Broad Autism Phenotypic Traits and the Relationship to Sexual Orientation and Sexual Behavior

Journal of Autism and Developmental Disorders
December 2018, Volume 48, Issue 12

Broad Autism Phenotypic Traits and the Relationship to Sexual Orientation and Sexual Behavior


Individuals with higher levels of the broad autism phenotype (BAP) have some symptoms of autism spectrum disorder (ASD). Like individuals with ASD, people with higher-BAP may have fewer sexual experiences and may experience more same-sex attraction. This study measured BAP traits, sexual experiences, and sexual orientation in typically developing (TD) individuals to see if patterns of sexual behavior and sexual orientation in higher-BAP resemble those in ASD. Although BAP characteristics did not predict sexual experiences, one BAP measure significantly predicted sexual orientation, β = 0.22, t = 2.72, p = .007, controlling for demographic variables (R2 change = .04, F = 7.41, p = .007), showing individuals with higher-BAP also reported increased same-sex attraction. This finding supports the hypothesis that individuals with higher-BAP resemble ASD individuals in being more likely than TD individuals to experience same-sex attraction.

Broad autism phenotype Sexual behavior Sexual orientation Same-sex attraction

At the Intersection of Neurodiversity and Gender Diversity

Journal of Autism and Developmental Disorders
December 2018, Volume 48, Issue 12, pp 3973–3973

At the Intersection of Neurodiversity and Gender Diversity

Gerrit Ian van Schalkwyk

The intersection of gender diversity and neurodiversity is a focus of considerable interest, discussion, and research. As reflected in the contents of this special issue, literature has explored this intersection both clinically and conceptually. The opinions, integrations, and new data presented in this volume support the presence of a complex relationship between gender diversity and neurodiversity, but suggest that the common framing of gender dysphoria and autism spectrum disorder being ‘comorbid’ is limited. As described in the letter by Diane Ehrensaft, individuals who are both gender and neurodiverse bring to life new conceptualizations of gender as fluid and intersectional. This articulates with the hypothesis that gender diversity is an expected outcome in individuals who are less contingent on social information for identity development. Correspondingly, Jack Turban argues as to the potential for gender diverse youth to have reversible challenges with social functioning related to their experience of minority stress, potentially limiting the value of screening and diagnostic instruments to make strict, categorical statements.

The important work by Nobili et al. is unique in this literature by its inclusion of a control group—remarkably, autism caseness based on Autism Quotient scores was found to be comparable in both the transgender and cisgender group, although the subgroup of transgender individuals assigned female at birth were twice as likely to have clinically significant scores. Consistent with the argument made by Turban, this difference was related mainly to the presence of additional social difficulties in this group, highlighting both the limitations of screening instruments, and the need for more detailed conceptualizations that better capture the complexity of these youths’ experience and identity.

Where better to start than with the individuals themselves? Strang et al. describe a remarkable study that sought to understand the experience of gender in youth with diagnoses of Autism Spectrum Disorder through an in-depth qualitative approach. The participants are given a clear voice in a study that emphasized the challenges of both defining and communicating gender identity for neurodiverse youth, and the need for clinicians to create supportive spaces in which this can occur. Participants showed discomfort at simplistic reframing of their gender as a superficial ‘obsession,’ describing an altogether different and more complex character for their gender experiences.

Additional manuscripts in this issue describe a range of related issues, including sexual orientation, gender differences and ASD symptoms, and gender preferences in friendships for children with ASD. Overall, this special issue touches on key themes at the intersection of gender and ASD and should prove to be an invaluable resources to clinicians and researchers hoping to further their understanding in this area. It is clear that gender is an important issue in understanding neurodiversity, and that individuals who are neurodiverse may place importance on gender. Further work is needed to understand how these individuals may be best supported, but enough is known to provide a clear impetus for a clinical approach rooted in the knowledge that neurodiverse youth may have complex gender narratives that warrant affirmation and support.


Sexual orientation of 46, XX patients with congenital adrenal hyperplasia: a descriptive review.

J Pediatr Urol. 2018 Sep 19. pii: S1477-5131(18)30437-6. doi: 10.1016/j.jpurol.2018.08.004. [Epub ahead of print]

Sexual orientation of 46, XX patients with congenital adrenal hyperplasia: a descriptive review.

Gondim R1, Teles F1, Barroso U Jr2.

Author information ブラジル



Congenital adrenal hyperplasia (CAH) consists of a group of diseases characterized by an enzyme deficiency, particularly 21-hydroxylase deficiency. The condition may present in the simple virilizing form or in the salt-wasting form, with varying degrees of genital ambiguity. The non-heterosexual orientation is used in gender studies fields and includes bisexual, homosexual, Lesbians, gays, bissexuals, transgender, intersex, and others.


The objective of this study was to evaluate the frequency of non-heterosexual orientation in patients with CAH, in an attempt to identify biological factors possibly associated with this occurrence.


This was a descriptive review of observational studies on the sexual orientation of patients with CAH published between 1985 and 2016, as listed in PubMed.


