A measure of overall well-being in children with autism has not yet been devised, and little is known about how well-being might relate to autism severity. Previous studies show an inverse relationship between the severity of autism symptoms and adaptive functioning (Paul, Loomis, & Chawarska, 2011), but this relationship has typically been quantified as a correlation between one composite from the Autism Diagnostic Observation Schedule (ADOS; Lord et al., 1989) and a summary score from the Vineland Adaptive Behavior Scales (VABS; Sparrow, Cicchetti, & Balla, 2005). This approach excludes other measures of well-being (e.g., ratings of quality of life and behavioral or emotional regulation) and the restricted, repetitive behavior domain. It also does not take into account the planned revisions to the Diagnostic and Statistical Manual of Mental Disorders to include sensory sensitivity symptoms in the restricted, repetitive behavior domain.
Objectives:
This study examined well-being and its relationship to autism symptoms by using multiple assessments to derive latent variables for both constructs and generating a structural equation model of associations among these variables.
Methods:
Data collected as part of the Autism Treatment Network at the University of Rochester Medical Center were analyzed for 147 children (M = 7.15 ± 2.7 years; 123 males). Autism symptom severity was operationalized as a latent construct comprised of the communication, social, and stereotyped behavior and restricted interests total scores from both the ADOS and the Autism Diagnostic Interview-Revised (ADI-R). Well-being was operationalized as a latent construct with VABS adaptive behavior composite, total problem score from the Child Behavior Checklist (CBCL; Achenbach & Rescorla, 2001), and psychosocial total score from the Pediatric Quality of Life Inventory (PedsQL; Varni, Seid, & Rode, 1999). Sensory symptoms were quantified by using the total score from the short form of the Sensory Profile (Dunn, 1999). Analyses were conducted with AMOS (Arbuckle, 2010) to generate measurement and structural models for this dataset.
Results:
In the measurement model that best fit this data, (1) autism symptom severity was represented by all three ADOS total scores and only the social interaction score from the ADI-R and (2) CBCL and PedsQL represented the latent construct of well-being. This model fit the data very well χ2 (12) = 16.69, p = .162, RSMEA = .052. A structural equation model tested the predictive value of autism and sensory symptoms and revealed that sensory symptoms (β = -.692), but not autism symptoms (β = -.006), significantly predicted well-being.
Conclusions:
This study suggests that well-being in children with autism may be conceptualized as a latent variable comprised of behavior regulation and quality of life and may be predicted by sensory symptoms. While autism symptoms were not significantly associated with well-being in this study, it should be noted that Ray-Subramanian, Huai, and Weismer (2011) found that the relationship between autism symptoms and AF was non-significant after controlling for age and IQ. This study highlights the importance of utilizing additional assessments to model the relationship between autism symptoms and well-being, as both are multi-dimensional constructs.
See more of: Core Symptoms
See more of: Symptoms, Diagnosis & Phenotype