Recognising autism in girls within the education context: reflecting on the internal presentation and the diagnostic criteria

ABSTRACT

It is widely acknowledged that girls are less likely to receive an autism diagnosis compared to boys (Gould, J. 2017. “Towards Understanding the Under-Recognition of Girls and Women on the Autism Spectrum.” Autism 21 (6): 703–705), with growing recognition that girls are underserved by the current clinical diagnostic criteria, potentially due to differences in presentation (Duvekot, J., J. van der Ende, F. C. Verhulst, G. Slappendel, E. van Daalen, A. Maras, and K. Greaves-Lord. 2017. “Factors Influencing the Probability of a Diagnosis of Autism Spectrum Disorder in Girls Versus Boys.” Autism 21 (6): 646–658). Lack of awareness of the female profile of autism among teachers and other professionals can have implications in terms of recognition, diagnostic referrals, formal identification and support. A narrative review approach was adopted to review extant literature pertinent to sex differences in autism, using the diagnostic criteria outlined in DSM-5 as a framework (American Psychiatric Association. 2013. Diagnostic and Statistical Manual of Mental Disorders. 5th ed. Washington: American Psychiatric Association). The aim of this paper is to raise awareness of the varied presentation exhibited by some autistic girls. The findings reveal differences in the presentation of autism characteristics among girls with a more internal presentation apparent. The findings of the review indicate the need for revisions to diagnostic tools to reflect the more nuanced female presentation, increased training and guidance for professionals working with autistic girls and continued research. It is hoped that this review will further knowledge in this area to support the recognition of autism in females and mitigate the risk of females being missed, misdiagnosed or diagnosed later in life.

KEYWORDS:

• Autism
• female
• gender differences
• presentation
• teacher knowledge

Introduction

Autism Spectrum Disorder¹, henceforth autism, is characterised by differences in social communication and interaction and restricted interests and repetitive behaviours (APA 2013). Historically, autism was perceived as a ‘male disorder’ (Estrin et al. 2021) and up until recently it could be argued that the production of knowledge about autism could be regarded as the production of knowledge about male autism (Moore et al. 2022; Saxe 2017). This perspective has been challenged of late with attention drawn to the androcentrism of the diagnostic criteria (Kopp and Gillberg 1992) and the sampling bias in research (Kirkovski, Enticott, and Fitzgerald 2013).

Sex² differences in the prevalence of autism have been evident since the seminal reports into autism were published and the initial conceptualisations of autism were largely informed by research on male dominated samples (Asperger 1944; Kanner 1943). This has ultimately influenced the development of an autism profile based on male presentations. It is this profile that has informed the diagnostic criteria and screening tools used today (Carpenter, Happé, and Egerton 2019), however, it has been suggested that some autistic females do not fit this profile of autism (Wing 1981). As a result, it has been noted that the diagnostic process may lack validity in the identification of autistic girls (Cheslack-Postava and Jordan-Young 2012). It can be argued that a gender disparity exists in autism diagnosis with a review of epidemiological studies (Fombonne 2005) indicating male to female ratios ranging from 1.4:1 to 15.7:1, with 4:1 often cited (Fombonne 2009). More recent statistics estimate that the gender ratio may be closer to 3:1 (Loomes, Hull, and Mandy 2017). It is reported that girls with a co-occurring intellectual disability are more likely to be diagnosed with the ratio reducing to 2:1 (Fombonne 2003). Meanwhile, those with an average to higher IQ or with no language delays may go unidentified (APA 2013) with a more pronounced gender gap evident for this cohort of girls (Fombonne 1999; Fombonne 2009; Pisula et al. 2017; Tierney, Burns, and Kilbey 2016).

It is apparent that there exists a gender bias in autism diagnosis with autistic females being missed, misdiagnosed or diagnosed later in life (Carpenter, Happé, and Egerton 2019; Gould and Ashtown-Smith 2011). There is evidence to suggest that the disparity in diagnosis places the individual at risk of not getting access to the necessary support (Gould and Ashtown-Smith 2011), and can have significant implications for development and life outcomes (Taylor and DaWalt 2020). This paper will consider the gender bias in autism with reference to the education context. While a diagnosis is ultimately made by multidisciplinary teams, teachers play an important role in identifying children for a referral for a diagnosis (Farmer et al. 2003; O’Byrne 2018). Research states that autistic boys are referred for diagnosis 10 times more often than autistic girls (Wilkinson 2008). The aforementioned raises questions in relation to teacher awareness of how autism may present in girls. It is necessary to reflect on whether gaps in teacher knowledge and understanding are playing a role in the under identification of autistic girls.

There has been a positive shift in autism research to include a focus on sex/gender differences in autism in recent times (Mandy and Lai 2017). This research has enabled the development of a more comprehensive profile of autism in females (Wassell and Burke 2022), a profile that is not currently reflected in the DSM-5 or ICD-11 criteria. This may be one reason why girls remain under researched, under-identified and under-supported (Haney and Cullen 2017). Furthermore, it could be argued that there exists a discrepancy between developments in research and practice on the ground, which may account for inconsistencies in teacher knowledge.

This paper focuses on exploring an internal presentation of autism often linked to females diagnosed at Level 1 (APA 2013), who may display average to higher IQ with no language delays. The purpose of refining the discussion with reference to this cohort is that it is typically reflective of those who are educated in mainstream education settings. The overarching aim is to demonstrate the ways in which this internal presentation can be mapped onto the diagnostic criteria to better support educators recognise and support autistic girls in the education context.