Various studies have been performed to establish the relationship between CAH and non-heterosexual orientation. Non-heterosexual orientation is more prevalent in patients with more advanced Prader stages and those with the null and I2-splice genotype.


The prevalence of homosexuality and bisexuality is greater in patients with CAH in relation to the general population.

Copyright © 2018 Journal of Pediatric Urology Company. Published by Elsevier Ltd. All rights reserved.

Congenital adrenal hyperplasia; Prader; Sexual orientation


Gender Identity and Sexual Function in 46,XX Patients with Congenital Adrenal Hyperplasia Raised as Males.

Arch Sex Behav. 2018 Oct 5. doi: 10.1007/s10508-018-1299-z. [Epub ahead of print]

Gender Identity and Sexual Function in 46,XX Patients with Congenital Adrenal Hyperplasia Raised as Males.

Apóstolos RAC1,2, Canguçu-Campinho AK1, Lago R1, Costa ACS2, Oliveira LMB1, Toralles MB1, Barroso U Jr3,4.

Author information ブラジル


In individuals with congenital adrenal hyperplasia (CAH) and 46,XX karyotype, androgens produced by the adrenal glands during the intrauterine development promote virilization of the genitals, which may even result in the development of a well-formed penis. Some of these children with late diagnosis are registered as males after birth. After obtaining approval from the internal review board, we evaluated gender identity and sexual function in four 46,XX severely virilized patients with CAH, who were originally registered and raised as males, assisted in our Disorders of Sexual Development Clinic. The evaluation consisted of questionnaires to assess gender identity and sexual activity and interview with the multidisciplinary team that provides care for these patients. The patients underwent surgery to remove uterus, ovaries, and remaining vaginal structures, in addition to implantation of testicular prosthesis and correction of hypospadias, when necessary. All four patients have developed a clear male gender identity, and when evaluated for sexual activity, they have reported having erections, libido, orgasms, and sexual attraction to women only. Two of these 4 patients had satisfactory sexual intercourses when assessed using the International Index of Erectile Function questionnaire. The other two patients who never had sexual intercourse reported not having a partner for sexual activity; one is 18 years old, and the other is 14 years old. This study showed that this group of 46,XX severely virilized patients with CAH, registered and raised as males, adapted well to the assigned male gender, with satisfactory sexual function in patients who had sexual intercourse.


Congenital adrenal hyperplasia; Disorders of sex development; Gender identity; Intersex


Gender Identity in Autism: Sex Differences in Social Affiliation with Gender Groups.

J Autism Dev Disord. 2018 Apr 28. doi: 10.1007/s10803-018-3590-1.

Gender Identity in Autism: Sex Differences in Social Affiliation with Gender Groups.

Cooper K1, Smith LGE2, Russell AJ2.

1 Centre for Applied Autism Research, Department of Psychology, University of Bath, Claverton Down, Bath, BA2 7AY, UK. k.cooper@bath.ac.uk.
2 Centre for Applied Autism Research, Department of Psychology, University of Bath, Claverton Down, Bath, BA2 7AY, UK.


High rates of gender variance have been reported in autistic people, with higher variance in autistic females than males. The social component of gender identity may be affected, with autistic females experiencing lower identification with and feeling less positively about their gender groups than controls. We measured gender identification, gender self-esteem, and aspects of gender expression (masculinity and femininity) in autistic natal males and females, and controls (N = 486). We found that autistic people had lower gender identification and gender self-esteem than controls, and autistic natal females had lower gender identification than autistic natal males and natal female controls. In conclusion, autistic people, particularly natal females, had lower social identification with and more negative feelings about a gender group.

Adults; Autism; Gender; Self-esteem; Sex differences; Social identity

※ 22% (Dewinter et al. 2017) もしくは33% (Bejerot and Erikson 2014; George and Stokes 2017)の自閉症圏の生来女性が,自閉症圏の生来男性の8% (Dewinter et al. 2017)もしくは22% (George and Stokes 2017)と比べてジェンダーバリアンスが多い。

※ジェンダークリニックを訪れる人で自閉症の診断が特定される人は,一般人口の1%と比べて,5.5%から26% (Pasterski et al. 2014; Kaltiala-Heino et al. 2015)と顕著に高い。


Autism spectrum disorder (ASD, referred to as “autism” in this paper) is characterised by difficulties with social communication, alongside repetitive, restricted behaviours, interests or activities, as outlined in The Diagnostic and Statistical Manual of Mental Disorders (5th ed.; DSM–5; American Psychiatric Association 2013). Autism is reported to affect around 1% of the population, with a disproportionately higher number of natal males diagnosed with the condition compared to natal females (Baird et al. 2006). Some recent research has found sex differences in the autism phenotype (e.g. Mandy et al. 2012; Lai et al. 2011; Hiller et al. 2014) but results have been mixed regarding the nature of such sex differences. For example, a systematic review and meta-analysis found boys had higher levels of repetitive behaviours after the age of 6 years but that there were no sex differences in social communication across the lifespan in autism (Van Wijngaarden-Cremers et al. 2014).

