Repetitive and Stereotyped Behaviors in Autism: A Longitudinal Exploration of the Repetitive Behavior Scale-Revised throughout Childhood

Background: Repetitive and stereotyped behaviors are a core domain of autism spectrum disorder (ASD); however, little is known about how these behaviors co-occur and evolve with time, and how they correlate with other developmental outcomes. Previous study results tend to differ depending on what measure is used. To date, the Repetitive and Behavior Scale –Revised (RBS-R; Bodfish et al., 2000) is the most frequently used in the literature. Mirenda et al. (2010) validated the use of this parent-report questionnaire in a large cohort of young children with ASD using confirmatory factor analysis based on models proposed in the literature. They reported that a three-factor model yielded the most parsimonious results and that the three factors correlated negatively with adaptive behaviors, while none correlated with IQ.

Objectives: The goal of this study was to explore the evolution of repetitive and stereotyped behaviors exhibited throughout childhood using the Mirenda et al. (2010) sample and to document the associations between the trajectories identified and developmental outcomes.

Methods: Data were drawn from the Pathways in ASD cohort (N=421). Participants were assessed using the Repetitive Behavior Scale-Revised at the time of diagnosis(T1), 6 months(T2) and 12 months(T3) later, and followed-up at ages 6 years(T4), 9 years(T5), and 11 years(T6). Assessments of developmental outcomes (Vineland Adaptive Behavior Scales for adaptive functioning, the Perceptual Reasoning Index of the WISC-IV for non-verbal IQ, and the Core Language score of the Clinical Evaluation of Language Fundamentals for language) were completed at T5 or T6 (only the most recent assessment was considered). To create a measure of the number of different behaviours exhibited that was independent of severity, 43 items on the RBS-R were recoded into dichotomous variables and summed. Scores 1-3 were assigned a 1 (exhibits behaviour), while items with a score of 0 remained (does not exhibit behaviour). Group based trajectory modeling analysis was then performed to describe children’s RBS-R trajectories using these summed scores. One way ANOVAs were used to examine whether trajectory groups were associated with developmental outcomes.

Results: A five-group trajectory model was determined the best fit. Trajectory 1(11.5%) had the lowest numbers of RBS-R exhibited behaviours throughout the time period, followed by trajectory 2(28.2%). They both tended to decrease gradually with time. Trajectories 3(19.5%) and 4(24.8%) had higher numbers of exhibited behaviours at T1, but trajectory 3 decreased considerably from T2 to T4 whereas trajectory 4 slowly increased during the time period. Trajectory 5(16.1%) had the highest numbers of exhibited behaviours at all time points and tended to decrease with time. Children in trajectories 4(increasing) and 5(highest) had lower adaptive functioning (p<.001). Children in trajectory 4 also had lower language levels (p<.05). Non-verbal IQ did not differ between groups.

Conclusions: These results are a first step toward a better understanding of how the numbers of exhibited repetitive and stereotyped behaviors change over time in autism. Such longitudinal approaches can be useful in understanding how this domain of symptomatology is linked to development. Further insight involving the role of behaviour severity, in addition to its appearance, is also warranted.