Methodology

A narrative review approach was employed for the current paper as it was deemed appropriate to synthesise the extant literature on gender differences in autism and more specifically the presentation of autism in females (West and Martin 2023). It is hoped that this review could provide clarity for professionals, and researchers alike, in relation to how the female presentation of autism maps onto the diagnostic criteria.

The narrative review approach facilitated the inclusion of a broad range of literature types including research studies, literature reviews papers, theoretical and discussion papers, books and book chapters (Coughlan and Cronin 2021). An initial literature search was conducted for articles discussing sex/gender differences in autism and the female presentation of autism. Key terms included in the search were ‘Autism’, ‘Autism Spectrum Disorder’, ‘Autism Spectrum’, ‘girls’, ‘women’, ‘female phenotype’, ‘autism diagnosis’, ‘sex/gender differences’. Additional articles were identified through snowballing (Wohlin 2014) and consultation with other researchers. In keeping with a narrative review design, the authors did not apply strict eligibility criteria (Ferrari 2015), with the search and selection phase focused on the relevance of the articles (West and Martin 2023). Relevant content from books was included and autobiographical accounts for autistic individuals in books and/or book chapters were cited to ensure the inclusion of the autistic voice.

Findings

The literature search identified a body of research describing sex/gender differences in autism. The following will discuss the concept of a female profile of autism and the implications of gender-biased knowledge in relation to this. The body of the review will then explore the characteristics of autism in girls with reference to the diagnostic criteria.

Autism and girls: a different presentation

There is some debate around the notion of whether there exists a female phenotype of autism or are the differences in presentation due to the socialisation of females compared to males (Kreiser and White 2014). While it is beyond the scope of this paper to explore this debate, it is apparent that autistic girls can exhibit a different presentation of autism compared to autistic boys (Gould and Ashtown-Smith 2011; Lai et al. 2011). It has been posited that some autistic girls experience a more internal presentation of autism whereby their autistic traits are more inconspicuous compared to a more ‘stereotypical’ presentation (Kirkovski, Enticott, and Fitzgerald 2013; Rivet and Matson 2011; Wassell and Burke 2022). In addition to this, there is also evidence to suggest that autistic girls engage with camouflaging or masking to hide their autistic traits and/or use compensatory behaviours to fulfil neurotypical social ‘norms’ (Attwood 2007; Gould and Ashtown-Smith 2011). While this practice is not exclusive to autistic females, it has been reported as more prevalent in females compared to autistic males (Hull et al. 2017).

The characteristics associated with the internal presentation of autism and engagement with camouflaging undeniably plays a role in the ability of others to recognise and support autistic girls (Carpenter, Happé, and Egerton 2019), which may increase the likelihood of autistic girls being diagnosed later or in some cases missed altogether (Gould 2017) These concerns are grounded in research with reports highlighting that autistic females are consistently diagnosed later than their male counterparts (Begeer et al. 2012; Happé et al. 2016; Russell et al. 2022), and that more autistic women are being referred for diagnosis as adults compared to autistic men (Giarelli et al. 2010). It is worthwhile to reflect on these findings with reference to the education context, considering the role of educators in recognising autism in girls, gender-biased knowledge and its implications.

Gender biased knowledge in the education context

Recognition precedes diagnosis (Driver and Chester 2021) and as previously discussed, teachers play a significant role in recognising and referring students for autism diagnoses. Research examining teacher knowledge in relation to recognising autism in girls is sparse, however, two studies, both adopting a vignette method, aimed to investigate whether teachers could identify autism in girls (Ward, Curtis, and Seehagen 2022; Whitlock et al. 2020). Contradictory findings presented, however, the limitation of one study may account for the differences. A United Kingdom study by Whitlock et al. (2020) suggests that primary school educational staff were more likely to identify autism in males than females and appeared to demonstrate greater awareness of the male presentation. Moreover, the results indicated that the participants were better able to identify autistic girls displaying a male-based, traditional profile of autism compared to girls with an internal presentation. Interestingly, a second study based on a cohort of teachers in New Zealand suggested that the perceived gender of the child did not influence the identification of autism, indicating that teacher knowledge was not gender biased (Ward, Curtis, and Seehagen 2022). However, it was noted that the inclusion of externalising behaviours, typically associated with the male-based profile, in the vignettes may have skewed the results. These studies imply that while teachers may have some understanding of autism, their knowledge is somewhat influenced by a gendered profile of autism, and they may be less aware of the internal presentation of autism and lack knowledge of how autistic girls can meet the diagnostic profile in a different way to autistic boys. A third study conducted by Gray et al. (2021) used a self-report questionnaire to gather the views of SENCos³ in the United Kingdom on their knowledge of the presentation of autism in girls and their ability to recognise autism in girls and support their needs. The majority of SENCos were aware that autism presented differently in males and females, however, when detailing the characteristics they associated with autism in girls, a gendered knowledge base became apparent (Gray et al. 2021).

If educators are unable to recognise autistic girls who experience an internal presentation of autism, disparities in referral rates will persist and girls will continue be marginalised in terms of accessing a diagnosis and in turn, post-diagnostic supports. Moreover, a lack of teacher knowledge and understanding in relation to autism in girls may result in inadequate or non-existent support provision which may have negative implications in terms of opportunity to benefit from, experience success in and thrive in their education. These concerns are reflected in research conducted with autistic girls in relation to their educational experiences. Difficulties accessing the curriculum, transitions across and within settings, the interactional and sensory environments and feelings of anxiety are consistently reported in literature (Cridland et al. 2014; Goodall and MacKenzie 2019; Jarman and Rayner 2015; Moyse and Porter 2015; Sproston, Sedgewick, and Crane 2017; Tomlinson, Bond, and Hebron 2021). Moreover, the significance of teacher knowledge and understanding in relation to autism in girls was identified as a significant factor influencing school experiences (Tomlinson, Bond, and Hebron 2020), consistent with the concerns of the author and the aforementioned research.

