The underdiagnosis of autism in girls is a story of gender inequality

Autism in women and girls is more prevalent than currently thought, researchers say, and intersects with a range of other experiences they face to compound problems
Oleksandr Rupeta/NurPhoto/Getty Images

Dena Gassner was 33 when she took her son to be checked out for autism. He was three years old and had been struggling to cope with ordinary situations, facing challenges with learning, expressing symptoms of distress and harming himself. She started searching for ways to help him, and began to notice that, in her own ways, she faced many of the same problems – in her struggle to manage routine tasks, in managing stress, and difficulty dealing with others.

It was then, after years of other diagnoses, from PTSD to ADHD, from clinical depression to OCD, and a cocktail of medications that proliferated as doctors attempted to counterbalance the effects of each, combining Ritalin with Lithium, Wellbutrin, narcolepsy medications and Valium, that Gassner, alongside her son, was diagnosed with autism. It was, she says, a huge relief – at last doctors had pieced together what had been assumed to be standalone problems to regard them as a single condition.

“It took a year and a half to detox from all the other medications,” she says, to begin to feel more 'normal'. Gassner had been sexually abused as a young person, which she attributes to failing to spot warning signs. This obscured more fundamental issues, causing physicians to focus on the “noise” it generated, but failing to spot the problems underneath: intellectual development out of step with her peers, organisational problems, and missing social cues.

With diagnosis, Gassner was able to meaningfully describe and control her condition. Before, she felt shame at her inability to “be productive” – she had grown up in a household where everything had to be put away neatly, and would break down in tears to announce that she was “so disorganised” as a way to excuse her inability to be similarly neat.

She now realises that she experiences a different form of memory processing. To Gassner, objects stop existing once they’ve been put away: she needs things laid out. It also helps to prepare scripts in her head before entering unfamiliar situations, which lessen the anxiety of having to explain herself to others in a language they will understand. Gassner’s experience is by no means unique. But why did it take so long for her to reach a diagnosis?

BSIP/UIG via Getty Images

Autism is an extremely complex condition which affects people in different ways. It manifests in difficulty reading and relating to others; restricted and repetitive behaviours; and in differences from the 'neurotypical' population in memory formation and in processing sensory information. There is at present no biological test for autism, but a test based on behavioural criteria.

At an incidence of 4:1 male to female, autism has primarily been seen as a male condition, and Hans Asperger, who first theorised the condition in 1938, initially thought it was exclusive to males. Autism rates have been rising as the condition has gradually been better understood, however: over the last eighteen years, diagnoses have risen threefold in the United States, according to the Center for Disease Control and Prevention.

With ‘higher-functioning’ forms of autism – a term that relates to higher IQs and greater ease adapting to the norms of the world we live in – the incidence is higher still, at 10:1 male to female, raising important questions about autism's relationship to sex and gender, and why it is that so few women have been seen to have autism.

Men do seem to have a higher predisposition to developing autism. Male foetuses seem to be more vulnerable to health problems in general. They have a higher mortality rate and are more prone to a range of health issues – ear infections, for instance. Meng-Chuan Lai, director of gender research at the Autism Research Centre in Cambridge, says this vulnerability may result from hormonal disturbances related to sex differentiation, immunological development, and development of the gut microbes – all mechanisms thought to relate to the development of autism.

Alycia Halladay, chief science officer of the Autism Science Foundation, proposes there may also be a ‘female protective effect’. “Girls with autism certainly have a higher genetic load,” Halladay says. “So you actually see more genetic mutations in girls with autism than you see with boys with autism,” implying that girls are protected genetically.

A lot of early research into autism was undertaken by Simon Baron-Cohen in the late 1980s. Baron-Cohen, who is director of the Autism Research Centre, is responsible for the ‘extreme male brain theory of autism’, suggesting that high levels of foetal testosterone – to which male infants are more exposed – predicted poor eye contact, poor social relationships and a lack of empathy, as well as a greater aptitude for working with patterns, systems and objects, as opposed to people.

The theory is that people with autism are late to develop a ‘theory of mind’ – assessing and responding to others’ emotional states – and have average to above average abilities understanding systems and machines with routine outputs, which are, according to Baron-Cohen, predominantly male characteristics. Autism, Baron-Cohen says, is an extreme expression of male biological predispositions.

