Screening for autism in adulthood – Exploratory data analysis for psychopathological self-assessment of 1382 individuals of a clinical population

Beck, Stephan Johannes (2022). Screening for autism in adulthood – Exploratory data analysis for psychopathological self-assessment of 1382 individuals of a clinical population. PhD thesis, Universität zu Köln. Open Access

PDF
Promotion_13_2_Stephan_ Beck_21_10_2022.pdf
Download (1MB)

Abstract

The aim of this study is to gain new insights into psychopathological self-assessment and hypotheses for improving the screening of autism based on the largest clinical sample of an outpatient clinic for adults with late diagnosed autism spectrum disorder (ASD) to date. This retrospective data analysis focuses on exploring the influence of the variables gender, age, intelligence quotient (IQ) and depressive symptoms on the widespread self assessment questionnaire, the Autism-Spectrum Quotient (AQ-50). In addition, a revised edition of the short version of the AQ-50 (the AQ-10 by Allison et al. 2012) will be created. For this purpose, 528 patients (366 males and 162 females) of an outpatient clinic with an ASD (ASD+) as well as 854 individuals (544 males and 310 females) in whom autism could be ruled out (ASD-) are included in this study. The age of all 1382 individuals ranges from 18 to 75, with a median age of 35. Depending on the topic and under consideration of the required statistical assumptions, Mann-Whitney U tests, Spearman’s correlations, biserial correlation, independent samples t tests, item-analyses via the Discrimination Index (DI) and Receiver Operating Characteristic (ROC) analyses are performed. The results of this study indicate that gender and age have a comparable sized effect on the self evaluation of autistic traits as the diagnosis ASD itself. After controlling for these variables, it seems that a more precise discrimination between individuals with and without an ASD can be achieved, as demonstrated in ROC analyses. IQ and severity of depressive symptoms does not seem to differ significantly between ASD+ and ASD-. Only in ASD+ self-reported autistic symptoms appear to correlate positively with the IQ. Depressive symptoms only seem to correlate with subscale ‘attention switching’, but not with the total score of the AQ-50. Regarding the subscales of the AQ-50, the results indicate that adults with an ASD score comparable in ‘attention switching’ and ‘attention to detail’ as their counterparts without an ASD. In contrast, there are significant differences in the response to subscale ‘imagination’ depending on the diagnosis ASD, although the underlying symptomatology of this factor is hardly represented in the common diagnostic systems. ‘Imagination’ also seems to be the only subscale that is not influenced by variables gender and IQ. Item analysis implies that for 9 items of the AQ-50, adults without an ASD score more frequently than their counterparts with an ASD and only 3 items of the original AQ-10 (by Allison et al. 2012) overlapped with the here created AQ 10 revised. Finally, ROC analyses suggest that the AQ-10-revised may differentiate better between ASD+ and ASD- than the original AQ 10 and AQ 50. However, the AQ-10-revised should be validated in further studies.

Item Type:

Thesis (PhD thesis)

Translated title:

Title	Language
Screening bei Autismus im Erwachsenenalter – Eine explorative Datenanalyse zur psychopathologischen Selbstbeurteilung von 1382 Individuen der Inanspruchnahme-Population einer Spezialambulanz für Autismus im Erwachsenenalter	German

Creators: