Considering Autism Spectrum Disorder in Females

Getting a diagnosis

Until recently, commonly used diagnostic criteria, such as the Diagnostic and Statistical Manual of Mental Disorders (DSM) and International Classification of Diseases (ICD), have not considered sex differences at all (Gould and Ashton-Smith 2011; Lord et al. 1982). The updated DSM-5 criteria (American Psychiatric Association 2013) show significant changes to the general criteria for Autism Spectrum Disorders (ASDs) with the removal of independent diagnosis for Asperger Syndrome and other changes to the required indicators. Also, there is mention of sex differences for the first time: ‘In clinic samples females tend to be more likely to show accompanying intellectual disability, which suggests that girls without accompanying intellectual disability or language delays may go unrecognized, perhaps because of subtler manifestation of social and communication difficulties.’ (American Psychiatric Association 2013). Mandy (2013) adds his own comment on the potential implications of this:

In this sense, the architects of the DSM-5 have laid down a challenge to researchers: Provide an account of the female phenotype, so that clinicians can learn to better identify, and help, females on the autism spectrum [without an accompanying intellectual disability]. (Mandy 2013)

The DSM-5 also includes the following new specifier, which may assist the female cause: ‘C. Symptoms must be present in the early development period (but may not become fully manifest until social demands exceed limited capacities, or may be masked by learned strategies in later life’ (American Psychiatric Association 2013).

The explicit recognition that autism may not be ‘fully manifest’ until life becomes unmanageable for an individual will, it is to be hoped, aid those seeking adult diagnosis who may, up until now, have been told: ‘You can’t have autism; you would have been diagnosed by now if you did’. Furthermore, the assertion that ‘learned strategies’ may be adopted, and the inference that they should be considered as part of the evidence of potentially autistic indicators, should aid more accurate diagnosis in both males and females, but may have particular relevance to the documented masking abilities of females.

The new DSM-5 criteria place a greater emphasis on restricted, stereotyped behaviours being observed, yet some research has found that females with ASD do not exhibit these behaviours to the same degree as males, or at least not in the same way, which may exclude some females from receiving a diagnosis of ASD (Mandy et al. 2012). Some autism research professionals do not feel that non-gender-specific criteria are necessarily in need of changing to better identify autism in females, but that the onus must be on clinicians to interpret them accurately with a good understanding of how autism may present in females. When training clinicians to use the Diagnostic Interview for Social and Communication Disorders (DISCO) assessment tool, Dr Judith Gould encourages clinicians to examine the data gathered with a broader view:

The criteria relate to the condition which are not gender specific. The point is how do clinicians interpret the criteria rather than changing the criteria? […] The key is asking the right questions and sadly that only comes from experience and knowledge about the female presentation of the condition. Educating professionals as we do in our diagnostic training is one of the ways forward. (Gould 2014)

Current diagnostic tests and tools do not consider sex differences and only follow the diagnostic requirements as laid down by the DSM criteria. Frazier (2014) suggests that sex-specific norms for some diagnostic processes might be helpful. The teaching of these methods should provide gender-differentiated examples to give new clinicians the breadth of the presentation of the autistic profile. This book attempts to do the same by providing first-hand and professional examples of how autism shows itself in and affects females.

The diagnostic process itself takes on a number of forms depending on the age and/or intellectual ability of the individual and the culture of the diagnostic service or clinician. There is no definitive test for autism and no standard way of carrying out an assessment. The outcome of the diagnostic assessment has a subjective element and is based on the quality and quantity of evidence gathered and the experience and opinion of the clinician or clinicians involved. This explains why the knowledge and experience of the clinician are essential both in asking the right questions to elicit the necessary data and in interpreting that data to provide an accurate outcome – processes that are particularly relevant in the case of females. Eileen Riley-Hall (2012) provides a detailed account of the diagnostic process (in the US) that she went through with her two daughters. Following the diagnosis, professionals should have good knowledge of local referral points for services and links to information about understanding the diagnosis for both the individual and their family members. Diagnosis is only the beginning of the journey, not the end.