In 2003, Baron-Cohen extended his autism research into a broader theory arguing that men and women are wired differently, and that these essential differences mean men are more attuned to the workings of systems and women to people and emotions.

“Leaving aside political correctness, there is compelling evidence for sexual dimorphism [binary difference between men and women] in the brain, cognition, and behavior... We conjecture that understanding sex differences in the general population has implications for understanding the causes of autism-spectrum conditions,” wrote Baron-Cohen along with coauthors Rebecca Knickmeyer and Matthew Belmonte in 2005. Baron-Cohen argues that males perform better in spatial navigation, where girls are more verbal earlier, and boys prefer playing with mechanical toys where girls elect to play with dolls.

Baron-Cohen says that his 2003 book The Essential Difference explained his “position was not an extreme biological determinist one, nor an extreme social determinist one... but is an interactionist and in [his] view moderate position.” That this terrain is fraught, he says, “speaks to the political sensitivities surrounding any gender research in psychology and neuroscience,” but affirms his position “that some gender differences in the mind and brain may be the result of an interaction between prenatal biology and experience.” Speaking to WIRED, Baron-Cohen said he "can understand this sensitivity, after centuries of historical discrimination against women, but we need to separate science from politics.”

His research has, however, come in for criticism. Neuroscientist Lise Eliot, who reviewed Baron-Cohen’s work on the extreme male brain theory and his theory of psychological sex differences in Pink Brain, Blue Brain, says the data to support the idea that testosterone determines brain differences in men and women “were extremely weak” – both in those with autism and in the wider population. “You’d also find a correlation between foetal testosterone and who’s wearing pink or blue Pampers, but that doesn’t mean the relationship is causal,” she writes.

Eliot, along with Sarah Richardson at Harvard University, questions the validity of investigating less complex mammals in the context of sex brain differences. In humans, disparities associated with gender “have a strong sociocultural component” which is extremely “challenging to model in animals,” if not impossible. Their research questions institutional requirements to investigate sex using animals, arguing it skews research and “entrenches the presumption that human brain and behavioral differences are largely biological in origin,” overlooking "potentially more powerful social, psychological, and cultural contributors".

Eliot draws attention to a widely-cited 2000 study by Baron-Cohen which looked at 102 newborn babies to see whether boys and girls spent the same amount of time looking at a mobile or a live human face. In this small study, boys spent 52 per cent of their time looking at the mobile and 46 looking at the face, while girls split their time 41 and 49 per cent between the mobile and the face. This was heralded as evidence “beyond reasonable doubt” that sex differences are innate. Yet other studies fly in the face of these findings, observing more time spent on faces by boys than girls, or more toy-time by both boys and girls.

Indeed, across the population, Eliot says, "the differences between men and women on average are very subtle, but there is this one major difference, which is that men’s bodies and brains on average are about ten per cent larger than females.” This difference in size, though, has been shown not to confer noticeable differences: “once you take account of brain size differences, almost all the other differences dissolve,” she says.

Eliot emphasises that research into autistic traits cannot be separated from wider considerations of how personality itself develops, especially in the context of gender, and that investigations into the role played by sex and apparent ‘biology’ are extremely difficult to undertake because of the intense plasticity of the human brain and the lasting impact of gendered socialisation which takes place from the moment babies arrive. “We say we’re looking at sex differences but what we’re really looking at, with boys and girls, men and women, are gender differences, because being born and labelled male or female changes one’s social and communicative experience from birth,” Eliot told WIRED.

For instance, it’s been documented that “mothers talk more to their daughters than to their sons,” which may contribute to girls’ better communicative development, and while there are differences in girls’ and boys’ eye-contact by four months old, there are none at birth, according to a 2004 study. “Early social development is rapid and intense and has – presumably – long-lasting implications if you think about the way the brain is built, and the way the initial connections are formed are going to lay the groundwork for all the later circuitry,” Eliot says.

The brain's synaptic density, for instance, quadruples in the first year of life, attesting to the sheer concentration of development in these early months and years. Because of how early they are born and how long their period of development relative to similar-sized mammals, human infants are extremely impressionable, depending upon intensive interaction for survival. Eliot’s work contends that disruption in these developmental process may provide a key insight into the development of autism, with implications for how we gender babies.

