Women Get Autism Too

Gender differences in autism – research review

Autism Spectrum Disorder (ASD) is a broad, complex and still, apparently, relatively unfathomable condition, although our knowledge and understanding about it are constantly increasing. Recent data from the Centers for Disease Control and Prevention in the US suggest that 1 in 68 people have an ASD (Autism and Developmental Disabilities Monitoring Network 2014). We are learning that autism is a more diverse disorder than was originally described by Dr Kanner more than 70 years ago. Yet despite all the recent breakthroughs in research, we still can’t explain it fully; we’re not even close. We know even less about whether and/or how gender affects individuals with ASD. Research that investigates and differentiates the male and female profile, presentation and experience of autism is fairly scarce, thereby perpetuating the myth that autism is a significantly male-dominated condition, or alternatively that gender is simply not a consideration.

Although the development of the diagnostic profile was based almost entirely on a male phenotype (Kopp and Gillberg 1992), it is important to note that mention was made of differences in the presentation of female autism by both Leo Kanner and Hans Asperger in the 1940s – perhaps it was not considered to be significant at the time. Kanner (1943) had noted that one girl did not understand games and pretended to be a dog, walking on all fours and making dog noises. Despite evidence that suggests that male and female brains in the general population have a different neurological profile (McCarthy et al. 2012; Ruigrok et al. 2014), it appears that researchers have hitherto thought that autism somehow overrides these differences and rather than having the neurological physiology and cognitive profile of a ‘man’ or ‘woman’, one is simply ‘autistic’. Personally, I think this unlikely, but this is how it has largely been viewed until fairly recently.

There appear to be several threads for discussion here, which current research has investigated to some degree:

•Is there a measurable and/or observable difference – biologically, neurologically, behaviourally and/or cognitively – between males and females with autism?

•How does that difference manifest cognitively and behaviourally?

•Do the current diagnostic criteria take into account any identified differences and accurately identify and diagnose males and females with autism?

•If not, is there potential for under-diagnosis of females?

•What do we need to do about any different manifestation in terms of diagnostic tools, diagnostic understanding and support?

Due to the paucity of women diagnosed with autism, it is difficult to find sufficient numbers of age- and developmentally matched females to participate in research that might show significant findings (Mandy et al. 2012). Put simply, if girls are not getting diagnosed because the criteria do not pick them up (because they weren’t considered in the creation of the criteria), there will be fewer diagnosed and included in research; research will therefore draw conclusions from smaller samples of the females who do match the male criteria…and on it goes…

Many esteemed individuals working in the autism field have commented and considered female ASD as worthy of differentiation from the standard profile; they include Tony Attwood, Simon Baron-Cohen, Svenny Kopp, Lorna Wing and Judith Gould, all of whom have put their substantial professional weight behind acknowledging and improving understanding and diagnosis for women on the autism spectrum. Practitioners are gathering data and developing lists of characteristics seen in women and girls (Marshall 2014). But a broader academic knowledge-base is thin on the ground. In contrast, the autobiographical and anecdotal published works on autism have a plethora of women with autism as contributors: Temple Grandin, Liane Holliday Willey, Wenn Lawson (formerly Wendy Lawson), Rudy Simone and Donna Williams, to name but a few. It seems that the women themselves have the words and the desire to speak out about their lives, but the ‘experts’ are not doing so to the same degree on their behalf in a scientific or practitioner context. These women are well-known names and faces in the autism world – confident public speakers and educators, sharing their unique perspective – but until now they have been considered largely indistinguishable from the men.

There are several different views that are worthy of consideration when discussing the male–female autism issue (Lai et al. 2011). One is that male and female autism may differ on a neurological and/or cognitive level (Carter et al. 2007). Another suggests that perhaps there are fewer females with autism because they are somehow protected against developing the condition (Jacquemont et al. 2014; Volkmar, Szatmari and Sparrow 1993), which may also explain why many neurodevelopmental conditions appear to affect more males than females (Zahn-Waxler, Shirtcliff and Marceau 2008). We also have to consider that the tools used for identifying autism may have a male gender bias and therefore are not ‘picking up’ female autism if it has a distinctly separate behavioural quality. ASDs are currently diagnosed behaviourally and observationally, therefore despite any potential similarities or differences on a cognitive or neurodevelopmental level, it may simply be that females present their autism differently through behaviours that are not included in the current diagnostic criteria (Kopp and Gillberg 1992). If this is the case, it is likely that clinicians making diagnoses need to view the current criteria more broadly to ensure that female behaviour is considered. More importantly, it may be necessary to develop a new set of diagnostic criteria that fully includes the female presentation of autism.Neurological profile

While sex differences in autism have been largely ignored (Bölte et al. 2011), there is a small amount of work that has considered neurological perspectives of differences in brain development and functioning as potential factors in explaining male and female autism. Lai et al. (2013) found that aspects of brain neuro-anatomy are sex dependent. Furthermore, this study also found minimal overlap between the neuro-anatomical features of males and females with autism, suggesting that males and females with autism may actually be neurally and cognitively distinct. Craig et al. (2007) found differences in the density of grey and white matter in areas of the brain linked to social behaviour deficits.

