Links between gender diversity and autism traits in non-autistic cisgender and transgender adults: Contributions of negative affect and alexithymia

Background

Much of the literature on Autism–gender diversity (GD) relies on self-report Autism screening questionnaires (e.g. Autism-spectrum Quotient; AQ-50) that may tap constructs other than Autism.

Aims

In two studies, we tested whether links between GD traits or categorical gender modality (cisgender/transgender) and Autism traits in non-Autistic cisgender and transgender adults reflect variance shared with alexithymia/negative affect.

Methods

Study 1 (N = 285 cisgender adults) measured Autism traits (AQ-50), alexithymia, negative affect, perceived discrimination, and dimensional GD. Study 2 (N = 208; 104 transgender and 104 cisgender) measured primarily Autism traits, alexithymia, and negative affect. Hierarchical regressions tested whether dimensional GD (Study 1) and being transgender(Study 2) predicted AQ-50 scores before and after other measures were included. Additionally, (a) mediations examined the interplay between alexithymia, negative affect, and Autism traits; (b) regressions tested whether dimensional GD or being transgenderpredicted “Autism-specific” AQ-50 variance, and; (c) groups in Study 2 were matched on alexithymia and negative affect, with differences in AQ-50 tested using independent t-tests.

Results

Both dimensional GD indices (Study 1) and being transgender (Study 2) predicted higher Autism traits, but adding alexithymia and negative affect eliminated the predictive effects. Mediation analyses showed partial, bidirectional alexithymia-negative affect associations with Autism traits. Importantly, Autism-specific variance on the AQ-50 was not predicted by dimensional GD in Study 1 or transgender identity in Study 2. Moreover, the predictive effect of being transgender in Study 2 disappeared after matching groups on alexithymia and negative affect.

Discussion

In non-Autistic cisgender and transgender people, associations between dimensional GD or being transgender and self-reported Autism traits became nonsignificant after accounting for alexithymia and negative affect. These findings caution against inferring Autism-specific links from screener totals. Clinically, elevated Autism traits alone in non-Autistic transgender people should not guide decisions.