Autism-Spectrum Quotient (AQ): A Preliminary Study of Its Diagnostic Validity in a Clinical Spanish Sample, More Than a Psychometric Test?

Background:  Autism is defined in terms of abnormalities in social and communication development, in the presence of marked repetitive behavior and limited imagination (American Psychiatric Association [APA], 1994). The cultural background can greatly influence how an individual perceives and presents with psychiatric symptoms, as well as impacts diagnosis and treatment.  Establishing who is “affected” and who is not, or the degree to which the accepted standardized diagnostic criteria for ASD condition are applicable to a given patient will lead to better diagnosis.

Objectives:  Screening for possibly affected individuals within the typical population, would allow us to refine this diagnostic tool in order to make a diagnosis more consistent with the intrinsic characteristics of ASD.

Methods: This study encompassed two different samples (ASD and TD); they were aged between 17 and 25 years. The first group (TD) consisted of 129 students from the La Laguna University, Tenerife, Spain. During the break of one of their classes, the students were asked to complete the AQ. Students were recruited from the fields of humanities (law and philology; n = 46) and natural sciences (including medicine; n = 83). All participants in the Autistic-high functioning group (n = 21) had a diagnosis and recruited through (APANATE, ALDIS, ASPERTEN). The Autism Spectrum Quotient (AQ) was developed by Baron-Cohen et al. (2001) and translated into Spanish by Betty Trabal, Editorial Amat, S.L., Barcelona, (2005).

Results:  The Total AQ score and its subscales were assessed for normality through an examination of absolute skew and kurtosis score for each variable; the data was significant not found to deviate from normality, *P = .04; R² = .13 (TD group) and *P = .03; R² = .08 (ASD group), respectively (Figure 1) . AQ Total values distribution were analyzed separately for both groups, TD and ASD, which found that was normally distributed, Kolmogorov-Smirnov P > 0.10 respectively (Figure 2). We found that TD individuals had a lower Total AQ score of M = 19 (SD = 7.07), compared to ASD individuals Total AQ score of M = 31.32 (SD = 5.74); this difference was significant, P < .0001 (Figure 3); The discrimination power of the AQ test case and control groups show a significant difference **P = 0.006 and both groups show a Gaussian approximation, R² = 0.9955 for TD and R²= 0.9859 for ASD. The mean comparison for gender and AQ Total score show that TD males scored slightly higher than TD females and that ASD males scored higher than females on Total AQ score (Figure 4). Furthermore our results revealed that was significant difference for gender by diagnosis on the subscales such as social skills, communication, imagination, and attention switching (Figure 5).

Conclusions:  The Spanish version of the Autism Spectrum quotient (AQ) has showed satisfactory levels of internal consistency. The results confirmed the AQ Total and Subscale scores. Finally, gender differences were obtained. These findings support the use of the Spanish version of the AQ for the assessment ASD.