Factors Associated with Parents' Ratings of the Severity of Autism Spectrum Disorder: A Population Study

Background: There is currently little consensus on how severity of a child’s autism spectrum disorder (ASD) should be measured (Bernier, 2012). Definitions of severity have been based on criteria such as the number and intensity of core symptoms, the child’s overall level of functional impairment, and the child’s ASD diagnosis subtype (e.g. Asperger’s Disorder, Autistic Disorder). Additional definitions have focused on the presence of secondary diagnoses, including intellectual disabilities, behavioral problems and learning disabilities. Despite this lack of a standardized definition, parents are readily able to answer a question asking them to describe the severity of their child’s ASD. The goal of the current study is to examine the basis for a parent’s judgment of ASD severity.

Objectives: Identify child and family factors that are associated with a parent’s severity rating of their child's ASD.

Methods: Data come from two national surveys, the 2009-2010 National Survey of Children with Special Health Care Needs (NS-CSHCN) and a follow up survey to the NS-CSHCN, the 2011 Survey of Pathways to Diagnosis and Services (“Pathways”).  Parents in households with a child diagnosed with an ASD between the ages of 6-17 were eligible for the current study.  To be included in the final sample the parent had to have completed both surveys, including a mailed module in the Pathways survey (n=967). Parents answered questions about the severity of their child’s ASD (mild, moderate or severe), autism symptoms (utilizing the Children’s Social Behavioral Questionnaire (Hartman, 2006)), the impact of the child’s condition on the family, and the impact of the child’s condition on the child’s social and behavioral health (utilizing the Strengths and Difficulties Questionnaire (Goodman, 1997)). A survey weighted measurement model was created utilizing Mplus software to generate latent factors of child impact, family impact and child symptoms. Additional analyses were completed in Stata 12.0. Associations between demographic characteristics and level of parent-reported severity were calculated utilizing weighted χ² tests. Separate weighted multivariate logistic regression models, adjusted for demographics of the child and family were fitted to compare different severity groups by latent factors.

Results: Chi-square tests revealed that children with a mild severity rating were younger than children with a moderate or severe severity rating. Children with a mild ASD were less likely to have anxiety, depression, behavioral and conduct problems or an intellectual disability when compared to children with more severe ASD.  In adjusted models, children with moderate ASD had more severe symptoms (OR=2.16, 95% CI:1.38-3.41, p=.001) and family impacts (OR=1.79, 95% CI:1.26-2.54, p=.001) than children with mild ASD.  Meanwhile, children with severe ASD had higher family impacts (OR=2.80, 95% CI:1.66-4.72, p<.001) than children with moderate ASD, but not higher symptoms.

Conclusions: A parent’s conceptualization of their child’s ASD may primarily vary as a function of the impact of the child’s disability on the family, and to a lesser extent on the impact the child’s disability has on the child. Therefore, a parent’s conceptualization of severity may not always coincide with objective or clinical assessments focused on symptoms or functional impairment (Perrin et al., 1989).