Taken together, these findings indicate that gaps in teacher knowledge impact autistic girls’ school experiences in terms of recognition of characteristics, understanding of needs and the provision of support. The author concludes that specific guidance in terms of how the characteristics of autism present for autistic girls is required for teachers to ensure equitable and appropriate provision for autistic girls.

Interrogating the dyad of difference in autism diagnosis

Autism is diagnosed from observation of specific behaviours as detailed in the diagnostic criteria.⁴ These ‘behaviours’ exist in two core domains⁵, namely differences in social communication and interaction and restricted, repetitive patterns of behaviour, interests, or activities (RRBI).

Despite the disparity between the sexes in the prevalence of autism, it is only in recent years that research has placed a specific focus on sex/gender differences in autism (Lai et al. 2015). This research has revealed differences in the clinical presentation of autism between the sexes (Bond and Hebron 2019; Gould and Ashtown-Smith 2011; Lai et al. 2015). However, there is no consideration afforded to these potential differences in the diagnostic criteria (Giarelli et al. 2010; Gould and Ashtown-Smith 2011). It is clear that the historical male bias in autism knowledge together with the lack of consideration of sex differences in presentation in the DSM-5, have hugely influenced the diagnostic process for autistic females (Estrin et al. 2021; Mandy and Lai 2017). Moreover, there is evidence to suggest that some professionals, in both clinical and education contexts, are unaware of how autism may present in autistic girls compared to their male counterparts (Estrin et al. 2021; Gray et al. 2021; Kopp and Gillberg 1992; Ward, Curtis, and Seehagen 2022; Whitlock et al. 2020), adding a further barrier for autistic girls in terms of referral and diagnosis.

The following section will explore the way in which autistic girls can meet the different criterion within the two core domains of difference, as identified in the DSM-5 (APA 2013) to support recognition and identification of autism in girls.

Social communication and interaction

Social Communication and Interaction is the first domain of difference and individuals must experience differences in all three diagnostic criteria (Table 1 – Supplementary Material). There is a body of evidence to support the differences in social communication and interaction between autistic males and females, however, there are some conflicting findings (Andersson, Gillberg, and Miniscalco 2012; Kirkovski, Enticott, and Fitzgerald 2013; Rivet and Matson 2011; Van Wijngaarden-Cremers et al. 2014). The criteria within this domain will now be discussed individually with reference to extant literature on gender differences in autism.

Social and emotional reciprocity

Reciprocity refers to the back-and-forth action between communication partners required for social communication and interaction. A review of gender differences in autism carried out by Rivet and Matson (2011) claimed that autistic girls exhibit fewer difficulties with reciprocal interaction than boys, while a systematic review and meta-analysis by Van Wijngaarden-Cremers et al. (2014) found no significant differences in terms of social interactions and communication. However, considering that autistic girls with an average IQ are diagnosed at a later stage/misdiagnosed, it was suggested by the authors that the studies included in the meta-analysis may have missed this cohort and therefore a true conclusion could be that females with ASD show better social communication and social behaviour than males with autism (Van Wijngaarden-Cremers et al. 2014). This was also alluded to in another study of pre-school autistic children where it was suggested that perhaps girls with a higher IQ, better language and superficially better social skills will only be identified later in life (Andersson, Gillberg, and Miniscalco 2012). This theory has been supported more recently by findings from other researchers (Dean, Harwood, and Kasari 2017; Lai et al. 2017). A study by Dean, Harwood, and Kasari (2017) posited that the social challenges of girls appeared less obvious than those of boys, stating that autistic girls ‘looked like’ typically developing girls from the outside. It is worthwhile considering this idea with reference to the work of Lai et al. (2017) and their conceptualisation of camouflaging as a discrepancy between the ‘external’ behavioural presentation and the individual's internal status. This emphasises the importance of professionals communicating with autistic girls and learning more about how they feel in relation to social situations rather than observations alone. Therefore, despite the difficulties of autistic girls being less obvious, the social domain is identified as a core area of difficulty for autistic girls (Gould 2017). These finding allude to the ability of autistic girls to mask their difficulties with social communication and interaction which may result in them being less likely to be identified by professionals, in turn leading to underdiagnosis in this population (Lai et al. 2017). Furthermore, as autistic girls are often able to hide their difficulties with social reciprocity to onlookers, the importance of teachers being aware of the difference in the way boys and girls meet this criterion differently is significant (Hiller, Young, and Weber 2014).

It is widely reported that autistic girls are more socially motivated than autistic boys (Lai et al. 2015) and have demonstrated a greater ability to engage in reciprocal conversation (Hiller, Young, and Weber 2014). Girls may experience greater success in conversation as they are more likely to engage with interest sharing (Hiller, Young, and Weber 2014). Autistic females also appear to communicate well and can make good eye contact and engage with appropriate body language (Lawson 2017). This outward appearance of ‘good communication’ is grounded in neurotypical expectations and may not come naturally or automatically to autistic girls (Mandy and Lai 2017). However, it has been noted that autistic girls are at an advantage to boys here as they are experts in observing and copying what they think/feel they should be doing (Mandy and Lai 2017; Wassell and Burke 2022). Reflections provided by autistic female adults indicated that they demonstrated an ability to ‘learn’ the social rules of neurotypical interaction and ‘mimic’ the expected social behaviours (Milner et al. 2019). This ability to engage with social imitation is evidenced in the accounts of many autistic female adults as they reflect on their experience pre- and post-diagnosis (Baldwin and Costley 2016; Bargiela, Steward, and Mandy 2016; Tierney, Burns, and Kilbey 2016), while insights shared by autistic girls (Buckingham 2019; Salter 2019) detail the processes involved in meeting social expectations.

