Persistence of Repetitive Behaviors in ASD

Background: Repetitive behaviors (RBs) are a defining feature of autism spectrum disorder (ASD) and have been studied extensively over the past decade (Leekam 2011). The use of RBs as a phenotype in genetic studies of ASD has expanded beyond an overall score (Cuccaro 2003; Szatmari 2006) but remains somewhat limited (Shao 2003; Smith 2009). Given evidence for trajectory of symptoms as an important phenotype in various diseases, exploration of change in RBs may represent a novel way to incorporate RBs into genetic studies of ASD.

Objectives: The primary objective of this study is to examine the pattern of changes in RBs using two empirically identified repetitive behavior factors, Insistence on Sameness (IS) and Repetitive-Sensorimotor Behaviors (RSMB). The goal of this study is to determine whether change in RBs constitutes a viable phenotype.

Methods: Our dataset consisted of 989 unrelated individuals with ASD ranging in age from 48-239 months (M=116 months, sd=50) who were participants in a larger study of autism genetics. The sample was predominantly male (84%). The primary data for analysis consisted of 13 items from the ADI-R Interests and Behavior section and one item from the Language and Communication Functioning section (Verbal Rituals). For each item, Ever and Current ratings were used to develop the change score (i.e., Change=Current-Ever). Composite scores were calculated for IS and RMSB. Change scores for these composites served as our primary outcome of interest. Given the scoring convention used in the ADI-R, change scores reflect only whether current behavior has improved. Age, sex, language level, and adaptive level were incorporated into our analyses. Descriptive data for repetitive behavior items and composites were generated. To test for differences in degree of change for the respective composite scores we used a mixed models approach.

Results: Among the 14 RBs, the most frequently occurring (based on Ever ratings) were Repetitive Use of Objects (83%), Undue General Sensitivity to Noise (79%), and Unusual Sensory Interests (76%). At the item level, the most change was observed for Repetitive Use of Objects (48%) while the least change was noted for Resistance to Trivial Changes in the Environment (14%) and Unusual Preoccupations (16%). We tested the association of age, sex, and adaptive level on our composite scores of interest (i.e., changes in IS and RSMB) and found that age at exam was significantly associated with both change in IS and RSMB; however, sex showed no association to change in IS but was significantly associated with the RSMB change score. Similarly, adaptive level was significantly associated with the IS changes score but not RSMB. Using a mixed model approach and adjusting for sex, age at exam, and adaptive level, we found no differences in the amount of change in IS and RSMB.

Conclusions: Change in repetitive behaviors vary across the different types of repetitive behavior. This preliminary analysis sheds light on patterns of change in RBs in ASD that may be useful in developing novel phenotypes. Certainly, the role of age, sex, and developmental level are critical influences on such phenotypes.