The Social Phenotype of ASD and Parent Report of Joint Attention in School Aged Children

Background: Joint attention, or the propensity to share a common referent with other people, is a major dimension of the phenotype of ASD. It is central to preschool assessment with modules 1 and 2 of the revised Autism Diagnostic Observation Schedule (ADOS-R) Gotham et al. 2008, 2009). However, joint attention is not included in module 3 and 4 applications of the ADOS-R with individuals with more advanced development. This is not because joint attention ceases to be major feature of ASD as children development. Rather it is because it is unclear how to measure joint attention in older children.

Objectives: The first goal of this study was to develop a theory based measure of joint attention for older higher functioning children with ASD (HFASD). The second goal was to provide data on its sensitivity and specificity for identifying school-aged children with higher functioning ASD.

 Methods: As part of a longitudinal study of higher functioning children with ASD a parent report measure called the Social Development and Joint Attention Rating (SDJAR) scale was developed. This included a pool of sixty items that tapped three subdomains of joint attention in older children: Nonverbal Shared Experience (S/he makes eye contact with you when something in the environment interests him/her; Joint Action (S/he works cooperatively in groups of more than one other child to achieve a common goal); and Verbal Shared Experience (S/he shares exciting events with you that happened in school). Parent report data on SDJAR were collected from 38 parents of children with HFASD (Age = 11.0, SD = 2.1 years; IQ = 101, SD = 13.4) and 36 parents of age and IQ matched comparison group (Age = 11.5, SD = 2.2; IQ = 106, SD = 16.1) that included children with ADHD symptoms (N = 21) or typical development (N = 15). Data were also collected on the ADOS-R, Social Communication questionnaire and the Social Responsiveness Scale.

Results: Forty-nine questions of the SDJAR item pool constituted a scale with significant internal consistency, Chronbach’s Alpha = .88, p < .001.  Confirmatory factor analysis revealed one factor that explained 56% of the variance across these items.  A discriminant function analysis revealed the SDJAR factor score displayed a sensitivity of 82% for the identification of children with ASD and a specificity of 86% for the identification of ADHD and Typical children. Additional analyses indicated that sensitivity and specificity of the SDJAR exceeded those of parent report on SCQ and SRS in this sample. The SDJAR was not significantly correlated with these measure, but was correlated with the ADOS in the both the ASD and ADHD samples (see Figures 1 and 2).

Conclusions:  Joint attention remains a prominent dimension of the phenotype of ASD in school-aged higher functioning children. Indeed, in this study, parent report of joint attention related behaviors was more strongly related to ADOS based sample classification of higher function children with ASD versus  ADHD and TD controls than was the case for commonly used parent report measures of ASD symptoms.