For some time there has been a general view that females with autism are somehow more impaired than males with autism and that perhaps the neural or cognitive ‘damage’ that has to occur in order for a female to get autism needs to be greater than for a male. One study found that girls with ‘high-functioning’ autism who attended a clinic were ‘more neuro-cognitively affected’ (Nyden, Hjelmquist and Gilberg 2000, p.185) than boys with the same diagnosis attending the same clinic. The girls were seen to have more ‘extensive deficits’ in theory of mind and executive functioning than the boys. The study suggests that girls attending the clinic may be in greater need of support than the boys (Nyden et al. 2000). Volkmar et al. (1993) found that in IQ testing, larger numbers of females with autism were at the lower end of the IQ scale and had learning disabilities than males, which appears to support the suggestion above that in general autism needs to be ‘worse’ for it to manifest in women, with fewer women appearing in the intellectually higher-functioning category of autism. Other work has similar findings (Lord, Schopler and Revicki 1982; Tsai and Beisler 1983; Wing 1981).

One of the most powerful theories in autism, which has had an impact on both public perception and sex differences, is the work of Professor Simon Baron-Cohen and other researchers from the Autism Research Centre in Cambridge, UK, and relates to levels of testosterone, androgen and other neurobiologically occurring chemicals. Several studies focus on the ‘extreme male brain theory’ of autism developed by Baron-Cohen (2002) and how this manifests in women. Baron-Cohen’s original work did not largely differentiate between males and females with autism, but he and others have since taken this further. The concept of masculinised behaviour in females with autism is one of the more researched areas in gender difference, which has had mixed reactions from the autism community. According to androgen theory of autism (Ingudomnukul et al. 2007), autism is partly caused by elevated foetal testosterone levels. These levels are correlated both positively and inversely with a number of autistic characteristics, including eye contact, vocabulary and social relationships. Further research revisiting these ideas (Bejerot et al. 2012) confirms that testosterone levels in women with autism were higher than control samples and that these women displayed more masculinised characteristics. It also found that men with autism presented more feminised characteristics, indicating that rather than women with autism being more masculinised per se, both genders may be more androgynous and represent a ‘gender defiant disorder’ (Bejerot et al. 2012, p.9). They go on to suggest that, ‘gender incoherence in individuals with ASD is to be expected and should be regarded as one reflection of the wide autism phenotype’ (Bejerot et al. 2012, p.9). The notion that females with autism present a profile – physically, cognitively and/or behaviourally – less stereotypically feminised than that of neurotypical (NT) females is one that will be explored more fully in other chapters of this book as it is something recognised anecdotally by many women with autism I have met. The ‘tomboy’ is recognised as a potential diagnostic indicator, although not all women with autism fit this profile.

Behavioural manifestation

There has not been a great deal of research into gender-differentiated behaviour manifestation, and certainly no large-scale studies (Giarelli et al. 2010). Anecdotally, as we will see later in this book, professional and personal accounts testify to distinct differences in visible presentation between males and females with autism, but this has not been quantified or discussed formally to any great degree. Research that has taken place so far has focused on very specific elements of the current diagnostic criteria when comparing gender performance. I have outlined a few studies below by way of illustration. Available research in this area is very hard to find, simply because it is scarce.

Knickmeyer, Wheelwright and Baron-Cohen (2008) examined sex-typical play in girls with autism. Their research found that girls with autism did not show a preference for female-typical items when engaged in play that did not involve pretence. This may show evidence of the hypothesis of masculinisation in girls with autism; alternatively, the researchers suggest these girls may be less susceptible to social factors that influence toy selection by girls and boys. However, if this were the case, we may have expected them to find the same lack of preference for gender-based non-pretence toys in boys with autism, but this was not the case; boys showed a preference for sex-typical non-pretence toys. When looking at games that did involve pretence (imagination), boys with autism showed virtually no interest, whereas girls with autism were very much engaged in imaginary play, as per typically developing girls. Imagination, usually considered to be an area of difficulty in individuals with autism, may not be affected in the same way in girls as it is in boys, with girls often retreating into fantasy worlds as respite from the stress of being in the real one. The Knickmeyer et al. (2008) study also cites research that suggests that pretence may be a skill that is developed via nurture more in girls than in boys: parent–daughter pretence play is more likely than parent–son pretence play.

