Identifying Autism Spectrum Disorders In Adults with Intellectual Disability: The Validity of the Social Communication Questionnaire

Background:  Identifying autism spectrum disorders in both children and adults who present with diverse symptoms is a crucial step in understanding the nature of these disorders and allocating resources to individuals who may need specific treatment and support for the associated challenges.  Individuals with ASDs often require extensive life-long support, and continued assessment throughout the lifespan is essential for maintaining appropriate care and treatment.

Continued research on the validity of ASD screening measures is important in identifying individuals who may need further assessment for autism, especially in at-risk populations, such as individuals with intellectual disability.  Many adults who have autism spectrum disorders were misidentified as having other disorders before the increased awareness of ASDs and the introduction of more sophisticated diagnostic tools.  While the psychometric studies of the SCQ in children and adolescents provide evidence that the SCQ is a useful and valid screening measure, no independent studies on the validity of the SCQ among adults with ASDs were found. 

Objectives: This project examined the validity of the Social Communication Questionnaire (SCQ) in a sample of 69 adults with a prior diagnosis of 1) intellectual disability (ID) and an autism spectrum disorder (ASD), 2) intellectual disability (ID) presenting with moderate-severe challenging behavior, or 3) intellectual disability (ID) presenting with mild-no challenging behavior. 

Methods: Adults with intellectual disability (ID) and/or their guardians were recruited from agencies in Central Ohio and provided assent/consent to allow researchers to review agency records to determine prior diagnosis of autism spectrum disorders.  Participants with ID nominated support staff to complete several measures to assess ASD symptoms (the SCQ and a DSM-IV Symptom Checklist), behavior problems (the Aberrant Behavior Checklist-Community), and adaptive functioning (Adaptive Behavior Assessment System – Second Edition) of participants with ID.  The discriminative validity of the SCQ Current version was assessed by examining sensitivity and specificity.  Receiving Operating Characteristic (ROC) analyses were conducted to identify optimal cut-off scores of the SCQ Current. 

Results:  The SCQ Current version, as rated by support staff, yielded a sensitivity of .76 and a specificity of .79 at the cutoff score of 15 proposed by the authors.  However, the optimal cutoff score in this sample was 12, which yielded a sensitivity of .86 and specificity of .65.  The ROC analysis of SCQ Current total scores in the total sample yielded an area under the ROC curve (AUC) of .72.  Analyses were repeated in a subset of participants in the ID-only group with high and low levels of behavioral problems, and comparable sensitivity, specificity and AUC values were found. 

Conclusions: Although the SCQ Current version was not specifically designed to screen for ASDs in adults, it may be a useful tool for screening individuals whose early developmental history is unavailable.  A lower cutoff score than the authors originally proposed is recommended for use in adults, which is consistent with research indicating that autism symptoms improve with age.  Behavior problems did not appear to affect the discriminative validity of the SCQ Current in this sample.