BSIP/UIG via Getty images

Despite certain differences between men and women, some scientists are unconvinced that they fully explain the disparity in autism diagnoses. Lai argues, for instance, that the perception that the condition mostly affects men has led to a bias in the diagnosis. Lise Eliot agrees: “From the outset, [autism] was defined as more prevalent in males. Furthermore the different way boys and girls are socialised means they express autism in different ways, which has led to diagnostic criteria that fails to spot some female patients.

Social camouflaging is a coping mechanism used by autistic children as a way of adapting to the ‘neurotypical’ world through mimicking behaviours, preparing ‘scripts,’ such as Gassner describes, and using other fitting-in strategies. Because girls experience socialisation which emphasises fitting in more in general, masking can often be more pronounced, making them more likely to evade diagnosis. “Women are taught to fit into situations. It isn’t acceptable for girls to be blunt,” Gassner says. Exploratory research into this camouflaging suggests that autistic women camouflage more than autistic men, that it is related to reduced wellbeing, and that it decreases as time passes after diagnosis.

Will Mandy, a clinical psychologist specialising in autism in girls and women at UCL, explains that where younger girls may have flown under the radar, their social world becomes immensely more complex around the age of 12-13. It’s common that additional mental health difficulties creep in for girls with autistic traits during this period, as “the demands being made on their social skills are outstripping their social capacities,” Mandy says.

The struggle to cope with this increasingly nuanced social world can be immensely stressful for girls with undiagnosed autism, and Mandy suggests this may be one of the reasons for the additional issues in terms of anxiety, OCD, ADHD, eating disorders and depression. It is striking, he says, that regularly across the scientific literature, at least a quarter of women who are suffering with anorexia nervosa appear to have high levels of undiagnosed autistic traits as well. In addition, there is some research suggesting autistic girls may be more vulnerable to sexual abuse, as Gassner's experience anecdotally corroborates.

Indeed, many women and girls who suspect they have autism have reported huge hurdles in accessing appropriate care. Mother and autism activist Mandy Chivers spent seven years fighting for doctors to take her seriously, and Lily was ten by the time doctors finally agreed on a diagnosis of autism.

Lily, who is now twelve, is finally in a school which specialises in helping autistic kids, but her road to this point has been marked by misunderstanding and mistreatment. At age six, Lily was locked in a cupboard by a teacher; in another school, which claimed to be able to serve autistic needs, a teacher asked Chivers to make her daughter promise not to have meltdowns in school, which are a common experience for autistic children when they feel overwhelmed.

Medical professionals refused to help Lily, apparently reluctant to give a diagnosis to a girl. Chivers says that people from the council regularly responded with, “oh, she’s pretty – oh, she might grow out of it – oh, she doesn’t really look autistic”. By contrast, Chivers’s son, who is also autistic, only took three years to be diagnosed. Lily kept being prescribed drugs to treat ADHD. “My daughter was so high she got about five days’ worth of words out in about two minutes,” Chivers says, but was told by the NHS that if they refused the drugs they’d be discharged from the system. One professional asked Chivers if she knew what “kids in special schools looked like” when she insisted Lily needed to be sent to a specialist school.

She is now in a “lovely school” exclusively for autistic kids where her needs are prioritised. But, her mother says, she is one of four girls and 55 boys in the school. Chivers was determined to address the inequality at the heart of Lily’s story: to find a way of supporting other parents struggling to get the appropriate care for their daughters.

Four and a half months ago, she set up FIGS – Fighting Inequality for Girls on the Spectrum. It now has 1850 members across the country, and works to advocate and raise awareness. Other online initiatives, such as the #doilookautisticyet hashtag, are helping to emphasise the stereotyping inherent in the present approach.

“I think we’ve exaggerated the gender difference – that actually, because we’ve systematically underestimated the true prevalence of autism in females,” many girls are going without the care they need, Mandy says. It’s time for a more accurate portrait of the condition in women and girls, which accounts for the way they are socialised. “I don’t want to see any more children being failed by a system that fails to recognise them, because the diagnostic tools fail to recognise the different presentations. It can’t go on,” Chivers says. “These children need to be supported, their needs nurtured, their strengths encouraged, and they will go on to be productive members of society.”

This article was originally published by WIRED UK