The extreme male brain theory of autism (Baron-Cohen 2002) proposes that the cognitive profile of autistic people is characteristically ‘male’ which can be “defined psychometrically as those individuals in whom systemising is significantly better than empathising” (Baron-Cohen 2002, p. 248). This profile has been labelled as male due to evidence of sex differences in these cognitive abilities in the general population, with women outperforming men in empathising and men outperforming women in systemising skills (although these sex differences are modulated by cultural norms and individual motivation1). Physiological sex differences also support this theory; Ecker et al. (2017) found that natal females whose brain anatomy was more similar to typical natal males, based on measures of cortical thickness, were three times more likely to have autism than natal females with more typically female brain anatomy. Therefore autism has been described as a case of the extreme male brain. It is possible that receiving a diagnosis of autism, with traits perceived to be male, has implications for how autistic natal females feel about their biological sex and gender. There is some research evidence to suggest that autistic natal females may not identify as readily with conventional female gender norms. Qualitative research findings suggest that autistic girls and women prefer socialising with boys and men rather than women (Bargiela et al. 2016; Cridland et al. 2014), and do not readily identify with the construct of femininity (Kanfiszer et al. 2017). Gender variance is defined as gender identity or gender expression which does not conform to traditional masculine or feminine gender norms (see Table 1 for definitions relating to gender in this paper), and sex differences in the prevalence of gender variance in autism have been reported. Studies have reported 22% (Dewinter et al. 2017) and 33% (Bejerot and Erikson 2014; George and Stokes 2017) of autistic natal females are gender variant compared to 8% (Dewinter et al. 2017) and 22% (George and Stokes 2017) of natal males. Another study found no sex differences in gender variance in autistic children (Strang et al. 2014), although this result could be confounded by unequal group sizes, as this study compared 24 natal females to 123 natal males.

When an individual is distressed by the experience of their gender identity opposing their biological sex, this is termed gender dysphoria. Gender dysphoria is characterised in the Diagnostic and Statistical Manual of Mental Disorders (5th ed.; DSM–5; American Psychiatric Association 2013) by the presence of two of the following criteria; incongruence between one’s sex and expressed gender; a strong desire to transition away from one’s current sex; a strong desire to transition to the other sex; a desire to belong to and/or be treated as belonging to a different gender group; a conviction that one has the typical feelings and reactions of the other gender. Studies identifying autism diagnoses in those attending gender clinics have found rates of autism diagnoses between 5.5–26% (Pasterski et al. 2014; Kaltiala-Heino et al. 2015), notably higher than the 1% prevalence of autism more generally (e.g. Baird et al. 2006).

High rates of gender variance have been reported in both autistic natal males and females, suggesting that while natal females are more affected, autistic people as a group are more likely to express gender variance than typically developing controls. Studies investigating gender variance have found autistic people display less stereotypically male role behaviour compared to sex-matched typically developing controls (Stauder et al. 2011; Bejerot and Eriksson 2014). Studies using a range of methods have provided evidence for gender variance and gender dysphoria in autism across the sexes including case studies (Williams et al. 1996; Landen and Ramussen 1997; Mukaddes 2002; Perera et al. 2003; Kraemer et al. 2005; Tateno et al. 2008; Jacobs et al. 2014; Lemaire et al. 2014); studies identifying high rates of gender variance in participants with an autism diagnosis (Stauder et al. 2011; Bejerot and Eriksson 2014; Strang et al. 2014; Janssen et al. 2016; May et al. 2017; Dewinter et al. 2017; Kanfiszer et al. 2017); and studies of elevated levels of autistic traits or diagnoses in those attending gender clinics (de Vries et al. 2010; Jones et al. 2012; Pasterski et al. 2014; Skagerberg et al. 2015; Kaltiala-Heino et al. 2015; VanderLaan et al. 2015; Shumer et al. 2016); and narrative and systematic reviews of the literature (van Schalkwyk et al. 2015; Glidden et al. 2016; Van Der Miesen et al. 2016).

Gender dysphoria and gender variance can be conceptualised as being parts of the spectrum of gender identity, with all individual gender identities belonging somewhere on this spectrum. Gender identification is a distinct construct from gender identity, and relates to one’s social identity as a member of a gender group. We define gender identification (in the terms of social identity theory; Tajfel and Turner 1979) as psychological affiliation to a gender group (cf. Tajfel 1981; Table 1), for example, someone with the gender identity ‘woman’ strongly identifying with other women due to their shared gender group. The gender groups which are most dominant are those of ‘men’ and ‘women’, however gender is increasingly being conceptualised as a spectrum, rather than a binary construct. Therefore there are numerous gender groups an individual could belong to, and some individuals may consider that they do not belong to a gender group at all, but rather are “gender free”. It may well be that autistic people, who a more likely to be gender diverse, are less likely to feel a sense of affiliation to any gender group, be that “men”, “women”, “transgender people”, “nonbinary people” etc. This is because autistic people are more likely to belong to minority gender groups, with fewer members and therefore less access to other in-group members, which may reduce a sense of affiliation within groups. However it is also possible that individuals belonging to minority gender groups will affiliate more strongly and positively with in-group members in order to mitigate the effect of belonging to a stigmatised group, and so more research is needed to understand the relationship between autism, gender variance and social affiliation with a gender group. Research recruiting autistic participants, who are more likely to be gender diverse, should therefore measure social affiliation with any gender group, without constraining responses to gender normative groups, in order to accurately reflect social affiliations in a gender diverse community.

