ArticlePDF Available

Abstract and Figures

Autobiographical descriptions and clinician observations suggest that some individuals with autism, particularly females, ‘camouflage’ their social communication difficulties, which may require considerable cognitive effort and lead to increased stress, anxiety and depression. Using data from 60 age- and IQ-matched men and women with autism (without intellectual disability), we operationalized camouflaging in adults with autism for the first time as the quantitative discrepancy between the person’s ‘external’ behavioural presentation in social–interpersonal contexts (measured by the Autism Diagnostic Observation Schedule) and the person’s ‘internal’ status (dispositional traits measured by the Autism Spectrum Quotient and social cognitive capability measured by the ‘Reading the Mind in the Eyes’ Test). We found that the operationalized camouflaging measure was not significantly correlated with age or IQ. On average, women with autism had higher camouflaging scores than men with autism (Cohen’s d = 0.98), with substantial variability in both groups. Greater camouflaging was associated with more depressive symptoms in men and better signal-detection sensitivity in women with autism. The neuroanatomical association with camouflaging score was largely sex/gender-dependent and significant only in women: from reverse inference, the most correlated cognitive terms were about emotion and memory. The underlying constructs, measurement, mechanisms, consequences and heterogeneity of camouflaging in autism warrant further investigation.
Content may be subject to copyright.
1 –13
© The Author(s) 2016
Reprints and permissions:
DOI: 10.1177/1362361316671012
Autism spectrum condition/disorder (henceforth ‘autism’)
has a life-long impact on individual development. Adult
presentation and outcome vary substantially (Howlin and
Moss, 2012). Those who are diagnosed in childhood tend
to show reduced autistic symptoms over time, but only
a minority show satisfactory social functioning (Howlin
et al., 2013). In contrast to individuals who receive diagno-
ses in childhood, some individuals are only identified later
in life and may ‘fly under the radar’ for many years partly
because of learnt strategies to conceal social difficulties.
These late-diagnosed individuals tend to suffer from con-
current mental health challenges potentially related to long-
term stress in adaptation to daily life in the society (Lai and
Baron-Cohen, 2015). Given long-standing environmental
support but also pressure to ‘fit in’ with neurotypical social
communication, individuals with autism (irrespective of
timing of diagnosis) may develop coping strategies over
development. One such coping strategy is that they may
‘camouflage’ difficulties during social situations (Attwood,
2007) by either hiding behaviour that might be viewed as
socially unacceptable or artificially ‘performing’ social
behaviour deemed to be more neurotypical – they Pretend
to be Normal (Willey, 1999).
Quantifying and exploring camouflaging
in men and women with autism
Meng-Chuan Lai1,2,3,4,5, Michael V Lombardo4,6,
Amber NV Ruigrok4, Bhismadev Chakrabarti4,7,
Bonnie Auyeung4,8, Peter Szatmari1,2,3, Francesca Happé9
and Simon Baron-Cohen4,10; MRC AIMS Consortium
Autobiographical descriptions and clinician observations suggest that some individuals with autism, particularly females,
‘camouflage’ their social communication difficulties, which may require considerable cognitive effort and lead to increased
stress, anxiety and depression. Using data from 60 age- and IQ-matched men and women with autism (without intellectual
disability), we operationalized camouflaging in adults with autism for the first time as the quantitative discrepancy between
the person’s ‘external’ behavioural presentation in social–interpersonal contexts (measured by the Autism Diagnostic
Observation Schedule) and the person’s ‘internal’ status (dispositional traits measured by the Autism Spectrum Quotient
and social cognitive capability measured by the ‘Reading the Mind in the Eyes’ Test). We found that the operationalized
camouflaging measure was not significantly correlated with age or IQ. On average, women with autism had higher camouflaging
scores than men with autism (Cohen’s d = 0.98), with substantial variability in both groups. Greater camouflaging was
associated with more depressive symptoms in men and better signal-detection sensitivity in women with autism. The
neuroanatomical association with camouflaging score was largely sex/gender-dependent and significant only in women:
from reverse inference, the most correlated cognitive terms were about emotion and memory. The underlying constructs,
measurement, mechanisms, consequences and heterogeneity of camouflaging in autism warrant further investigation.
adults, autism, brain structure, camouflage, camouflaging, cognition, coping, gender, sex, sex differences
1Centre for Addiction and Mental Health, Canada
2The Hospital for Sick Children, Canada
3University of Toronto, Canada
4University of Cambridge, UK
5National Taiwan University Hospital, Taiwan
6University of Cyprus, Cyprus
7University of Reading, UK
8The University of Edinburgh, UK
9King’s College London, UK
10Cambridgeshire and Peterborough NHS Foundation Trust, UK
Corresponding author:
Meng-Chuan Lai, Child, Youth and Emerging Adult Program, Centre for
Addiction and Mental Health, 80 Workman Way, Toronto, ON M6J
1H4, Canada.
671012AUT0010.1177/1362361316671012AutismLai et al.
Special Issue Article
at University of Cyprus on December 1, 2016aut.sagepub.comDownloaded from
2 Autism
Examples of camouflaging include making eye contact
during conversation, using learned phrases or pre-prepared
jokes in conversation, mimicking other’s social behaviour,
imitating facial expressions or gestures, and learning and
following social scripts (Lai and Baron-Cohen, 2015). One
may also learn to consciously speak more quietly or not
to stand too close to another person or not to make per-
sonal remarks, perhaps following feedback that these
may be hurtful or uncomfortable for others or perhaps as
a conscious goal to model their behaviour on a neuro-
typical peer in order to gain greater social acceptance.
