31683
Sex Differences in Youth with ASD: Language Phenotype and Relation to Autism Behaviors from the ACE Gendaar Network
The GENDAAR study is a multisite collaboration focused on understanding sex-based differences in youth with ASD using multimodal behavioral measures, neuroimaging, and genetics. This analysis specifically focuses on sex-based language differences in youth with ASD and neurotypical development, and the relation to autism behaviors (as reported by parents or observed by clinicians). In this abstract, we focus on the Word Class Expressive (WC-E) domain of the Clinical Evaluation of Language Fundamentals (CELF), in which the participants explain why words go together; this task involves both word knowledge and social/cultural-understanding of similarities between words.
Objectives:
To understand sex differences in language ability in verbal youth with ASD and how these sex based differences in language ability impact assessment of autism symptoms.
Methods:
8- 18-year-old individuals participated at four sites as part of the ACE GENDAAR Network. This analysis included 157 youth with ASD (71 female) and 158 youth with neurotypical development (NT; 80 female). All individuals with ASD met criteria on the ADOS, ADI, and DSM and had Verbal IQ > 70. Participants with NT had no family history or concerns about ASD and had scores £ 11 on the SCQ and £ 60 on the SRS. Language was assessed using the DAS II (Word Definitions, Verbal Similarities) and the CELF-5 subdomains. Analyses used ANOVA and regression to investigate effects of participant sex and diagnosis on language skills; and the contribution of language ability to observed and reported autism behaviors.
Results:
On all standard language measures, the NT group had higher scores than the ASD group (ps<.05). There were no sex differences within the ASD group on the DAS Word Definitions or Verbal Similarities; but there were sex differences on the CELF, with ASD females demonstrating better language skills than ASD males (Total Score; Recalling Sentences; Formulating Sentences; WC-E; Word Definitions; ps<.05). In the ASD group, WC-E was predicted by the sex (t=1.96) but not age, autism symptoms, executive functioning (EF) or parental education. Within the ASD group, parent report of autism behaviors (via SCQ) was predicted by sex (t=-1.7), with additional variability explained (FD=5.0) by language WC-E scores (t=-2.5), EF (t=2.5), and parental education (t=-3.9). In contrast, observed autism severity (ADOS calibrated severity score) was predicted by sex (t=-2.7) with no additional contribution from language and other variables (FD=.84).
Conclusions:
In this preliminary analysis, female youth with ASD had better verbal skills than males with ASD but only on the CELF, suggesting that even when “similar” language domains are being assessed, subtle differences in the measures may identify different sex related functioning within ASD (and may account for previous inconsistencies in sex-based findings in ASD). Sex and language were related to parent report of autism behaviors but not clinician observation of autism severity. Additional analyses will explore sex differences in other aspects of language and probe why differences were observed on the CELF but not the DAS; discussion will focus on implications for identifying sex-based differences in other autism samples.