The Female Phenotype of ASD – Same or Different to the Male Phenotype?

Background: One of the most remarkable features of autism spectrum disorders (ASD) is the relation of male and female affected patients, which refers to 4-5:1 on average. In ASD without cognitive impairment the male:female relation is reported to be 10:1, while it is 1-2:1, in patients with cognitive impairment. With this in mind, a discussion arises scrutinizing whether the female ASD phenotype is identical with the male one and whether diagnostic instruments maybe underdiagnose female participants.

Objectives: The present talk aims to analyze this question by the help of a huge data base and with matched sample comparisons.

Methods: Using one of the largest internationally available samples, which is uniformly diagnosed according to gold standard (www.asd-net.de), data from N= 2613 patients (age 13.3± 10.4, IQ 97,4 ± 20.02) were analyzed, focusing on sensitivity and specificity values of the diagnostic algorithms (ADOS und ADI-R). Due to the huge sample size, an additional sample was generated, which was matched according to gender, diagnosis, the applied ADOS module, age and intelligence (resulting in 454 pairs). We determined items for both genders, which differentiate best between groups (ASD vs Non-ASD) via discrimination analysis.

Results: Comparisons across the whole sample resulted in the following pattern: females with ASD are significantly older than males with ASD, at the time of diagnosis. We observed no IQ differences between both gender groups. Differences according to the intensity of the ASD symptoms were solely observed for older patients (module 3 and 4): female participants show reduced impairments with respect to domains, which are relevant for diagnosis. This effect was especially pronounced during the application of module 3 with moderate effect size. But in the exactly matched sample no differences concerning intensity of symptoms were found between female and male ASD participants. Sensitivity measures of the ADOS cut-off values are reduced in female as compared to male patients, while specificity values are increased. Discriminant analyses show further that there are items like for example, eye contact, which discriminate between ASD and non-ASD independently of gender, while other items discriminate between ASD and non ASD as well, but are dependent on gender.

Conclusions: Interestingly, gender has no influence on specific domains of ASD, when patients are matched, suggesting that gender-depending differences on the ASD symptom-level may be rather attributed to age- and development-related factors and the IQ, than to a specific gender-related phenotype.