As with non-autistic individuals, natural personality differences also need to be considered in this domain appreciating that while some girls can be quite sociable and enjoy chatting, others may be passive and shy (Gillberg 2019; Wassell and Burke 2022). However, it is more widely reported that autistic girls often do not engage in meaningless chatting and often experience difficulties engaging with social niceties (Gould and Ashtown-Smith 2011). The complexity of masking, engagement in social imitation and individual preferences emphasise the importance of moving past surface-level observation and eliciting autistic girls’ inner experiences in order to understand differences in these areas. This, in turn, will help teachers to recognise and provide tailored support appropriate to the needs of autistic girls.

Developing and maintaining friendships

It has been acknowledged that autistic young people experience difficulties with developing and maintaining friendships (Carrington, Templeton, and Papinczak 2003). Given that gender differences in neurotypical friendships are well established with female friendships centred around emotional sharing and male friendships based on shared activities (Aukett, Ritchie, and Mill 1988), it is interesting to reflect on whether these gender differences play a role in how autistic individuals experience friendships. This criterion is extremely pertinent in the education context in terms of friendships within the classroom and the more unstructured period of lunchtime.

It is reported that autistic girls displayed significant difficulties in areas such as friendships, reciprocal interaction and group play but had fewer difficulties in communication compared to boys (Kirkovski, Enticott, and Fitzgerald 2013; Rivet and Matson 2011). It also appears that girls present with a different manifestation of friendship problems to boys (Hiller, Young, and Weber 2014). When taken together these findings indicate that while the ability to mask/imitate can facilitate the appearance of successful communication, the complexity of interactions involved in friendship, particularly female friendships, are significantly more challenging to navigate.

A study examining the social networks of school-aged children with and without autism, discovered that while boys were more overtly excluded, autistic girls appeared to be overlooked and not specifically accepted or rejected in social networks, raising concern in terms of the development of friendships (Dean et al. 2014). The above findings indicate that it may be easier to identify challenges in friendship in autistic boys compared to autistic girls, a point supported by further research by Dean, Harwood, and Kasari (2017). The social landscape of female friendships could be described as a protective factor with literature highlighting that the nature of these social groups concealed autistic girls by enabling them to float on the periphery (Dean, Harwood, and Kasari 2017; Kreiser and White 2014; Wassell and Burke 2022). Autistic girls tend to stay in close proximity to these social groups, something that may result in their difficulties going unnoticed by observers, particularly teachers in busy classrooms. This contrasts sharply with the social landscape of male social groups whereby researchers more easily identified boys with autism as they tended to wander alone separate from structured games or groups (Dean, Harwood, and Kasari 2017). This finding is significant in the education context as the class teacher may not always be the observer on yard and therefore it requires all teachers to be attuned to the more subtle challenges experienced by autistic girls.

While autistic girls’ experiences of friendship groups may go unnoticed by adults, their differences may not be hidden from their non-autistic peers (Dean, Harwood, and Kasari 2017). Analysis of time spent in joint engagement versus solitary, indicates that autistic girls stay near the social group and gain access to these groups through masking their social challenges. However, they do not appear to be able to maintain engagement with their typically developing peers, perhaps due to their difficulties with recognising and interpreting the social cues. This idea is reinforced by research conducted with autistic adults reflecting on their social experiences as adolescents who discussed the complexity of social rules and their implications for maintaining friendships (Baldwin and Costley 2016; Bargiela, Steward, and Mandy 2016). Despite this, research has clearly indicated that autistic girls’ desire friendship (Sedgewick et al. 2016; Sedgewick et al. 2019).

The above findings support the idea that autistic girls are more likely to be able to initiate interaction and indeed friendship than autistic boys (Hiller, Young, and Weber 2014). However, their difficulties appear to lie most substantially in the maintenance of these friendships (Gould and Ashtown-Smith 2011; Morewood, Tomlinson, and Bond 2019). While difficulties may not appear to be as obvious in the earlier years of the girls’ life, they become more apparent in adolescence as the nature of female friendships change throughout the school years (Attwood and Garnett 2019). Difficulties with friendship maintenance has been attributed to autistic girls’ needs for control over play, a challenge that is probably more prevalent in earlier childhood friendships (Hiller, Young, and Weber 2014). Additionally, non-autistic girls reach more complex social and linguistic stages earlier than boys (Barbu, Cabanes, and Le Maner-Idrissi 2011; Bond and Hebron 2019), and therefore it can be expected that the complexity of their friendships increases in line with this. This can have significant implications for autistic girls as they attempt to navigate these complex interactions. Complex interactions such as gossiping, exclusion and betrayal of trust are common in adolescent girls’ friendships, however, autistic girls experience more difficulties in relation to understanding and managing these interactions (Sedgewick et al. 2016). Further research by Sedgewick et al. (2019) revealed that autistic girls experienced more conflict than neurotypical girls and both autistic and non-autistic boys and this in turn resulted in anxiety for the girls, a finding that is echoed in first-hand accounts given by autistic females (Buckingham 2019; Milner et al. 2019). The demands of maintaining friendships are reinforced by autistic females who described having one or two close friends, acknowledging that friendships are arduous (Milner et al. 2019). This is consistent with observations from other researchers (Attwood and Garnett 2019; Gould and Ashtown-Smith 2011; Horlock 2019), and could be best explained by the Double Empathy theory (Milton 2012; Milton, Gurbuz, and López 2022).