As previously mentioned, females are significantly less likely to receive a diagnosis than males (Giarelli et al. 2010; Russell et al. 2011), and by using current diagnostic methods, some females may ‘look’ less autistic, but not actually ‘be’ or ‘feel’ any less autistic (Lai et al. 2011). Girls and boys may be identified for diagnostic assessment at a similar age, but boys are more likely to receive an autism diagnosis whereas girls may receive a different diagnosis despite both sexes displaying markers/signs clinically associated with autism (Giarelli et al. 2010). It may be that clinicians who are expecting autism to be more common in males are attributing certain behaviours to autism in boys and attributing the same behaviours to different conditions in girls. Equally, it may be that clinicians are expecting males and females with autism to behave in the same way, even though boys’ and girls’ behaviour is not considered to be identical in the typically developing population (McCarthy et al. 2012; Ruigrok et al. 2014). Head, McGillivray and Stokes (2014) studied social and emotional skills in males and females with autism and compared them to those in typically developing peers aged 12–16 years. They found that females with autism scored higher than males with autism, and at a similar level to typically developing males, but lower than typically developing females. They concluded that this could partly explain the potential under-diagnosis of females as social presentation is not currently gender-differentiated in the diagnostic criteria. They suggest that if a female with autism presents as more highly functioning than a male with autism, their abilities may appear superficially typical (and similar to that of a typical male) and therefore not be considered as notable if the clinician is using the male autistic social skills profile.

In children without an autism diagnosis, it is more common for girls to receive diagnoses of general developmental delay or seizure disorder – staring spells and seizure-like activity were five times more commonly diagnosed in girls than in boys (Giarelli et al. 2010). Kopp and Gillberg (1992) report seeing many girls in their centre who ‘do not clinically present the picture usually associated with autistic disorder’ (p.90). These girls fulfil the criteria in childhood, but as they mature they present a considerably different profile to their male counterparts. Kopp and Gillberg (1992) suggest that some clinicians do not believe that these adult women could ever have fitted the full autistic profile earlier in life. A potential factor in these girls not presenting sufficiently to receive a positive diagnosis may be that the diagnostic criteria do not require responses to be made within a given time-frame. One of the noted issues for girls with autism is the length of time it takes them to respond socially – they may get there in the end, but are noticeably slower than same-age peers (Nichols, Moravcik and Tetenbaum 2009). Clinicians’ attention can be diverted by the presentation of other co-morbid mental health conditions in teenage and adult life – conditions such as anorexia and anxiety disorders, which may be part of the autism profile but may not lead automatically to considering autism as the causal factor (Kopp and Gillberg 1992; Simone 2010). W. Lawson (1998) describes how it took 25 years to be diagnosed with Asperger Syndrome after being wrongly diagnosed with schizophrenia. The majority of the women participants questioned for this book had experienced years of mental health difficulties (mainly anxiety) and interventions before receiving an autism diagnosis.

[It took] five years or so to finally pinpoint the condition after I scored highly positive in a preliminary written questionnaire regarding AS [Asperger Syndrome]. The difficulty came in my other issues obfuscating the root cause of my behaviourisms. I’ve been called anywhere from bipolar to schizophrenic and have had medication provided thusly to combat those symptoms. (Woman with autism)

I have been misdiagnosed most of my life, including hearing doctors tell my mother that I had emotional problems, that I was trying to get attention, that I had nervous problems or that I was neurotic. Later I was diagnosed with learning disabilities in high school. In college, a doctor who initially thought I was psychotic eventually diagnosed me as having Severe Neurosis with Schizoid tendencies. (Woman with autism)

Nichols et al. (2009) suggest that routes to eventual autism diagnosis for a female can include the following steps, and that a fuller knowledge and understanding of these can aid the clinician in considering autism when assessing the person and can potentially facilitate more accurate diagnosis:

•Previous diagnosis of another disorder or several disorders, including Attention Deficit Hyperactivity Disorder (ADHD), anxiety, depression, obsessive-compulsive disorder, eating disorder. Nichols et al. suggest that the older the girl, the more diagnoses she is likely to have collected.

•A diagnosis of social anxiety or general difficulty in social situations.

•Adult women with a previous diagnosis of schizophrenia or psychotic disorder.

•Another family member has an ASD diagnosis.

•Presentation or apparent deterioration of capacity to cope as adolescence approaches, where social relationships become more nuanced and complex.

•In adolescence, girls may not demonstrate typically ‘female’ social interests, such as fashion and relationships.

