Autistic gender identity differences – Resisting the social schema?

The relationship between gender identities incongruent with assigned sex, and autism spectrum conditions (referred to collectively here as ‘ASC’) has predominantly been investigated by recruiting samples of gender variant... [ view full abstract ]

The relationship between gender identities incongruent with assigned sex, and autism spectrum conditions (referred to collectively here as ‘ASC’) has predominantly been investigated by recruiting samples of gender variant and transgender individuals through gender identity clinics (Pasterski et al, 2014; de Vries et al, 2010; Skagerberg, Di Ceglie & Carmichael, 2015) or the internet (Kristensen & Broome 2016, Jones et al, 2012). Numerous hypotheses intended to explain this correlation have been posited, from the controversial proposal of Williams, Allard and Sears (1996) that sensory differences and restricted and repetitive interests may lead males with autism to become preoccupied with a range of things which “happen to be predominantly feminine in nature” (and that this does not in fact constitute ‘true transgenderism’), to the rather more straightforward suggestion from Ansara and Hegarty (2011) that autistic peoples’ tendency to greater directness and below-typical concern for social norms simply increases the likelihood that they will disclose their identity despite the prevalence of cisgenderism and transphobia in society. The debate around the etiology of autism and the debate around gender in autism (including, but not limited to, trans* genders) intersect in significant ways, but the potential research benefits of that intersection are relatively un-tapped.

In this study we examine the rates of gender identities incongruent with assigned sex in a large sample of autistic adults, and whether and how these individuals’ autism severity (Autism Quotient; AQ) and sensory sensitivity... [ view full abstract ]

In this study we examine the rates of gender identities incongruent with assigned sex in a large sample of autistic adults, and whether and how these individuals’ autism severity (Autism Quotient; AQ) and sensory sensitivity (Sensory Perception Questionnaire; SPQ) factor scores differ from those of cisgender adults in the sample. Based on the prior investigations into this area, we expected both AQ and SPQ to be higher in the transgender section of the sample, and for transgender identities to be more prevalent, and less binary in this sample than in the population as a whole. (‘Binary’ in this context refers to identifying solely and entirely with one of the two traditionally defined genders, man or woman.)

675 adults (52% assigned female) with a diagnosis of autism completed the AQ, the SPQ and demographic questions including age, assigned gender, and gender identity. We categorized the participants as cisgender, binary-trans... [ view full abstract ]

675 adults (52% assigned female) with a diagnosis of autism completed the AQ, the SPQ and demographic questions including age, assigned gender, and gender identity. We categorized the participants as cisgender, binary-trans and nonbinary-trans based on their relative answers to the assigned gender and gender identity questions, and examined the distribution of such identities by assigned sex, the between-group differences in AQ and SPQ score with ANCOVA (age as the covariate) and the equivalency of their correlations within each group (Fisher r-to-z transformation).

The main outcome measures are: -The proportions of cisgender, binary-trans and non-binary-trans individuals in the sample, and within the assigned-male and assigned-female sections of the sample. -The between-group... [ view full abstract ]

The main outcome measures are:

-The proportions of cisgender, binary-trans and non-binary-trans individuals in the sample, and within the assigned-male and assigned-female sections of the sample.

-The between-group differences in rates of AQ and SPQ factor scores between cisgender and transgender participants.

-Whether there is a difference between the groups in the correlation between SPQ and AQ.

14.8% of the participants identified as trans; 22% of those assigned female at birth and 8% of those assigned male at birth were classified as transgender. Within those, only 4% and 12% respectively identified as... [ view full abstract ]

14.8% of the participants identified as trans; 22% of those assigned female at birth and 8% of those assigned male at birth were classified as transgender. Within those, only 4% and 12% respectively identified as binary.

Transgender participants scored significantly higher than cisgender participants on the AQ overall, and specifically within the ‘social skills’, ‘switching’, ‘imagination’ and ‘numbers and patterns’ factors (all factors except routine). Contrary to our prediction however, transgender participants scored lower than cisgender participants on the SPQ, with significant differences in the factors 'Vision' and 'Hearing'. Although these were small effects, they were consistently and highly significant (p’s<0.01). However, the correlation between AQ and SPQ was equivalent in both groups (z=0.02, p=0.984).

The high rate of gender nonconforming identities in our sample versus the general population supports the existing evidence that both autism diagnosis and AQ are strongly positively correlated with the likelihood of gender... [ view full abstract ]

The high rate of gender nonconforming identities in our sample versus the general population supports the existing evidence that both autism diagnosis and AQ are strongly positively correlated with the likelihood of gender nonconformity, and there are subtle but interesting differences between cisgender and transgender autistic people in the measured impact (by AQ and SPQ) of their condition. Furthermore, there is a reduced extent of sensory hypersensitivity in transgender autistic people than their cisgender counterparts. However, these differences operate independently. Additionally, the proportion of trans people identifying as non-binary appears to be high in this sample. Similarly, the rates of transgender identification between assigned genders is biased toward assigned-female. These findings prompt us to scepticism toward some prior hypotheses, and draw attention to possible interplays between the most explanatorily valuable aspects of others. For example, autistic hypersensitivity to stimuli may extend to internal stimuli including gender identity and/or somatosensory experiences associated with transgender identities, such as phantom genitalia (Ramachandran et al, 2007, 2008). These possibilities are explored in the context of the flattened-priors hypothesis of autism etiology (Pellicano & Burr, 2012).