Another consideration within the context of friendship is that of play differences and how they vary in autistic girls. It is reported that some girls demonstrate more appropriate play behaviour compared to boys, something that may be influenced by the presentation of RRBIs and/or their communicative abilities (Dean, Harwood, and Kasari 2017; Kirkovski, Enticott, and Fitzgerald 2013). However, other research reported that some autistic girls may prefer to play with boys as their play can be simpler and more logical (Attwood and Garnett 2019).

The concept of friendship is inextricable with education settings, and it is therefore imperative that teachers are alert to the varied profile exhibited by autistic girls in order to recognise difficulties that may be occurring in the social realm of the classroom and provide appropriate, contextually sensitive support.

Non-verbal communication

Differences reported in nonverbal communication correlate with the premise that girls experience a more internal presentation of autism (Wassell and Burke 2022), compared to the more overt behavioural presentation in boys (Hiller, Young, and Weber 2014). Results from one study by Hiller, Young, and Weber (2014) revealed that the skill of being able to integrate nonverbal and verbal gestures was predictive of being female and over one-third of girls in the study exhibited no difficulties in this area. Attwood and Garnett (2019) attends to this strength in autistic girls and attributes it to their capacity to imitate non-verbal actions in order to blend in with non-autistic individuals. This need to assimilate is often most prominent in friendship networks and therefore it has been reported that autistic girls can be more animated with their peers compared to when they are with family. This is probably done in an attempt to achieve social success.

Even though girls demonstrate a better use of social gestures, they experience similar difficulties in understanding and interpreting non-verbal behaviours as their male counterparts (Hiller, Young, and Weber 2014). Interpreting the subtleties in facial expression, voice intonation and reading body language were identified as potential difficulties in non-verbal communication for autistic girls (Kreiser and White 2014). Moreover, it was posited that difficulties in the area of non-verbal communication make it harder for autistic girls to interpret more subtle interactions such as eye-rolling, shared glances and giggling all of which are inextricably linked to adolescent female friendship groups (Dean, Harwood, and Kasari 2017). Understandably, these challenges may lead to social confusion which can support understanding of autistic girls’ difficulties with social interaction and more specifically maintaining friendships. Knowledge and understanding of these are important for teachers to assist in the identification and referral of students for diagnosis.

Insight from autistic individuals can help conceptualise the complexity of social interaction and communication differences. Autistic author, Holliday Willey (1999) juxtaposes her strengths in public speaking with her dislike of more intimate social situations, highlighting the sheer complexity of the domain of social interaction and communication and more specifically the stress associated with understanding and interpreting body language. In order to support girls, it is important that teachers are aware of the challenges autistic girls have in the area of social interaction and communication and understand the complexity of these differences for some individuals.

Restricted, repetitive patterns of behaviour, interests or activities

The second domain of difference refers to RRBIs. Specific information in relation to the criteria in this domain can be found in Table 2 in Supplementary Material. Interestingly, research has reported that the most consistent difference in presentation between genders occur in this domain (Duvekot et al. 2017; Knutsen et al. 2019; Kreiser and White 2014; Sipes et al. 2011; Van Wijngaarden-Cremers et al. 2014) with females displaying quantitatively reduced RRBIs overall (Allely 2019; Kirkovski, Enticott, and Fitzgerald 2013; Lai et al. 2015; Van Wijngaarden-Cremers et al. 2014). It is argued that RRBIs are more easily identified than difficulties in the domain of social interaction and communication (Allely 2019). However, the identification of characteristics in this domain is largely influenced by gender bias and the RRBIs of autistic girls are often not reflective of the traditional male-based presentation (Nowell, Jones, and Harrop 2019).

It should be considered that a milder presentation in this domain may in fact be mediated by the developmental quotient of the individual with those with an average developmental quotient exhibiting the least restricted, repetitive behaviours (Sipes et al. 2011). In light of this, an alternative perspective relating to the camouflaging hypothesis can be considered (Hull et al. 2017; Knutsen et al. 2019). Knutsen et al. (2019) examined sex differences in the RRBI domain between age- and intelligence quotient-matched female and male children diagnosed with autism using the largest known sample. This study revealed more similarities rather than differences between males and females. However, decreased repetitive interests and stereotyped behaviours were observed in younger females with lower support needs. This offers support for the phenomenon that females with an average-higher IQ are more motivated/effective at camouflaging their autistic traits (Attwood 2007; Gould and Ashtown-Smith 2011). This idea is further alluded to in the work of Hull et al. (2017) which noted that adults with no co-occurring intellectual disability can use compensatory strategies to camouflage their restricted, repetitive behaviours. However, the complexity and sophistication of this technique was acknowledged in that it requires high levels of self-awareness for individuals to recognise their RRBIs, develop a strategy to conceal them and finally perform the compensatory behaviours in the appropriate environments. Nevertheless, this offers a unique insight into the cohort relevant to this study. The latter debate questions whether the diagnostic instruments are sensitive to the varied presentation of repetitive behaviours in girls (Gould 2017). These findings, when taken together, raise concerns with regards to the identification and support of autistic girls and may go some way in explaining the under recognition of this group of individuals by teachers and other professionals. In light of this, a discussion of each criterion in the RRBI domain will follow.