The process of change for the individual begins with diagnostic measures and how clinicians are implementing them. If the tools are not accurate in identifying girls, then clinicians will have a hard job accurately assessing them. It has been suggested that sex-specific behavioural or cognitive autism diagnostic criteria may be necessary to accurately reflect the true sex ratio (e.g., Lai et al. 2011). Until the time when new criteria are developed that encompass female ASD, clinicians will have to fill in the gaps themselves, by developing a broader understanding of what to look for and who to listen to. Tony Attwood has developed a Screening Test for parents to support identification of girls with autism (Attwood 2013). The questions focus on specifically female presentations of autism, such as gender toy preference, fantasy worlds and relationships with both people and animals. The test is used as part of a broader diagnostic assessment and at least begins to ask the right questions of parents in order to build a more accurate female profile. Marshall (2014) has developed (and continues to do so) a list of characteristics commonly seen in women and girls with autism. Parents are not always experts in autism and will not know what relevant information to present during a diagnostic assessment. It is the clinician’s role to prompt parents’ recall in specific areas. Many of the questions asked in the Attwood screening test are not those commonly associated with male autism and are therefore not elements that a parent would necessarily think worthy of mentioning.

Gould and Ashton-Smith (2011) also outline a number of key differences in autistic features between girls and boys and recommend a broader view of the diagnostic classification to be taken by clinicians. Items such as increased interest in reading fiction and immersing themselves in fantasy worlds (with set rules that can be followed) are mentioned as needing special consideration. It could be said that no clinician should diagnose (or rule out diagnosis of) autism in females without reference to sex-specific diagnostic materials such as these. It is to be hoped that, in the future, formalised diagnostic tools will become widespread, but in the meantime clinicians must make informed judgements and adjustments to existing tools where necessary. An inexperienced clinician’s ‘tick box’ approach to the autism diagnostic criteria may be very harmful to individuals with a different behavioural profile. An extensive knowledge of autism and face-to-face experience of a significant number of women with autism are necessary. Reading autobiographies, blogs, watching YouTube videos made by women with autism, and spending a day or two at a special school or women’s support group, will all help to build the real-life knowledge essential to be a competent clinician.

The neurobehavioural team [Asperger Syndrome diagnostic service] only takes referrals from consultant psychiatrists in the mental health team, and the mental health team didn’t understand autism so wouldn’t refer me as they thought that autism was only so-called ‘classic’ autism and didn’t understand that it was a spectrum. (Woman with autism)

Assuming an obvious physical presentation of ASD in a female (or indeed in anyone) may be the first mistake made by an inexperienced clinician. A full picture of childhood behaviours, self-reports and cognitive assessments should form part of the process. As a woman with autism, my own outward observable behaviour in a structured interview situation would not necessarily lead anyone to believe that I have autism, although I can articulate my experience as a person with autism very clearly. There can be a tendency to ignore family and self-reporting, and for the clinician to prioritise what they can ‘see’. It is essential to listen to what people say and not make assumptions based on their appearance and presentation. In my professional work, providing non-clinical assessments for individuals who suspect they may have autism, it is extremely rare to find a person who is incorrect about themselves. If an adult has been brought for an assessment by a partner or family member who suspects they have autism, a higher incidence of negative assessment is seen, but when the person investigates and instigates the process themselves, they are usually correct. I don’t say this because I am personally seeking to increase the numbers of the autistic population in order to claim world dominance; it is because people do not seek out diagnosis lightly. They have done their research and come to their conclusion solidly before approaching anyone for a diagnosis. As professionals, we must be respectful of the fact that these people are generally not time-wasters. They have more knowledge of themselves – and often their condition – than we do. When they seek diagnosis, frightening comments of professional ignorance are reported: ‘[UK state diagnostic service] were a disaster – openly admitted they didn’t understand autistic girls’ (Eaton 2012, p.11), and ‘[…] the psychologist told me she didn’t know girls could get it [autism]’ (Eaton 2012, p.11).

We moved to The Netherlands two years later and she was eventually diagnosed with Asperger Syndrome there at the local Children’s Hospital. We thought the Dutch were very liberal and forward thinking but they, to our horror, suggested that A would be better off institutionalised! (Parent of girl with autism)

Increasing the autism education of professionals and clinicians working at all points along the diagnostic and support route has to be the number one priority. In many cases, sadly, this is the stumbling block to accurate diagnosis and appropriate support. A study of mental health professionals in one region of England found that 79 per cent of people rated their own knowledge of autism as ‘limited’ or ‘fair’, and 59 per cent said that they had not received any autism training within the past two years. The same study asked service users with autism for their perceptions of Mental Health Services and found that only 17 per cent felt that staff understood their needs relating to having autism and only 23 per cent were satisfied with the mental health support they received (Impact Initiatives et al. 2013). Even those who had a positive experience in the end often appeared to have overcome an obstacle course to get there.