Positive feelings about a gender group, or gender self-esteem, can be measured through self-report scales of group self-esteem (e.g. Luhtanen and Crocker 1992). For example, the gender self-esteem of an individual who identifies as non-binary would relate to how positively they view the group ‘non-binary people’. No quantitative research has investigated how strongly autistic people identify with a gender group, and how positively they feel about a gender group.

A sense of social affiliation with a gender group has been shown to be positively associated with psychological well-being in typically developing people (Good and Sanchez 2010). This finding extends to those with varied gender identities; a study with typically developing transgender women showed that positive feelings about gender identity were correlated with improved psychological well-being (Sanchez and Vilain 2009). In autistic people a sense of social affiliation with other autistic people has been found to be related to improved psychological well-being (Cooper et al. 2017). Furthermore, George and Stokes (2018) found that autistic individuals with gender dysphoric traits had poorer psychological well-being (higher stress, depression, and anxiety and lower well-being) than autistic people without these traits. Therefore social affiliation with a gender group presents as an important construct to measure in autistic people, a group high in gender diversity, who are known to be vulnerable to mental health problems (Hofvander et al. 2009).

In terms of sex differences in gender identification in the general population, natal females and males tend to have equivalent scores in terms of gender identification, i.e. strength of affiliation with gender identity (e.g. Schmader 2002). Furthermore, natal females tend to feel more positively about their gender group (have higher gender self-esteem) than natal males (Barker 2009), or have equivalent gender self-esteem to natal males (Foels and Tomcho 2005). However, no studies to date have investigated how gender and related constructs are conceptualised by autistic people, and because of this knowledge gap, it is not known if there are sex differences in gender identification and gender self-esteem in autistic people.
Aims and Hypotheses

Given the evidence suggesting high rates of gender variance and dysphoria in autism, the purpose of this research was to investigate whether having autism affects the extent to which natal males and females identify with a gender group, and how positively they feel about these gender groups. In this article, we test the proposition that the autistic natal females will have higher self-perceived gender variance than autistic natal males (gender identity different to biological sex, and masculinity and femininity ratings incongruent with biological sex), which will affect how attached they feel to a gender group (gender identification), and the extent to which they feel positively about their gender group (gender self-esteem).

We hypothesised that autistic participants would have lower gender identification and gender self-esteem than typically developing controls. We also hypothesised that participants who are not gender congruent would have lower gender identification and gender self-esteem than gender congruent participants. Furthermore, autistic natal females would consider themselves to be significantly more masculine and less feminine than typically developing natal females. Finally, we predicted sex differences within the autism group; autistic natal females would have lower gender identification and gender self-esteem than autistic natal males.


Gender Congruence and Autism

Autistic participants were significantly more likely to be gender incongruent (defined as selecting an ‘other’ gender label or a gender label different to biological sex) than TD controls, χ2(1, N = 485) = 45.98, p < .001. Further, autistic females were significantly more likely to be gender incongruent than autistic males χ2(1, N = 219) = 150.24, p < .001. See Table 2. Examples of ‘other’ gender labels that participants chose included ‘androgynous’, ‘alien’, ‘gender neutral’, ‘gender fluid’ and ‘not sure’.
Gender Transition and Autism

Autistic participants were significantly more likely to have or be planning a gender transition than TD participants, χ2(1, N = 485) = 12.40, p < .001, and autistic females were significantly more likely to have or be planning gender transition than autistic males, χ2(1, N = 219) = 11.85, p = .001.
Gender Transition and Gender Congruence

Gender incongruent participants were significantly more likely to have or be planning a gender transition than gender congruent participants, χ2(1, N = 486) = 189.97, p < .001.
Gender Identification and Gender Self-Esteem

We conducted two hierarchical multiple regressions, entering the factors sex (male or female), autism (Autism or TD), and gender congruence (gender congruent or gender incongruent) at Step 1, and the sex*autism, sex*gendercongruence and autism*gendercongruence interaction terms at Step 2, and the three way interaction term at Step 3, predicting gender identification and gender esteem, respectively. See Tables 3 and 4 for means, SDs and correlations for these variables. In the first model, predicting gender identification, we found that autism was significantly negatively associated with gender identification, β = − .399, t(482) = − 10.07, p < .001, with autistic participants (M = 4.18, SD = 1.32) scoring significantly lower than TD controls (M = 5.48, SD = 1.03). Gender congruence was negatively associated with gender identification, β = − .298, t(482) = − 7.54, p < .001, with gender congruent participants (M = 5.09, SD = 1.12) having significantly higher gender identification than gender incongruent participants (M = 3.42, SD = 1.34). There was no overall difference between males (M = 4.81, SD = 1.27) and females (M = 4.97, SD = 1.39) on gender identification, β = − .046, t(482) = − 1.20, p = .229.



