Children with autism spectrum disorder (ASD) experience internalizing and externalizing problems at higher rates than typically developing children, which could worsen social impairment. The present study compared impairment scores (social responsiveness scale, 2nd edition; SRS-2 scores) in 57 children (3-17 years, 82.5% male) with ASD, either with or without heightened levels of anxiety or ADHD symptoms, all per parent report. Children with heightened anxiety problems showed higher scores on four SRS-2 subscales (Social Cognition, Social Communication, Social Motivation, and Restricted Interests and Repetitive Behavior). Children with heightened ADHD traits showed higher scores on two subscales (Social Communication and Social Awareness). These findings suggest similarities and differences in how social deficits in ASD may worsen with anxiety or ADHD symptoms.

Previous work has found gender differences in restricted and repetitive behaviors and interests (RRBI) for autism spectrum disorder (ASD). Compared to girls, affected boys have increased stereotyped and restricted behaviors; however much less is known about gender differences in other areas of RRBI. This study aims to identify whether specific RRBI (i.e., stereotyped, self‐injurious, compulsive, insistence on sameness, ritualistic, and restricted), as measured by item‐level data on the Repetitive Behavior Scale‐Revised (RBS‐R), can distinguish girls from boys with ASD. Participants included 615 individuals with ASD (507 boys; 82.4%), ages 3–18 years of age (M = 10.26, SD = 4.20), who agreed to share data with the National Database for Autism Research (NDAR). Multivariate analysis of variance and discriminant function analysis (DFA) were used to determine whether item‐level RBS‐R data could correctly classify cases by gender. DFA results suggest that RBS‐R items significantly differentiate gender. Strongly differentiating RBS‐R items had greater success in correctly classifying affected boys (67.90%) than girls (61.00%). Items that best‐discriminated gender were heightened stereotyped behaviors and restricted interests items in boys and compulsive, sameness, restricted, and self‐injurious behavior items in girls. This study is the first to find that girls with ASD may have increased compulsive, sameness, and restricted RRBI compared to boys. Additionally, findings support heightened self‐injurious behaviors in affected girls. Future research should disentangle whether elevated rates of RRBI in girls are central to the presentation of ASD in girls or an epiphenomenon of the high rates of co‐occurring disorders (e.g., anxiety) noted in girls. Autism Res 2019, 12: 274–283 © 2018 International Society for Autism Research, Wiley Periodicals, Inc. Lay Summary This study is the first to examine a comprehensive measure of repetitive behavior in children with autism, with findings of increased compulsive, insistence on sameness, and self‐injurious behavior characterizing girls and increased stereotyped and restricted behavior characterizing boys. Future research should determine whether these elevated behaviors in girls are directly part of the autism presentation in girls or symptoms of co‐occurring psychopathology. It is important for autism diagnostic measures to best capture the types of repetitive behavior girls may demonstrate.

While the function of restricted repetitive behaviors (RRBs) in autism spectrum disorder (ASD) is unclear, RRBs may function as anxiety reduction strategies (Joosten et al. J Autism Dev Disord 39(3):521-531, 2009. Moreover, anxiety in ASD is associated with low social motivation (Swain et al. J Autism Dev Disord, 2015. The present study examined social motivation as a mediator between anxiety and RRBs in a sample of 44 children (2-17 years old; 80 % male) with ASD. The relationship between anxiety and IS, but not other RRBs, was partially mediated by social motivation. These findings suggest anxiety is linked to social motivation deficits in children with ASD, which may increase ritualized behaviors and difficulties with changes in routine. Implications are discussed for differing functions and treatment of RRB domains.