Using the Social Responsiveness Scale–Short Form to Assess ASD Characteristics in Children with Dup15q Syndrome

Background: The Social Responsiveness Scale (SRS) is a quantitative measure of autism symptoms, which is widely used in studies of autism risk and phenotypic characterization of associated genetic syndromes. Due to concerns over the vulnerability of the SRS to non-ASD traits such as cognitive and language abilities, Sturm and colleagues (2017) developed a short form SRS, using item response theory to identify the least vulnerable items. The resultant 16-item short form was unidimensional and un-biased by language or non-verbal cognitive ability. In order to further evaluate the performance of the SRS short form in children with cognitive and language impairment, we examined the SRS scores of children with duplications of 15q11.2-q13.1 (dup15q syndrome), which is one of the most common copy number variants associated with ASD and intellectual disability (ID). Dup15q syndrome is characterized social communication deficits, cognitive and language impairment, hypotonia, motor impairment, and epilepsy (Battaglia etal., 2010). These widespread and significant impairments present a challenge to accurate measurement of ASD characteristics.

Objectives: 1) Examine the association of short-form SRS scores with verbal and non-verbal cognitive ability, and other measures of ASD characteristics (SCQ, ADOS) in children with dup15q syndrome, 2) compare these associations in children with dup15q syndrome and age and IQ-matched children with idiopathic ASD and ID.

Methods: Participants included 31 children with dup15q syndrome (3-14 years; IQ M(SD) 35(22)), and an age and IQ-matched group of children with idiopathic ASD (N=18). Participants were assessed using the Autism Diagnostic Observation Schedule (ADOS) and cognitive/developmental assessment (Mullen Scales of Early Learning or Differential Ability Scales). Parents completed the SRS and Social Communication Questionnaire (SCQ). The SRS was scored with original long form T-scores and revised short form totals.

Results: In children with dup15q syndrome, both long and short form SRS scores were negatively associated with verbal IQ (r=-.61, p<.001; r=-.42, p=.02, respectively), and non-verbal IQ (r=-.7, p<.001; r=-.46, p=.009). Long and short form SRS scores showed moderate positive association with ADOS severity scores (r=.62, p=.004; r=.56, p.01) and SCQ scores (r=.69, p=.003; r=.57, p=.02). In contrast, neither long nor short form SRS scores were associated with verbal or non-verbal IQ in the ASD+ID group, but did show the expected positive association with SCQ and ADOS scores (r-values .65-.7, p-values .008-.02).

Conclusions: SRS short form scores did not show an association with language or cognitive abilities in children with ASD+ID, consistent with the results from Sturm and colleagues. However, in children with dup15q syndrome, SRS short form scores were negatively correlated with language and cognitive scores, albeit with reduced correlations strength. Encouragingly, SRS short form did show the expected positive association with other measures of ASD characteristics in the dup15q sample. In children with dup15q, the short form SRS represents an improvement over the long form in terms of reduced relationship to non-ASD characteristics, but was not independent of those characteristics as in the ASD+ID sample. The SRS short form should be interpreted with caution when used in genetic syndromes with a high degree of impairment across domains.