The purpose of this study was to investigate sex differences in autistic and TD participants in social affiliation and attachment with a gender group (gender identification) and positive feelings derived from a gender group (gender self-esteem). There was support for the hypothesis that autistic people would identify less with and feel more negatively about a gender group as there was a significant effect of diagnostic group on both gender identification and gender self-esteem, with autistic participants scoring lower on both measures compared to TD controls. There was support for the hypothesis that gender incongruent participants would have lower gender identification and self-esteem than gender congruent participants. There was also support for the hypothesis that autistic females would have higher masculinity and lower femininity compared to TD females. Finally, there was support for the hypothesis that autistic natal females would have lower gender identification and gender self-esteem than controls, because while TD females had higher gender identification than TD males, the opposite pattern was found in the autism group, with autistic females having significantly lower gender identification than autistic males. There was no sex*autism interaction for gender self-esteem, but autistic participants had lower gender self-esteem than typically developing people, and females had lower gender self-esteem than males, and so autistic females had the lowest gender self-esteem. Also of relevance were the significantly lower rates of gender congruence, and higher rates of gender transition in all autistic participants, which were found to a greater extent in autistic females. The findings suggest that even gender congruent autistic people have lower gender identification and gender self-esteem than typically developing people.

We found that autistic natal females are particularly prone to lower social affiliation to a gender group, and greater variance in their gender expression, with these participants reporting lower femininity and higher masculinity than autistic males. There are two possible interpretations of this finding; the higher gender diversity in autistic natal females make it more difficult for this group to select a gender group to identify with, or they have selected a gender group and struggle to identify with other members of it. The finding that natal females are particularly affected fits with qualitative studies which found that autistic women often prefer to socialise with men (Bargiela et al. 2016). Research focusing on the lived experiences of autistic natal females regarding the development of their gender identity and feelings about gender groups would help to further explain these findings.

This study extends the previous findings regarding sex differences in gender identity in autism, demonstrating that the social component of gender identity is also affected in autistic people. This suggests that the higher variance in gender identities found in autistic people is associated with a lower sense of affiliation with and more negative feelings about gender groups as compared to controls.

There is increasing evidence that autism is a condition with high levels of gender variance (Bejerot and Eriksson 2014), and that gender identity is variant in autistic males as well as females (Bejerot et al. 2012; George and Stokes 2017). The current study replicated these findings, showing high levels of gender transition and incongruence in the autistic participants compared to TD participants. While not an objective of the study, we also found higher rates of diversity in sexual orientation in autistic compared to TD participants. Interestingly, autistic males considered themselves significantly less masculine than TD males, in line with previous findings (Stauder et al. 2011; Bejerot and Eriksson 2014). This would not be expected within the extreme male brain theory, however masculinity refers to the social norms associated with the gender expression of males, and it could be that autistic males rated themselves as lower in masculinity due to their awareness of being different to TD males. Masculine gender expressions may be seen as a marker of ‘typical’ social behaviour, and so autistic males may report lower levels of masculinity due to an awareness of their differences in social communication compared to other males. Therefore, according to these results both autistic females and significantly, males, appeared to display gender variance. The results of this study further suggest that autistic people are less likely to identify with a gender group, and see their gender groups more negatively than TD individuals. This is significant because gender identification and gender self-esteem are associated with psychological well-being in the TD population (Good and Sanchez 2010; Sanchez and Vilain 2009). Furthermore, there is evidence that gender dysphoria is associated with poor mental health in autistic people (George and Stokes 2018), and social identity processes are relevant to psychological well-being in autistic people (Cooper et al. 2017), and so the low gender identification and self-esteem found in autistic participants could impact on quality of life in this group. However, we did not measure psychological well-being in this study and cannot draw conclusions regarding the link between gender identification and well-being. An alternative hypothesis is that autistic people are less constrained by the gender norms of the typically developing population, and the high levels of gender diversity in autistic people are in fact associated with improved mental health. Further research is needed to unpack the relationship between gender identification, gender self-esteem and psychological well-being in the autism community.