Repetitive behaviours

There exists a body of research indicating that females demonstrate less unusual stereotyped and repetitive behaviours⁶ (Kreiser and White 2014; Rivet and Matson 2011; Van Wijngaarden-Cremers et al. 2014). This is further supported by findings of a study that considered the factors influencing the probability of a diagnosis of ASD in girls versus boys and found that restricted, repetitive behaviours were less strongly associated with autism in girls than boys (Duvekot et al. 2017). The aforementioned raises questions in relation to the awareness of professionals working with autistic girls and their ability to recognise what constitutes repetitive behaviours in this cohort.

Research by Hiller, Young, and Weber (2014) revealed that 27% of girls in their study did not meet this criterion compared to only 6% of boys. A further study surveyed clinicians involved in diagnosis and found that perceived sex-related differences were predominantly in the area of repetitive behaviours (Jamison et al. 2017). An emerging pattern between diagnosis age and the RRBI domain was also noted with clinicians reporting that females exhibit less ‘severe’ behaviours in relation to stereotyped or repetitive motor movements during the school-age/adolescent period (Jamison et al. 2017). Other research reflects on the diagnostic process proposing that it is the more passive, non-problematic behaviours observed in autistic girls compared to the overt behaviours exhibited by autistic boys that may lead to reduced rates of referral for diagnostic assessments by teachers (Kreiser and White 2014). This observation only further accentuates the importance of supporting teachers to recognise the varied autism profile associated with girls. These findings question whether the presentation of restricted, repetitive behaviours in girls is qualitatively different compared to boys, a hypothesis supported by observational clinician accounts (Attwood 2007; Bond and Hebron 2019; Gould and Ashtown-Smith 2011).

It has been noted that the repetitive behaviours of autistic girls may look very different to the stereotypical presentation associated with autistic boys i.e. lining up toys and engaging with echolalia (Hiller, Young, and Weber 2014; Wassell and Burke 2022). Instead, some autistic girls may engage in more inconspicuous repetitive behaviours for example re-reading the same book, re-watching the same film or re-doing homework. All of these behaviours can fulfil a role of comfort and predictability, both of which are associated with engagement in repetitive behaviours (Bond and Hebron 2019) and are supported by the detailed autobiographical accounts of autistic girls (Salter 2019).

The above discussion highlights the potential varied presentation and emphasises the importance of widening teacher knowledge to enable them to look beyond the stereotyped repetitive behaviours associated with autism. It is necessary for teachers to first appreciate the function of repetitive behaviours before then recognising behaviours which may fulfil this function for autistic girls.

Inflexibility

Inflexibility is also discussed in the context of a preference for or insistence on sameness. It is reflected in this domain through a demonstration of a strong preference for environments and behaviours that are predictable for the individual and a resistance to change (Boucher 2017). An early study into autism alluded to the ‘aggressive’ and ‘domineering’ behaviours that may be exhibited by autistic boys in demonstrating their insistence on sameness, meeting this criterion in a more obvious manner (Kopp and Gillberg 1992). It was indicated that this behavioural style was not evident in the sample of autistic girls in the study and as a result it was acknowledged that the difficulties that some girls experience in this area may go unnoticed. Differences in presentation of behaviour are also discussed in the context of the externalising behaviours of boys compared to girls’ tendencies to internalise their difficulties (Bond and Hebron 2019). These differences have resulted in teachers reporting fewer concerns about autistic girls as they do not outwardly express their difficulties (Bond and Hebron 2019). However, it is important to recognise that while an outward demonstration of this criterion may not be as explicit in girls, that is not to say that autistic girls do meet this criterion.

Insight from an autistic female author Cook O’Toole (2018) emphasised that taking a more holistic view of the criteria is required to appreciate that the outward presentation may differ between males and females. Regardless, the function of these behaviours may be the same. She reflected on the way order and facts provided security and predictability to her life, contrasting this with the unpredictable and changeable nature of social interactions. Inflexibility can influence the manner in which autistic girls play with their peers and can be evidenced through a tendency to be overly dominant or controlling, a nod towards the need to control the play (Attwood and Garnett 2019). Controlling behaviour is also referenced in relation to interactions with peers, so moving from play activities into interactions more broadly (Bond and Hebron 2019; Gould and Ashtown-Smith 2011). In addition to this, it is reported that autistic girls may tend to be perfectionistic (Lai et al. 2015). When taken together and considered outside of the stereotypical connotations associated with this criterion, these traits can demonstrate a clear preference for sameness and a resistance to change. Autistic girls who exhibit a different profile of traits in relation to these criteria may still require the support of their teachers despite the lack of an outward presentation, therefore highlighting the importance of an awareness of the internal presentation of autism.

Interests

Research has acknowledged that the manifestation of restricted interests is characteristically different in girls compared to boys (Hiller, Young, and Weber 2014). These interest differences may be more difficult for professionals to identify and have been suggested as a reason for the under-identification of autistic girls (Attwood 2007). In a controlled study conducted by Solomon et al. (2012), marginally significant differences were identified with boys presenting with more ‘obvious’ restricted interests than girls. The use of the word obvious refers to interests that were more typically associated with the ‘male as norm’ presentation of autism. This finding endorsed the findings of another review paper; however, it was acknowledged that the choices available on the parent report forms could have been a limitation of this study (Rivet and Matson 2011). A further study by Hiller, Young, and Weber (2014) revealed that the interests of girls were most commonly reported as being in the ‘seemingly random’ category. It is noteworthy to reflect on the category label of ‘seemingly random’ (examples: animals, rocks) as it appears that the interests of girls were categorised under this label as they did not reflect the other categories which could be argued as being stereotypical of male based interests. These findings correlate with other research which has revealed that the restricted interests of autistic boys are often object based and technical whereas the interests of autistic girls are often people/animal oriented (Lai et al. 2015).