Another study looked specifically at one element of behaviour and compared response inhibition in males and females with autism (Lemon et al. 2011). Individuals were asked to respond to a light being switched on by pressing a button as quickly as possible. The study found that females with autism were significantly slower to stop responding (i.e. to inhibit their responses) than either males with autism or NT males and females. Males with autism did not show any difference in response inhibition times to NT males and females. This study only focuses on one small area of behaviour, but may suggest that females with autism have a different neurobehavioural profile to males with autism. The consequences of impaired inhibitory control include impulsiveness, risk-taking and general executive dysfunction, including planning and decision-making. This may also impact on other social difficulties, such as appropriate behavioural responses, particularly when under stress (Lemon et al. 2011).

Carter et al. (2007) looked at sex differences in toddlers with autism and found that girls scored more highly in visual reception than boys, while boys scored more highly in language, motor and social competence than girls. They report statistically significant cognitive and development profile differences between girls and boys with autism aged 1–3 years. A study of a similar age group (Hartley and Sikora 2009) found many parallels in the male and female profile, but some ‘subtle but potentially important differences between the male and female ASD phenotype’ (p.179). Boys showed more repetitive, restricted behaviours than girls, and girls showed more communication deficits, sleep problems and anxiety. When looking at an older age group (3–18 years), teachers reported that males with autism externalise their social problems more readily than females and therefore are more likely to be identified as needing support in the classroom (Mandy et al. 2012). This work also agreed with Hartley and Sikora (2009) that there were subtle differences between the sexes. They found that girls had more ‘emotional problems’ (Mandy et al. 2012, p.1310) and better fine motor skills. As with other studies, restricted, repetitive behaviour was seen less in girls than in boys.

Bölte et al. (2011) found that executive functioning was scored more highly by girls with autism, and attention to detail was scored more highly in boys with autism. This may add weight to the suggestion that boys with autism show more stereotypical ritualised behaviours than girls with autism (Carter et al. 2007). It may also support the idea that women with autism have more ‘life events’ than men with autism (van Wijngaarden-Cremers 2012) as their better executive functioning may enable them to be more active in life in general (this does not mean that their decision-making is always effective).

Rather than the ‘extreme autistic aloneness’ personality mentioned by Kanner (1943, p.242), Kopp and Gillberg (1992) saw girls in their study being more ‘clinging’ (p.96) to others, imitating their speech and movements without a deeper understanding of the silent laws of ordinary social interaction. They also saw more repetitive questioning and ‘almost constant use of language’ (p.97), which is not commonly expected in the typical (male) profile of autism. This verbosity can appear social and interactive, but may be, on examination, scripted, learned or largely self-centric in nature. Comments on females’ ability to appear socially typical and mask their autistic behaviours occur in the literature time and time again (Attwood 2007; Gould and Ashton-Smith 2011; Lai et al. 2011), but as yet, only a few professionals in the field are acting on this to improve the situation by changing diagnostic processes. Kopp and Gillberg’s (2011) later work found that the following behaviours were much more typical of girls with autism than boys: ‘avoids demands’, ‘very determined’, ‘careless with physical appearance’ and ‘interacts mostly with younger children’ (pp.2881–2882).

Lai et al. (2011) found that on comparing boys and girls with autism, both were ‘equally autistic’ (p.5) as children, but that as adults the females showed fewer social communication difficulties, suggesting that they may have learned compensatory strategies and may be more motivated to do so throughout their lives in order to appear more ‘socially typical’ (p.6). The men in this study had not followed the same trajectory into adulthood. This masking of autistic characteristics has been documented by others in the field (Attwood 2007; Gould and Ashton-Smith 2011) and will be discussed in much greater detail throughout this book. It is suggested that women with autism are able to apply the systematic nature of their autistic brain (Baron-Cohen 2002) to the study and replication of people skills in order to imitate and participate socially. However, the mechanical (rather than intuitive) basis of these strategies means that at times of stress, in unexpected situations or after a period of time, it may be impossible for them to be maintained (Lai et al. 2011). For some women this can mean that they present a very capable front that cannot be maintained beyond certain limits, after which it collapses (and sometimes so does she). What has not yet been measured is the toll that this socially typical facade takes on the individual over a period of time.

It has also been suggested that women with autism are better at masking their autistic features due to better self-awareness and self-referential abilities (Attwood 2007; Lai et al. 2011). This increased ability would support women’s reported improved understanding of what is required socially and how to meet these expectations. Self-reporting of autistic characteristics was also seen to be higher in women, despite fewer behaviours being observed when tested with clinical autism diagnostic tools (ADOs). This suggests that despite less obvious observable characteristics, women have the same (if not higher) perception of their autism, and its impact, as men. In relation to this, it has been suggested that in studies that require parent observation and evaluation of autistic characteristics, parents may have higher expectations of the social and communication behaviour of girls than they do of boys, which may skew any findings (McLennan, Lord and Schopler 1993).