There are features of autism which may contribute to this greater rate of gender diversity and lower gender identification across the sexes, however there is limited research available about how autistic individuals think about gender as a concept as compared to typically developing individuals, and so the following theories may or may not relate to gender identity in this group. Deficits in self-categorisation may pose challenges to autistic individuals in identifying with a gender label (Skorich et al. 2016), as individuals who struggle to place themselves in social categories may develop more idiosyncratic gender identities compared to typically developing people who are more readily able to categorise themselves within a gender group. Deficits in social communication may result in less knowledge and understanding of gender norms, or a freedom from gender norms allowing for more diverse gender expression, and a lower level of social reciprocity and affiliation with members of a gender group. Biological factors may also play a part, with factors such as prenatal testosterone linked to the development of autism and gender diversity (Baron-Cohen et al. 2005). Furthermore, for autistic people who have non-traditional gender identities, accessing and socialising with minority gender groups may be a particular challenge, as these are smaller and therefore less accessible groups. Indeed, George and Stokes (2017) found significant associations between all subscales of the Autism Quotient (AQ) and a measure of gender dysphoria, the strongest associations being with the social and communication subscales of the AQ. This suggests that the social communication deficits present in autism may contribute to differences in gender identity, or vice versa, the experience of gender dysphoria could contribute to social difficulties. This fits with the current findings that social identification with a gender group was lower in autistic participants compared to controls. Furthermore, gender self-esteem is likely to be lower if one belongs to a group which is stigmatised by society, and unfortunately gender variant individuals are more likely to be bullied (Russell et al. 2011) and ostracised (Carter and McCloskey 1984). Autistic people are also at higher risk of being bullied than typically developing people (Schroeder et al. 2014), and so an autistic person with gender dysphoria must contend with two stigmatised identities. Further research is needed into the experiences of autistic people with a range of gender identities, to investigate how their gender identity developed in relation to others of the same or different genders, and how this has impacted on quality of life.

Our findings also show that autistic natal females had significantly more diverse gender identities compared to autistic males. Furthermore, they tended to identify less strongly with a gender group than autistic males. It was noteworthy that the higher gender identification for TD females as compared to TD males found in this study was inverted for the autistic females, who had lower gender identification than autistic males. The extreme male brain theory has highlighted biological factors related to the development of autism and the male sex, for example the high levels of prenatal testosterone in autism (Baron-Cohen et al. 2005). It is possible that these biological factors impact on the gender identity of females. There were high levels of gender incongruence in autistic natal females compared to males. However, autistic females who were not gender congruent were much more likely to identify outside of a gender binary, with just 7% identifying as male and 26.5% identifying with a non-binary ‘other’ identity. This suggests that while it is possible that biological factors associated with autism may impact gender identity in females more than in males, this does not lead to a higher masculine gender identity, but to individuals identifying outside of the gender binary. This is despite autistic females rating themselves as significantly more masculine than TD females.

One limitation of this study was the possible bias in the sample. Participants self-identified as having a formal diagnosis of autism, and so it is possible that some participants did not have a diagnosis from a medical professional. Furthermore, the average age of ASD diagnosis was in early adulthood, and there was an unusually high rate of gender and sexual orientation diversity and mental health problems in both the autism and TD sample, suggestive of selection bias. There was a discrepancy in mental health diagnoses between autistic and TD participants; this may in part be related to differences in rates of gender dysphoria between these groups (George and Stokes 2018). Some clinicians have posited that there may be some false diagnoses of autism in those with gender dysphoria due to high levels of social anxiety in this group; it was beyond the scope of this research to investigate this. Another limitation relates to the gender identification and self-esteem measures. We asked participants to respond to items referring to their relationship with and thoughts about their “gender group”, and this was left open to their interpretation. Therefore for those participants who were in the process of transitioning between gender groups, it is not clear which gender group they responded for. However, gender congruence was included in the analysis as an independent variable, and we still found that the sex*autism interaction term was a significant predictor in three of four analyses. Furthermore, social affiliation to any gender group is of interest, given the findings that belonging to a wide range of social groups has positive effects on physical and psychological well-being in the typically developing population (Jetten et al. 2012), and that affiliation with other autistic people is associated with higher psychological well-being in autistic people (Cooper et al. 2017). However, the relationship between affiliation to gender groups and well-being in the autism community is unknown and would benefit from further investigation.

A strength of this study is its focus on the social processes related to, but distinct from gender identity, specifically gender identification and gender self-esteem, which to the authors’ knowledge have not yet been quantitatively investigated in autism. Furthermore, many autistic participants opted to describe their gender using the ‘other’ option, revealing that autistic people frequently identified outside of a gender binary, an option which previous surveys have not reported.

Given evidence of differences in gender identity in this group, it is important to consider the way that autistic people are supported with gender issues. Health services are likely to come into contact with autistic people who are gender diverse for numerous reasons, for instance for physical health care, mental health care, or for support around gender identity and transition. It is important that healthcare professionals working with this group are able to make reasonable adjustments to the care they give this group to ensure that treatment is accessible and effective. Furthermore, psychosocial interventions are likely to be useful for a group who have a lower sense of social affiliation to gender groups. One such intervention would be peer support groups of gender diverse autistic people where group members can share their stories and experiences of being autistic and gender diverse and support other group members. Future research should prioritise the voices of individuals who are members of both the autism and gender diverse communities through qualitative research exploring their experiences of belonging to these groups, and outcome research investigating the efficacy of psychosocial support groups and gender specific interventions for this group.

