Non-Binary Gender Is Associated with Higher Autistic Traits

Background: There is a relationship between autistic traits and gender: it is well-established that they are higher in males than females and ASD  is more commonly diagnosed in males.  There is now also evidence that individuals attending gender identity clinics for gender re-assignment score highly on measures of autistic traits, but those self-identifying as of non-binary gender need not be seeking gender re-assignment, and the incidence of autistic traits in this broader case of non-binary gender experience is unknown 

Objectives:  

This study aims to provide initial data on autistic traits in those who experience their gender as non-binary. 

Methods:  

An online survey advertised to students at University of Reading and through social media sites used by individuals identifying as of non-binary gender.  The Autism Spectrum Quotient (AQ) was used as a measure of autistic traits. Respondents were asked to report their gender and male, female or other.  The other category offer a list of further options based on self-identification: agender,androgyne, bigender, genderfluid, genderqueer, neutrois, pangender, trans*female, trans*male, trans*person, two-spirit, and a “prefer not to say” option.  Respondents were also asked basic demographic questions, including age, country, first language, educational level and fluency in English and whether they had a diagnosis of ASD.

Results:  

329 adults started the survey, 39 individuals completed gender questions but not AQ. 273 completed the AQ and the gender questions, of these 40 respondents reported receiving an ASD diagnosis and 12 chose not to disclose this information.  221 respondents reporting no ASD diagnosis were included in the analysis:  N=33 male, N=119 female, N=69 non-binary.  Mean (SD) AQ scores for male = 22.9 (8.2), female = 17.4 (8.2), non-binary = 28.3 (7.4). Multinomial logistic regression was used to look at the association between gender and AQ score, controlling for reported ASD diagnosis and demographic variables.  The model was significant (χ²(12) = 91.0, p < .001).  Only AQ score was a significant predictor of reported gender.  An increase in AQ score was associated with an increased likelihood of reporting male gender relative to female (odds ratio = 1.08; 95% confidence interval = 1.037-1.136). An increase in AQ score was also associated with an increased likelihood of respondents their gender as non-binary relative to female (odds ratio = 1.183; 95% confidence interval = 1.124-1.245).

Conclusions:  We replicate the association between higher ASD traits in males relative to females and extend the association to adults who self-identify as non-binary.  This is consistent with reports from clinical samples and suggests, further, the broader experience of non-binary gender is also associated with higher reported autistic traits.  The causes of this association are unclear, but suggestions have included: prenatal androgen exposure and interpreting a problem relating to others, and associated feelings difference, as a problem of being in the wrong gender. Differences in interoception reported in ASD may also be implicated in gender dysphoria.  It is also possible that the association we find is because everyday social interactions experienced by individuals identifying with non-binary gender results in increased reporting of autistic traits.