My aunt said I was an odd child, and she wasn’t altogether wrong, but I’ve come to understand a kinder explanation for the way I am in the world. I was the child standing alone in the playground, the one who didn’t have friends. Often lonely, at age 11 I made a conscious effort to copy other girls in the hope I would fit in. It set up a pattern that persisted throughout my entire life – trying hard and still feeling I was failing. It took a toll on my mental health and for many years I struggled with depression, anxiety and obsessive compulsive disorder (OCD).

I reflected on my childhood in personal therapy for over ten years, but could never make sense of it through the usual frameworks of attachment or trauma. It was only in my late 50s when I discovered I was autistic that things began to change. The diagnosis brought understanding and a powerful sense of belonging. It also brought a profound sense of loss for all the years of struggle and what might have been possible. I wrote Other Wise Hearts,1 a novel inspired by the real-life experiences of late identified autistics, in the hope that girls and women might see themselves reflected in the characters and that this might bring them insight and relief from suffering. Because sadly, autism in girls and young women is still being missed.

The male-model problem 

While evidence suggests the number of girls and women who are autistic is equal to that of the number of boys and men, diagnosis in early years is at a rate of one to four, meaning that three out of every four autistic girls are misdiagnosed or missed altogether at a crucial time in their development.2 The disparity is rooted in early research conducted largely by male clinicians, such as Leo Kanner and Hans Asperger. Their work focused predominantly on white boys, and this shaped the diagnostic criteria that underpin contemporary medical models of autism. 

Later theories, such as the ‘extreme male brain’, which suggested autistic minds are more analysing and systemising than empathising,3 reinforced male stereotypes. Autism became synonymous with the socially oblivious male, possessing savant-like abilities in mathematics or science with little emotional insight. 

Masking and the female phenotype 

It wasn’t until the late 2000s that investigation into a female presentation gained any real traction, with mainstream recognition only emerging after 2015. The presentation of autistic females does not fit male-based models. Autistic girls and women are more socially motivated, demonstrate fewer externalising behaviours and are more likely to experience internalising difficulties such as anxiety and depression. A key factor is masking, sometimes called ‘camouflaging’. Masking is an umbrella term that describes different strategies that autistic people use to try and blend in with non-autistic peers. It can include suppressing stimming behaviours and the adoption of behaviours learnt from observing others. This might be forcing eye contact, copying facial expressions, or moving and speaking in the way non-autistics are seen to do. 

Autistic girls and women often mask because they have learnt that their differences may be received negatively. Non-autistics can engage in similar behaviours to make a good impression socially. However, this social presentation differs from autistic masking, which is more likely to be effortful and challenging to identity and selfesteem. Additionally, when an individual believes they are unlikely to be successful in their attempts to find acceptance, they experience anxiety and this expectation of failure is more likely in the case of autism. It is perhaps ironic that for many autistic people, this strategy has had limited success and, despite continued efforts, relationships still break down. 

Two important concepts may help explain this experience. Recent studies show that neurotypical observers make rapid judgments of others based on subtle non-verbal and expressive cues, often within seconds of meeting them. These studies showed that neurotypicals rated autistic children and adults less favourably using these thin-slice judgments. Less favourable impressions included measures such as likeability and desire to interact further – meaning they would be less inclined to sit with, talk to or engage with an autistic individual.

Autistic people are often believed to lack social skills. However, research shows social difficulties between autistic and non-autistic people are bidirectional, meaning misunderstandings are mutual with non-autistics struggling just as much to understand autistics as the other way round. This is known as the double empathy problem, which suggests breakdown in social communication in cross-neurotype interactions arises from mismatch in styles and not because of an autistic deficit.

As one young woman I spoke to for my book said, ‘Masking is an exhausting self-preservation tool. It harms mental health and puts our lives at risk. How exactly do you deal with something that saves us and risks our life at the same time.’

For autistic girls and women, masking does not occur in a social vacuum. Gendered socialisation may have an impact, with some girls and women feeling an expectation to be socially accommodating, while also receiving powerful messages that friendship determines worth. This suggests that what appears as increased social motivation may not reflect intrinsic drive but social conditioning. 

Within this context, girls and women may feel pressure to interpret painful social experiences as misunderstanding rather than harm. Autistic girls and women may report being acutely aware of their social environment, registering the emotional tone of interactions, subtle exchanges of looks, and not only what is said, but what is left unsaid. As such, bullying can be insidious and difficult to identify or articulate. A lack of validation and acknowledgement of hurt can have a devastating impact on self-esteem and mental health. 