The interests of autistic girls are often viewed as more ‘typical’ and are described as similar to those of their non-autistic peers, for example interests in animals, books, celebrities (Gould and Ashtown-Smith 2011; Winter-Messiers 2007). However, the interests of autistic girls differ from their peers in terms of their intensity and the time they engage with them (Attwood 2007). This is a subtle difference and could go unnoticed by teachers due to lack of awareness and knowledge of the female profile of autism. Narrative descriptions from an autistic girl (Salter 2019) provide useful lens into a more female perspective whereby music was identified as a ‘restricted interest’ with examples provided from early years to young adulthood to support this. Insights from an autistic perspective emphasises a varied presentation where the interest was perhaps less obvious than typically associated with autism. It is apparent that although qualitatively different from autistic boys, autistic girls demonstrate equally passionate interests. Alerting teachers to the subtle differences in the interests of autistic girls and autistic boys may better equip them to focus their professional observations.

Sensory differences

The final consideration in the RRBI domain is that of sensory processing. Limited research examines the sensory differences and experiences of autistic girls and women and therefore little is known in relation to whether gender differences exist in this domain. In an investigation carried out by Lai et al. (2011) of the similarities and differences between the sexes in age- and IQ-matched adults with autism, it was found that females experienced more lifetime sensory symptoms, however, it was advised that this result should be viewed as preliminary as the Autism Diagnostic Interview-Revised (ADI-R), a standardised, semi-structured interview schedule based on the DSM-IV and ICD-10 diagnostic concepts was not designed to be specifically sensitive to sensory symptoms and merely provides summary information. More recently, in a study by Duvekot et al. (2017) examining whether behavioural characteristics influenced the probability of an autism diagnosis in a sample of children aged 2.5-10 years old, findings acknowledged that sensory symptoms were positively associated with an autism diagnosis in both boys and girls. Moreover, it was suggested that the presentation of RRBIs did not largely contribute to a diagnosis in girls (Duvekot et al. 2017).

Despite the inconsistent findings and limited research in this area, it is important to appreciate that sensory differences appear to be an important feature of autism for autistic individuals and are commonly reported among autistic women (Baldwin and Costley 2016; Bargiela, Steward, and Mandy 2016; Tierney, Burns, and Kilbey 2016). Research has acknowledged that schools can be challenging sensory environments for autistic girls (Goodall and MacKenzie 2019; Milner et al. 2019; Moyse and Porter 2015), therefore necessitating teachers to work with autistic girls to support their sensory needs. In addition, further research is needed to ascertain gender differences in relation to this criterion to enhance understanding and create more inclusive sensory environments.

Future directions

This narrative review has examined literature pertinent to the sex/gender differences in autism and more specifically explored the female profile of autism. The diagnostic criteria for autism as identified in the DSM-5 (APA 2013) was used as a framework for interrogation. The findings have indicated differences in the presentation of autism between the sexes. The research points towards a more internal presentation of autism in girls with their characteristics less obvious compared to their male counterparts. The findings of this review have significant implications for policy, practice and research in terms of recognition of and support for autistic girls and developments in autism research to improve the outcomes of autistic females.

Awareness of a varied presentation of autism is imperative for accurate identification of and support for females. In light of this, the information presented in this review requires action on behalf of policy makers in terms of the diagnostic process and post-diagnostic supports. It is accepted that the diagnostic criteria outlined in DSM-5 (APA 2013) are broad, however, it could be argued that their application to the varied and perhaps more nuanced female presentation could be restricted by the knowledge of professionals involved in the diagnostic process. An expert knowledge base in the presentation of autism in females along with the ability to see beyond the androcentrism of the diagnostic process and accompanying screening tools is necessary in the current context to identify of autistic girls. As a result, there is an inherent need to provide clarity with regards to how the characteristics of autistic girls meet the diagnostic criteria. While there is an emerging body of research on the female profile of autism being published, it is important this material is being disseminated and used to advance the diagnostic process. It is recommended that this research should be used to inform revisions on the current diagnostic criteria and screening tools used in the identification of autism. In addition, it is also necessary to consider the important role of the education context in identification and support of autistic girls. There is a need for guidance documents to be issued to educational contexts to ensure the needs of autistic girls are being attended to. While 2022 saw the publication of the Autism Good Practice Guideline in the Republic of Ireland, minimal information was presented in relation to the varied presentation of autism in girls. It is important that guidance appreciates the diversity of autism and its presentation in order to challenge male-biased perceptions of autism.

