Skip to main content

Gender Variance, Autism, and Birth-Assigned Sex Differences: A Meta-Analysis

23 July 2018



Gender variance is an umbrella term used to describe gender identity, expression, or behavior that falls outside of culturally defined norms associated with a specific gender (Simons et al., 2014). Researchers are increasingly examining the co-occurrence of autistic symptoms and gender variance in clinical populations. The current literature does not often directly examine differences between birth-assigned males and females when examining these two clinical realms in relation to each other. Although a shift in birth-assigned sex ratios has been identified in several studies with gender clinics finding an increase in birth-assigned female referrals (Aitken et al., 2015), this occurrence has not been researched in the context of the increasing correlation between gender dysphoria and autism. No meta-analyses have been conducted that synthesize results across studies; therefore, the current study aims to examine the co-occurrence of autistic traits and gender variance in clinical populations as well as examine differences among the birth-assigned male and female participants.


Electronic databases, journals, and reference sections were searched for primary studies and chart reviews assessing the co-occurrence of autistic traits and gender variance. Search terms ‘autism,’ ‘asperger,’ ‘autism spectrum,’ ‘autistic,’ ‘transgender,’ ‘gender dysphoria,’ ‘gender variance,’ ‘gender nonconformity,’ ‘gender identity disorder,’ and ‘transsexual,’ were used. To be included in the meta-analysis, primary studies needed to include individuals presenting to a clinical setting with both autistic traits and gender dysphoria or variance in gender identity. However, participants in the primary studies did not need a DSM-IV or 5 diagnosis of Autism Spectrum Disorder or Gender Dysphoria.

Results and Conclusions

Data from 11 primary studies or chart reviews were collected using a coding manual to capture study descriptors in addition to effect size information. The main variables coded included, natal sex, mean age, autistic traits, and gender variance. After coding was completed, an intrarater reliability analysis was conducted to ensure reliable and accurate coding of the primary studies, and to resolve any discrepancies before performing statistical analyses.

It was expected that a significant correlation between the presence of gender variance and symptoms of autism would be found. Given the increase in birth-assigned females presenting to gender clinics, it was also expected that there a significant positive effect for birth-assigned females endorsing gender variance would be found when compared to birth-assigned males. In addition to the test of the average effect, moderators of the effect sizes were explored.

Implications of the study, including directions for future research and clinical practice are provided.


Files are restricted to Pacific University. Sign in to view.