A questionnaire measuring six adult autism spectrum problem domains by self- and other-report

Background:  Psychometrically sound questionnaires measuring adult autism spectrum spectrum problems are scarce but much-needed for 1) epidemiological research, 2) measurement of subthreshold ASD problems, and for 3) profiling scores on different ASD problem domains in patients with a formal ASD diagnosis.

Objectives:  To develop a questionnaire measuring AS problems as they are seen in adults on the spectrum. We aimed for a multidimensional questionnaire in correspondence with the heterogeneous nature of the disorder and the multiple problem domains that can be discerned. Our aim was further to measure both the patient’s perspective and that of an important other (a parent or a spouse) for a comprehensive picture of the AS problems. Without compromising on psychometric quality, we additionally aimed for a brief questionnaire for practical usefulness.

Methods:  We built on our previous work on the Children’s Social Behavior Questionnaire which we developed for children and adolescents (CSBQ; Hartman et al. 2006). We used a top-down rational approach of item writing, taking the CSBQ core problem domains and corresponding items as a starting point (for lifespan continuity) but expanding on this by formulating multiple adult (developmentally appropriate ) equivalents of the CSBQ behaviors. A comprehensive item pool of 90 items was subjected to principal component analysis. Items were selected with a factor loading >.3 on its main factor and a minimum difference of .2 with a possible secondary factor. These requirements had to apply to both the self- and other-report data. Findings are based on 1143 self-report and 644 other-report questionnaires from 6 outpatient clinics in the Netherlands.

Results:  We started analyzing per ASD domain: social, communication, and stereotypic behavior, respectively, in order to make a first selection of adequate items. Results were highly similar for self- and other report. For the social domain the data indicated a two factor structure capturing amount of social contact and quality of social contact. For the communication domain we were able to distinguish between the understanding of social exchanges and acts that are at odds with social rules and conventions. For the stereotypies domain we could differentiate between sensory stimulation and motor stereotypies on the one hand and multiple aspects of rigidity and insistence on sameness on the other hand. Items with aforementioned pattern of factor loadings on both the self- and other-report versions were retained for a second set of analyses across the three domains. This yielded the expected six factor structure in both the self- and other-report version where a total of 38 items (6-7 item per factor) fulfilled aforementioned criteria. Further work on the psychometric qualities of these scales is underway.

Conclusions:  We developed a self- and other-report questionnaire of adult AS problems which differentiates between the following domains: reduced contact, reduced empathy, reduced interpersonal insight, violation of social conventions, insistence of sameness, and sensory stimulation/motor stereotypies. The instrument is short and easy to apply and yields a score profile among these six problem domains both from the perspective of the patient and from someone close.