Risk Factors for Aggressive Behavior In Youth with Autism Spectrum Disorder

Background: Children with autism spectrum disorders (ASD) frequently exhibit aggressive behavior (Fox et al., 2007), demonstrating higher prevalence rates than typically developing or intellectually disabled individuals (Kanne & Mazurek, 2010). In youth with ASD, research has linked aggressive behavior to deleterious sequelae such as poorer quality-of-life (Gardner & Moffatt, 1990) and greater family stress (Lecavalier et al., 2006). Despite this burden, there is a dearth of research identifying risk factors for aggression in the ASD population. The few studies that have examined determinants of aggression have yielded inconsistent findings. Surprisingly, this body of research has paid little, if any, attention to the role of other psychiatric comorbidities in the development of aggression in this population.

Objectives: To examine the demographic, cognitive, linguistic, and psychiatric risk factors for aggression in youth with an ASD.

Methods: Cross-sectional data were derived from the a local registry project housed in an urban outpatient pediatric clinic. Children were evaluated by trained diagnosticians using DSM-IV-TR criteria and the Autism Diagnostic Observation Schedule (ADOS; Lord et al, 2002).The sample consisted of 170 children, ages 2 to 17.9 years (M = 6.26).

After bivariate relationships were established, stepwise multiple linear regression analyses, using the stepwise regression function in Minitab 15.0 (State College, PA), were performed to identify predictors of aggressive behavior. Aggression, including both verbal and physical, was measured as a continuous variable using the aggressive behavior syndrome scale raw score from the Child Behavior Checklist - Parent version (CBCL-P; Achenbach and Rescorla, 2001). Six predictors were examined:1) demographics, 2) standardized IQ scores, 3) expressive language as measured by the Vineland Adaptive Behavior Scale II (Sparrow et al., 2005), 4) psychiatric comorbidities as assessed by the CBCL-P, 5) repetitive behaviors as measured by the ADOS, and 6) ASD severity using ADOS raw scores (see Gotham et al., 2006). Analyses were conducted separately for children < 6 (n = 96) and 6-18 years given the separate CBCL modules for each age group.

Results: Lower parental education in the < 6 group and younger age in the 6-18 group were associated with increased aggressive behavior (both p < .05). For both age groups, repetitive behaviors, ASD severity, expressive language, and IQ were not associated with increased aggressive behaviors (all p > .05). Results from the multivariate model indicated that anxious/depressed, attention, and somatic problems were associated with increases in aggression among children < 6 years. For children 6-18, significant predictors included anxious/depressed, attention problems, and younger age (all p < .05).

Conclusions: Unique to the literature, this study demonstrates the role of internalizing and attention problems, as well as somatic complaints in younger children, in the development of aggressive behavior in youth with ASD. No significant associations were found between factors widely assumed to contribute to aggression in youth with ASD (i.e., ASD severity, cognitive, and language factors). Taken together, these data highlight the need for both further research and a thorough psychiatric evaluation and intervention plan to prevent or diminish aggressive behavior in children with ASD.