In sum, this study explored the social aspects of gender identity in autistic people, focusing on gender identification and gender self-esteem. The results corroborate those of previous studies which found high rates of gender variance in autism, with females particularly varied in their gender identities. It further extends these findings, suggesting that there are sex differences within the autism population, different to those found in the typically developing population, and that autistic individuals have weaker identification with gender groups, and feel less positively their gender groups than TD controls.


Sexuality in autism: hypersexual and paraphilic behavior in women and men with high-functioning autism spectrum disorder.

Sexuality in autism: hypersexual and paraphilic behavior in women and men with high-functioning autism spectrum disorder.

Dialogues Clin Neurosci. 2017 Dec;19(4):381-393

Authors: Schöttle D, Briken P, Tüscher O, Turner D


Like nonaffected adults, individuals with autism spectrum disorders (ASDs) show the entire range of sexual behaviors. However, due to the core symptoms of the disorder spectrum, including deficits in social skills, sensory hypo- and hypersensitivities, and repetitive behaviors, some ASD individuals might develop quantitatively above-average or nonnormative sexual behaviors and interests. After reviewing the relevant literature on sexuality in high-functioning ASD individuals, we present novel findings on the frequency of normal sexual behaviors and those about the assessment of hypersexual and paraphilic fantasies and behaviors in ASD individuals from our own study. Individuals with ASD seem to have more hypersexual and paraphilic fantasies and behaviors than general-population studies suggest. However, this inconsistency is mainly driven by the observations for male participants with ASD. This could be due to the fact that women with ASD are usually more socially adapted and show less ASD symptomatology. The peculiarities in sexual behaviors in ASD patients should be considered both for sexual education and in therapeutic approaches.


To our knowledge, this is the first study to explore gender-specific aspects of hypersexual and paraphilic fantasies and behaviors in a cohort of high-functioning individuals with ASD in comparison with a matched control group. Our main findings are that individuals with ASD show more hypersexual and paraphilic fantasies and behaviors than HCs.

Previous research suggested that in individuals with ASD, although mainly regarded as being heterosexual, 18 there were higher rates (up to 15% to 35%) of homosexual or bisexual orientation than in the non ASD population. 14,64 In the present study also, fewer individuals with ASD reported being heterosexual than HCs; however, it has to be noted that all HCs were heterosexual and are thus not comparable to the general population. In the Global Online Sexuality Survey, a total of 10% of participants indicated being homosexual. 65 Different assumptions have been made about the broader range of sexual orientation in the ASD population. Maybe gender is not that relevant in choosing a partner, due to limited access to romantic or sexual relationships and limited experience and sociosexual exchange with their peers. In combination with less sexual knowledge, this could lead to a restricted understanding of sexual orientation or preference. 33,35,37 Furthermore, there is evidence that ASD individuals are possibly more tolerant toward same-sex relationships, 15 and it could be possible that ASD individuals choose their sexual preferences more independently of what is socially accepted or demanded, maybe partly due to a lower sensitivity to social norms or gender roles. 15

Significantly more HCs than individuals with ASD reported being in a relationship with marked genderspecific differences. More women than men with ASD were in a relationship. The results of other studies examining gender differences in relationship status are inconclusive, but there is some evidence that although men desire dyadic relationships more than women, ASD women are more often in a romantic and sexual relationship. 11,31 This could be due to the ASD women's ability to call on more advanced coping strategies (eg, imitating the social skills of their non-ASD peers), leading to less impairment in social functioning. 33-36 Regarding the frequency of sexual behavior, women with ASD reported more solitary than person-oriented sexual behavior and less desire to have sexual intercourse with a partner than their non-ASD female counterparts. A similar pattern was found in ASD males, which is in line with other studies. 12,23,24,33

However, disregarding social norms together with the frequently found restricted social skills and the sensory hyposensitivities or hypersensitivities could also increase the risk for engaging in nonnormative or quantitatively above-average sexual behaviors. 22,38 Underscoring this assumption, we found that hypersexual behaviors were more frequently reported for ASD individuals than HCs; however, these differences were mainly driven by the male ASD patients, and no differences between the female groups were observed. On the basis of precise operationalization of hypersexual behaviors, previous studies have found prevalence estimates ranging from 3% to 12% for healthy male subjects. 66-68 In an online survey of almost 9000 German men, Klein and colleagues found a prevalence of hypersexual behaviors (defined as more than seven orgasms per week over a period of 1 month) of 12%. 69 Clearly, this indicates that more male ASD subjects in our studyshowed hypersexual behaviors than these population-based estimates. So far, only Fernandes and colleagues have assessed hypersexual behaviors in ASD individuals and found lower rates than we did. 70 Of the 55 high-functioning male ASD individuals assessed, 7% reported on hypersexual behaviors, defined as more than seven sexual activities per week, and 4% were engaged in sexual activities for more than 1 hour a day, which is clearly below the numbers found in the present study. However, Fernandes et al did not mention how they defined sexual activities, and it is possible that the participants in their study only rated dyadic sexual activities, explaining the lower number of hypersexual behaviors. 70 The possible causes of the higher rates of hypersexuality in ASD men remain unclear, but it can be hypothesized that they are a part of the repetitive behaviors or influenced by sensory peculiarities. Because we did not differentiate between person-oriented and self-oriented sexual behavior, the higher rate of hypersexual behaviors in the ASD men could also be an expression of excessive masturbation, which has been found in other studies and case reports. It was suggested that excessive masturbatory behavior could reflect the desire to be sexually active although not being able to achieve this because of problems engaging in a dyadic sexual relationship due to limited social skills. 14,46-48,52 With regard to women, much less research has been conducted about the frequency of hypersexual behaviors, and due to small sample sizes, prevalence estimates range from 4% to 40% in the general population. 60 In the German validation study of the HBI, 4.5% of the almost 1000 women included scored above the proposed hypersexuality cutoff. 59 As part of the DSM-5 field trials for hypersexual disorder, it was found that 5.3% of all patients seeking help at a specialized outpatient care center were women, 43 indicating that the rate of hypersexual behaviors might be much lower in women than in men. As female ASD patients seem to be better socially adapted and usually show less-pronounced ASD symptomatology (eg, less repetitive behaviors), it is not surprising that hypersexual behaviors in the present study were also found less frequently in female than in male ASD individuals.

