The Role of Child and ASD Characteristics in Caregiver-Reported Anxiety: Correlates and Cluster Analyses from a Large Pooled International Database of Autistic Young People

Background: Anxiety is one of the most common co-occurring conditions in ASD. However, studies that have investigated associations between child characteristics, ASD symptoms and anxiety are mostly of small sample sizes, tend to report inconsistent findings and have mostly explored overall (total) autism symptom severity scores. More in depth analyses of group clusters based on anxiety and ASD symptomatology profiles would enable greater understanding of the relationship between specific ASD characteristics and anxiety.

Objectives: This study pooled together data from 13 existing studies to examine the relationships between child characteristics (age, gender, cognitive/ adaptive functioning and ASD traits) and anxiety in a large sample. We also aimed to combine a variable-centered and person-centered approach in order to provide better understanding of the relationship between individual differences across anxiety subtypes with age, gender and core ASD traits.

Methods: A large international sample of 870 children and adolescents with ASD (763 males, 107 females, mean age 11.6 years, SD=2.77) was pooled together from 13 studies in the UK, Singapore and the US, all of which used the Spence Children’s Anxiety Scale- Parent Version (SCAS-P) total and subscale scores. Several other different measures of functioning and ASD characteristics were employed by the different studies (i.e. WISC, WAIS, SCQ, SRS, etc.) and harmonized scores were created for these. Regression and moderation analyses using child characteristics and ASD traits as predictors were conducted for SCAS-P total and subscale anxiety scores. Latent profile analysis (LPA) with item level SCAS-P data will be used to identify profiles (groups) of individuals with distinct constellation of ASD symptoms. Models with 1-8 profiles will be estimated and the decision on the optimal number of profiles will be guided by the combination of information criteria, likelihood-based statistics and entropy. Derived anxiety based profiles will be further defined by exploring their association with core ASD traits, age and gender.

Results: Age was positively associated with social and generalized anxiety. No consistent gender differences were found. ASD symptoms, especially Repetitive and Restricted Behaviors and Interests (RRBIs), were positively associated with anxiety. At the item level, autistic symptoms relating to adherence to routine, intense special interests, sensory hypersensitivity and being aloof/ not showing feelings, together with participants’ chronological age, were the most consistent predictors of SCAS-P anxiety symptoms. Tentative support that some moderation effects exist was also found, but effects were small. Cluster analyses are currently being carried out and these will also be reported.

Conclusions: The findings of this study strengthen existing literature examining the specific relationships between ASD symptomatology and anxiety in young autistic people. Implications of identifying distinct profiles/ groups and how this could enable more individualised targets for anxiety interventions are also discussed.