Teachers and school staff are critical stakeholders in terms of the identification of autism and will often be involved in the referral process. Moreover, they play a significant role in supporting the individual post-diagnosis. This review highlighted the variable knowledge of professionals in relation to autism in girls. It is apparent that there is a lack of knowledge on the presentation of autism in females among professionals with a gendered-knowledge base evident. This review revealed a varied autism profile and highlighted the subtleties of the internal presentation in girls. As a result, it is of critical importance that professionals are aware of the sex differences in autism to improve the recognition of and support for autistic girls. In summary, the findings of this review emphasise the need for training for teachers and other school staff. It is imperative that providers of Initial Teacher Education (ITE) review course content to ensure input is provided in relation to the female profile of autism. Considering the significance of this issue, it is necessary to go beyond ITE and ensure all in-service teachers are provided with the necessary knowledge to competently support autistic girls. It is important that individuals and training bodies facilitating professional development for teachers on the topic of autism update their training to ensure it is inclusive of the female presentation of autism. Furthermore, it is believed that there is a need for all school contexts to receive some input which specifically addresses the female autism profile. It is hoped that the findings presented in this paper will raise awareness of how autism may present for girls so that practitioners can be supported to identify individuals who experience a more ‘internal’ presentation of autism, facilitate appropriate support and improve outcomes for autistic individuals.

Finally, this review supports the need for further research. While there exists a growing body of research that examines sex differences in autism, there is need for ongoing research exploring how autism presents and is experienced differently in females. Furthermore, there is potential for further empirical research to examine the current diagnostic criteria and process through the lens of gender. It could be argued that much of the current research is limited due to its reliance on diagnosed samples, who, by definition meet the current criteria for autism. However, as discussed, the androcentrism of the current diagnostic criteria are perceived to be a barrier for autistic females in obtaining a diagnosis (Duvekot et al. 2017). Considering this, the use of population-based samples in research could prove powerful in furthering knowledge of the female profile of autism. Lastly, valuable learning could be garnered from research with female-specific samples and from larger research studies with balanced sex samples.

Conclusion

While there is a growing body of research in the area of sex/gender differences in autism, there continues to be a lack of awareness of the female presentation of autism. The lack of clarity in relation to how autistic girls may meet the current diagnostic criteria has resulted in persistent under-identification of autistic girls. It is important to acknowledge that recognition precedes diagnosis and teachers are commonly one of the first professionals involved in this process. Despite this, research suggests that teachers are not equipped with the knowledge base to recognise autism in girls, resulting in autistic girls going unrecognised and unsupported. The author concludes that it is imperative that the internal presentation of autism, commonly associated with autistic girls, is acknowledged as an alternative profile to the traditional profile of autism which is grounded in a male-dominated research base.

This paper endeavoured to bring together insights garnered from both a quantitative and qualitative research perspective in order to provide a comprehensive understanding of the presentation of autism in girls and detail how the profile of characteristics can be viewed with reference to the diagnostic criteria. It is accepted that the narrative review methodology adopted for this paper has its limitations. Broad details are provided in relation to the search strategy, however, this paper did not follow an official method for searching and retrieving the literature, a common criticism of the narrative review approach (Coughlan and Cronin 2021; Paré et al. 2015). Moreover, as with any narrative review, the sources used were selected based on perceived relevance and a systematic and comprehensive search of all literature was not carried out nor were sources given equal weight (Paré et al. 2015; West and Martin 2023). Nonetheless, it is hoped that this review will provide education practitioners and allied professionals, as well as researchers with a foundational understanding of the presentation of autism in girls.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This work was supported by Irish Research Council [grant number GOIPG/2022/1084].

Notes on contributors

Aoife Munroe

Aoife Munroe is a PhD Candidate in the Department of Educational Psychology, Inclusion and Special Education in Mary Immaculate College (MIC), Limerick, Ireland. Aoife is an Irish Research Council Government of Ireland Scholar, and her PhD research is focused on autistic girls’ experiences of mainstream education provision at post-primary level in the Republic of Ireland. Aoife is a qualified primary school teacher and has worked as a Lecturer in Special and Inclusive Education at St. Angela’s College, Sligo and Mary Immaculate College, Limerick.

Michele Dunleavy

Michele Dunleavy is an Assistant Professor in the Department of Educational Psychology, Inclusive and Special Education, Mary Immaculate College (MIC), Limerick, Ireland. Before joining the staff in MIC Michele worked for many years as a teacher in mainstream primary, Special Education Teacher, special schools, and special classes. Additionally, Michele was seconded, as an Advisor to the Special Education Support Service where she worked with teachers from the primary and post-primary sector. Her main research area is Autism and in particular how schools support the access of autistic students to relationship and sexuality education (RSE).

Notes

1 The adoption of neurodiversity affirmative language is preferred by the author and used where possible, however, due to the nature of the review topic, less preferred terminology will be presented, particularly in relation to the diagnostic criteria, for the purpose of clarity in the discussion between how the characteristics experienced by autistic girls may meet the criteria as identified in DSM-5 [1].

2 Sex/gender is used interchangeably to refer to the inter-related concepts of sex (biological) and gender (sociocultural).

3 SENCo stands for Special Educational Needs Coordinator. SENCo's are school staff members with specific responsibilities in special education provision.

4 For the purpose of this paper, diagnostic criteria as outlined in DSM-5 [1] will be adopted, however, The International Classification of Diseases (ICD) criteria are comparable to those of DSM.

5 In order to obtain a diagnosis, it is necessary that differences exist in all three areas of the social communication and interaction domain and in at least two out of four criteria outlined in the RRBI domain (See Tables 1 & 2), have a negative impact of daily life and must be present in early childhood (although they may not become fully evident until the social demands exceed the individual's capacity).

6 Repetitive behaviours refer to any repetitive or unusual speech, movements or use of objects. Typical examples include motor stereotypies (rhythmic, fixed movements such as hand flapping), lining up of toys, organisation of objects in a fixed manner or according to pattern, use of repetitive phrases, echolalia, scripted language, noise making, repeated engagement in an activity, etc.