So far, there are almost no existent systematic studies about paraphilias in the ASD population 64,70 ; most of the information comes from case studies. Moreover, almost all case studies addressed paraphilic behaviors in male ASD individuals with some kind of cognitive impairment; thus, comparison with findings from the present study is clearly limited. In the study of Fernandes and colleagues (to our knowledge the only previous study that addressed paraphilias in high-functioning ASD men), the paraphilias found most frequently were voyeurism and fetishism. 70 Voyeuristic fantasies and behaviors were also among the most frequently found paraphilias for ASD men and women in the present study. Furthermore, frequently reported paraphilias were masochistic and sadistic fantasies and behaviors. Again, this could be an expression of the pronounced hyposensitivity in the ASD population, indicating that such individuals need above-average stimulation to become sexually aroused. Furthermore, Fernandes et al found that the occurrence of a paraphilia was associated with more ASD symptoms, lower levels of intellectual ability, and lower levels of adaptive functioning, pointing out that lower cognitive abilities seem to be an important factor in the etiology of paraphilic fantasies and behaviors in ASD. 70 It can be hypothesized that awareness of social norms and behavioral self-control is even lower in ASD individuals with cognitive impairments, explaining the higher rate of paraphilic behaviors. Although many ASD individuals in the present study had paraphilic fantasies, considerably fewer individuals actually showed overt paraphilic behaviors, supporting the suggestion that high-functioning ASD individuals could have higher self-control abilities than ASD patients with cognitive impairments. Information on paraphilias in the general population is also scarce, with most of the studies involving men, mainly recruited in clinical or forensic settings. 71 In the general population, the prevalence rate of any paraphilia is assumed to be between 0.4% and 7.7%. 72-75 Also, using the QSEB, Ahlers et al found a rate of 59% for any paraphilic fantasies and a rate of 44% for any paraphilic behavior in their general-population sample of 367 German men, with the most common paraphilic fantasies being voyeuristic (35%), fetishistic (30%), and sadistic (22%) fantasies. 61 In the present study, especially for male ASD individuals, the rates of paraphilic fantasies and behaviors were higher than the prevalence estimates found in most of the general-population studies. Again, we found pronounced gender differences in the frequency of paraphilic fantasies and behaviors in our ASD population. A possible explanation for these differences could be that a stronger sex drive in ASD men could mediate the existence of paraphilias via a heightened energy in acting out their sexual interests or that those with a high sex drive more easily habituate to certain activities, thereby leading them to strive for novel activities. 71,76,77 Furthermore, hypersexuality could also lead to lower baseline sexual disgust or aversion toward paraphilic fantasies or behaviors clarifying the link between the higher rate of hypersexual, as well as paraphilic, behaviors. 77

The results of our study are limited because they are solely based on self-report, and one cannot be sure that all participants were diagnosed by a trained psychologist or psychiatrist. However, all ASD participants scored above the cutoff value of the German version of the AQ, ensuring that they showed pronounced ASD symptomatology. Furthermore, all participants were recruited through ASD self-help groups or ASD outpatient care centers, indicating that their contact with the medical system was due to their symptomatology. Our study results are also limited by the potential that individuals with a higher interest in sexuality-related issues, and perhaps also having more sexual problems, were more likely to volunteer to participate, thus affecting the study population. This could have led to an overestimation of the actual rate of hypersexual and paraphilic fantasies and behaviors in the ASD group. Nevertheless, if true, this should also have occurred in the HC group.

The present study is the first to examine hypersexual and paraphilic fantasies and behaviors in a large sample of high-functioning male and female ASD individuals in comparison with a matched control group, showing that although ASD individuals have a high interest in sexual behaviors, because of their specific impairments in social and romantic functioning, many of them also report some sexual peculiarities.