sensorimotor component as opposed to those characterized by an insistence on sameness–are prevalent among children on the autism spectrum as young as age two and are more common among these children than same-age peers with typical development and those with developmental disabilities who are not on the autism spectrum (Richler, Bishop, Kleinke, & Lord, 2007). The degree of RRBs present at age two has also been shown to predict expressive language level at age four (Paul, Chawarska, Cicchetti, & Volkmar, 2008). However, much of the existing research on RRBs in young children has been based on parent-reported behavior (e.g., Bishop, Richler, & Lord, 2006; Richler et al., 2007) and has not examined change in RRBs over time during early childhood. Examining clinician-observed RRBs and changes in RRBs during early childhood can further enhance our understanding of RRBs in young children on the autism spectrum. Objectives: The objectives of this study were to examine factors associated with (a) clinician-observed RRBs on the ADOS at ages two and three, and (b) changes in the degree of clinician-observed RRBs on the ADOS over this one-year period. Methods: Participants were 102 children (mean age at Visit 1 = 31 months; mean age at Visit 2 = 44 months) on the autism spectrum who are part of a longitudinal study of early language development. Autism spectrum diagnoses were determined using comprehensive diagnostic evaluations that included the ADI-R and ADOS. Measures used in the current study (ADOS; Mullen; Vineland-II; and Preschool Language Scale-4^th Edition) were administered at Visit 1 and Visit 2. Results: At Visit 1, the degree of RRBs observed was negatively correlated with receptive language skills, as measured by parent report on the Vineland-II (r = -.26, p<.05) and performance on the PLS (r = -.39, p<.01). RRBs were also negatively correlated with performance on the Mullen Visual Reception scale (r = -.34, p<.01). At Visit 2, observed RRBs continued to be negatively correlated with receptive language skills on the Vineland-II (r = -.33, p<.01) and PLS (r = -.36, p<.01), but were not significantly associated with performance on the Mullen Visual Reception scale. The degree of RRBs observed at Visit 2 was also negatively correlated with expressive language skills, as measured by the PLS (r = -.25, p<.05). A regression analysis of standardized residual change scores revealed that growth in receptive language skills over a one-year period, as measured by the PLS, significantly predicted reduction in the degree of observed RRBs, when controlling for changes in expressive language and cognitive skills. Conclusions: RRBs among children on the autism spectrum were found to be negatively associated with receptive language skills at ages two and three. An increase in receptive language skills over this one-year period predicted a decline in observed RRBs. Associations between RRBs and expressive language and cognitive skills were less clear at these young ages.