The Autism Spectrum Quotient-Revised: A measure to better identify the Autism Spectrum Disorder presentation in females?

Abstract

Literature reflects a growing concern that females may be "slipping through the net" because current understanding of the autistic phenotype and diagnostic approaches are derived primarily from research on males. The present study sought to develop a psychometrically sound revision of the Autism Spectrum Quotient (AQ), to improve its sensitivity in identifying adult females with high levels of Autism Spectrum Disorder (ASD) traits. In phase one of the study, 30 experienced ASD practitioners rated how relevant they thought 20 newly developed screening items were to ASD in females. Fifteen of these items were then added into the existing AQ, comprising the AQ-Revised. Alongside measures assessing convergent and divergent validity and levels of distress and impairment, 1007 non-ASD (55% female) and 45 ASD (51% female) participants completed the AQ-Revised. The AQ-Revised and the AQ were found to be psychometrically sound and both fit a 2-factor structure, based on factor analyses. No significant gender differences were apparent in non-ASD participants on the total AQ-Revised, however in the ASD group females scored significantly higher on the total AQ-Revised and AQ scores compared to males. The AQ-Revised was found to discriminate well between ASD and non-ASD cases and, with a cut off score of 29, captured 130 more females high on ASD traits than the existing AQ. These women scoring above 29 on the measure were found to have significantly higher levels of distress and functional impairment than those scoring below 29, but reported levels of distress and impairment were equivalent to women scoring above the recommended AQ cut off (32). The utility of the revised instrument for females and the concept of gender-specific ASD screening are discussed.