The mental health cost 

The psychological consequences can be profound. A very high proportion of autistic people experience at least one mental health problem. Sustained camouflaging, chronic social misunderstanding and repeated experiences of exclusion are associated with anxiety, depression, loneliness and autistic burnout. It is also strongly correlated with suicidality.6 As autistic people are more likely to experience victimisation, it is unsurprising many autistics also experience trauma. 

An overlap in symptoms can create diagnostic confusion. Hypervigilance, dysregulation, shutdown and avoidance may be caused by trauma but may also reflect autistic stress responses. This means, in some cases, trauma may be missed and, in others, autism may be overlooked. In my clinical experience, most autistic girls and women report traumatic experiences. A both/and approach can be helpful as it reduces the risk of behaviours being misattributed to parenting or family dynamics when they may instead reflect trauma, autistic stress responses, or both. 

Misdiagnosis contributes to autism in girls and women being missed. This extends to anxiety, depression, personality disorders and eating disorders. For example, emotional intensity, dysregulation and relational difficulties may be misinterpreted as a personality disorder, rather than understood in the context of autistic overwhelm, sensory stress, chronic masking and the double empathy problem. Eating disorders, particularly avoidant/restrictive food intake disorder (ARFID), may be rooted in autism rather than body image concerns, illustrating sensory sensitivities, a need for predictability around food and interoceptive differences. 

However, in most cases mental health problems are not intrinsic to autism but are socially mediated, arising from repeated experiences of relational trauma, exclusion, identity invalidation and systemic misunderstanding. 

Suicide risk 

Autistic people are at increased risk of suicidality, with up to 66% of autistics having thought about suicide in their lifetime and up to 35% having planned or attempted it.7 A large Swedish population-based study found autistic individuals without intellectual disability had more than a tenfold increased risk of death by suicide compared to the general population. Those without intellectual disability and autistic women are at most elevated risk.8 This risk may rise further in those described as twice exceptional (2E), that is autistic individuals with an IQ over 120, challenging the assumption that higher IQ is a protective factor. These findings have implications for safeguarding and the early recognition of autism, especially in girls. The need for improved recognition is supported by a growing concern that autistic traits may be overrepresented in those who die by suicide, implying lack of recognition may lead to missed opportunities to provide targeted support and save lives. 

Mental health problems and trauma can increase suicidal behaviour in autistic people but being autistic in itself is a risk factor. Or perhaps it would be more accurate to say, being autistic in a neurotypical world is a risk factor. 

Vulnerability 

The risk to mental health and implications for suicidality are not the only concerns. Autistic people are more vulnerable to victimisation including physical, sexual and psychological violence, traditional bullying and cyber bullying, with analyses showing up to 92% of autistic people will experience multiple forms of victimisation across their lifespan.9 Women are at significantly increased risk of violence, as are those who experience mental health problems, are economically disadvantaged, or belong to racial or gender minority groups.9 These realities make safeguarding especially important. Autistic traits may interact with the social environment in ways that increase vulnerability. For example: 

  • Difficulty in interpreting intentions and a desire for belonging may make autistic girls less likely to see potential dangers in a social situation
  • Fewer social networks, or social exclusion, may limit access to otherwise readily available information and opportunities to confide or find support
  • Masking may conceal distress, leading to underestimation of suicide risk
  • Prosopagnosia (often called ‘face blindness’), a condition in which there is difficulty recognising faces out of context, may contribute to stranger danger
  • Difficulties imagining and anticipating consequences may have implications around risk-taking behaviours and longer-term planning 
  • Sharing personal information may increase vulnerability to manipulation and bullying
  • Rejection may lead to intense feelings of distress or hopelessness. 

Therapeutic implications

As both a therapist and an autistic woman who has accessed personal therapy, I have experienced first-hand how easily therapeutic connection can be misunderstood across neurotypes. For me, eye contact requires masking. It is not a genuine expression of connection, but a performance of it. It created a painful paradox in therapy. The more authentic I tried to be, the more disengaged I appeared; the more I masked to preserve connection, the less authentic I felt. Situational mutism, where I couldn’t always access speech, added another layer. I sensed my quietness was misunderstood and I worried about being perceived as boring and difficult to connect with. Looking back, I can see my therapist and I were operating across a neurocognitive divide.

Despite experiencing more mental health difficulties, autistic adults are 25% less likely to show improvement from standard psychological therapies and 34% more likely to see deterioration in symptoms compared with matched non-autistic patients.10 There are things we can do to support autistic clients and help ensure they have a positive experience of therapy.

  1. Language is important. Many autistic people prefer identity-first language, describing themselves as ‘autistic’ because they view autism as an integral part of their identity.
  2. Reduce performative social demands. Eye contact, for example, may require masking, rather than signal genuine connection. 
  3. Remember the double empathy problem. Disconnection may reflect a neurocognitive mismatch rather than resistance or lack of engagement.
  4. Be mindful of therapeutic goals that aim to help children ‘fit in’. They may reinforce camouflaging rather than support the development of an authentic identity.
  5. Be wary of defaulting to attachment or parenting explanations. Autism may provide a better framework for understanding a child’s experiences.
  6. Be aware of masked distress. Apparent coping may conceal significant distress.
  7. Some autistic boys may present with social and behavioural traits more commonly observed in autistic girls.
  8. Prioritise safeguarding. Autistic girls and women face elevated risks of suicide, victimisation and exploitation.

The challenges facing autistic girls and women are great. Outdated medical models, masking and increased social motivation converge to increase their vulnerability to poorer physical and mental health outcomes. Autistic people who belong to other marginalised groups, including girls and women of colour, LGBTQ+ individuals and those from lowerincome backgrounds may experience compounded disadvantages due to intersecting forms of marginalisation. It is therefore essential to think about intersectionality when considering autistic experiences and to remain vigilant to the ways bias may shape understanding and support.

Recognition of autism can be transformative and early identification matters. It can be lifesaving. It also means autistic girls and young women get the best possible chance to make meaning of their social experiences and sensory realities, and in doing so, develop a secure sense of self-identity which will support them throughout their lifetime and allow them to thrive.

References

1 Graham H. Other wise hearts. London: Little Steps Publishing; 2026.
2 Fyfe C, Winell H, Dougherty J, Gutmann DH, Kolevzon A, Marrus N, Tedroff K, Turner TN, Weiss LA, Yip BHK, Yin W, Sandin S. Time trends in the male to female ratio for autism incidence: population based, prospectively collected, birth cohort study. BMJ 2026;392:e084164.
3 Baron-Cohen S. The extreme male brain theory of autism. Trends in Cognitive Sciences 2002; 6(6): 248–254.
4 Sasson NJ, Faso DJ, Nugent J, Lovell S, Kennedy DP, Grossman RB. Neurotypical peers are less willing to interact with those with autism based on thin slice judgments. Scientific Reports 2017; 7:40700.
5 Mitchell P, Sheppard E, Cassidy S. Autism and the double empathy problem: implications for development and mental health. British Journal of Developmental Psychology 2021; 39(1): 1–18.
6 Cassidy SA, Gould K, Townsend E, Pelton M, Robertson AE, Rodgers J. Is camouflaging autistic traits associated with suicidal thoughts and behaviours? Journal of Autism and Developmental Disorders 2020; 50(10): 3549–3562.
7 Newell V, Phillips L, Jones C, Townsend E, Richards C. A systematic review and meta-analysis of suicidality in autistic people without intellectual disability. Molecular Autism 2023; 14(1): 16.
8 Hirvikoski T, Boman M, Chen Q, D’Onofrio BM, Mittendorfer-Rutz E, Lichtenstein P, Bölte S, Larsson H. Individual risk and familial liability for suicide attempt and suicide in autism. Psychological Medicine 2020; 50(9): 1463–1474.
9 Cooke K, Ridgway K, Pecora L, Westrupp EM, Hedley D, Hooley M, Stokes MA. The prevalence and risk factors of autistic experiences of interpersonal violence: a systematic review and meta-analysis. Research Square 2022.
10 El Baou C, Bell G, Saunders R, Buckman JEJ, Mandy W, Dagnan D, O’Nions E, Pender R, Clements H, Pilling S, Richards M, John A, Stott J. Effectiveness of primary care psychological therapy services for treating depression and anxiety in autistic adults in England: a retrospective, matched, observational cohort study of national health-care records. The Lancet Psychiatry. 2023; 10(12): 944–954.