Autobiographical descriptions and clinician observations
often suggest that camouflaging unfortunately comes at a
cost: it often requires substantial cognitive effort, can be
exhausting and may lead to increased stress responses,
meltdown due to social overload, anxiety and depression,
and even a negative impact on the development of one’s
identity (Attwood, 2007; Boyd et al., 2011; Lai et al.,
2011; Simone, 2010; Willey, 1999; Williams, 1992).
Camouflaging may also play a role in the observed
male-preponderance in autism prevalence, if it is the
case that females are more likely or more motivated to
camouflage, and thereby go undetected and undiagnosed
for longer. Thus, the male-preponderance may reflect
aetiological sex/gender differences, but may also be a
product of under- or misrecognition of autism in females,
potentially associated with gender stereotypes and the
historically male-based behavioural characterization of
autism, with insufficient acknowledgement of how
females may present some behaviours characteristic of
autism in a qualitatively or quantitatively different way
from their male counterparts – camouflaging being one
example (Lai et al., 2015). Population-based data show
that females are often diagnosed at later ages (Begeer
et al., 2013; Giarelli et al., 2010; Rutherford et al., 2016;
Shattuck et al., 2009) and less easily than males with
autism (Russell et al., 2011; Wilson et al., 2016), unless
there are concurrent behavioural or cognitive challenges
(Dworzynski et al., 2012). One of the potential reasons for
this may be the heightened tendency to camouflage diffi-
culties in many females on the spectrum: when difficulties
in social interaction and communication are masked, their
signs of autism are less likely to be picked up by families,
teachers or primary care providers in order to trigger an
assessment. If the diagnostician further misses signs of
camouflaging, superficially ‘typical’ non-verbal skills and
social manner may be wrongly taken as evidence to rule
out the presence of autism (Lai and Baron-Cohen, 2015).
In the 1980s, investigating sex/gender ratio in the
autism spectrum, Wing (1981) wrote that ‘The possibility
that girls with the triad of impairments who had higher
levels of intelligence were missed in the search for cases
has to be considered’ (p. 134). Ten years later, Gillberg
(1991) noted that ‘Asperger syndrome can occur in girls
[…] on the surface, symptoms of impairment of social
interaction might be less conspicuous than corresponding
symptoms in boys’ (p. 129). He suggested that girls might
have more advanced social skills to conceal their autistic
characteristics. Attwood (2007) also pointed out in his
highly influential book The Complete Guide to Asperger’s
Syndrome that ‘Some girls and women with Asperger’s
syndrome, and adults of considerable intellectual ability,
can be more difficult to diagnose due to an ability to cam-
ouflage their difficulties’ (p. 40). Women with autism and
their parents regularly echo this observation and consider
camouflaging as one of the major reasons females on
the autism spectrum often go under-recognized until they
can no longer compensate (Ernsperger and Wendel, 2007;
Hendrickx, 2015; Lawson, 1998; Miller, 2003; Simone,
2010; Willey, 1999; Zaks, 2006).
Clinicians and researchers have also increasingly
described camouflaging in females, in contrast to males,
with autism (Attwood, 2006; Gould and Ashton-Smith,
2011; Kopp and Gillberg, 1992; Lai et al., 2015; Marshall,
2015). Recent large-scale, population-based epidemio-
logical studies (many of them included active case ascer-
tainment) show a 2–3:1 male-preponderance compared to
the widely cited 4–5:1 ratio from earlier studies (Baxter
et al., 2015; Idring et al., 2012; Kim et al., 2011; Mattila
et al., 2011; Zablotsky et al., 2015), possibly suggesting
better recognition of females in recent years owing to
increased clinical awareness or more sensitive measure-
ment. Longitudinal population-based studies in the Nordic
regions particularly confirm this trend (Jensen et al., 2014;
Kocovska et al., 2012). Improving our understanding
of camouflaging, along with other possible ‘female-
phenotypes of autism’, may further facilitate the identifi-
cation of masked symptoms and difficulties and enhance
timely diagnosis and support.
Although camouflaging has been frequently described
as a major characteristic of women with autism (e.g. by the
findings from the European Union (EU)-funded Autism in
Pink project,, it has received
surprisingly limited systematic scientific investigation. In
a recent qualitative study, Tierney et al. (2016) interviewed
10 teenage girls with autism on the social challenges
associated with adolescence and analysed the data using
Interpretative Phenomenological Analysis. These teenage
girls reported that they developed explicit strategies to
manage social relationships, in particular imitation and
masking. Hiller et al. (2014) compared school-age boys
and girls who were clinically diagnosed with autism and
discovered that they met the clinical criteria in somewhat
different ways. Notably, some differences may underpin or
reflect higher camouflaging in females. For example, girls
were more likely to be able to integrate non-verbal and
verbal behaviours, to have better imagination (at least at
face value), to maintain a reciprocal conversation and
to initiate (but not maintain) friendships. These character-
istics seem to have ecological impacts, as school teachers
at University of Cyprus on December 1, 2016aut.sagepub.comDownloaded from
Lai et al. 3
reported far fewer concerns about girls than boys with
autism regarding their social skills, friendship and exter-
nalizing behavioural problems (Hiller et al., 2014).
Head and colleagues found that teenage girls with
autism scored higher on the Friendship Questionnaire
(indicative of better/more friendship) than did teenage
boys with autism and at a level comparable to that of
typically developing teenage boys. One interpretation is
that females with autism may ‘develop a capacity to
camouflage or hide their social insecurities in order to
fit in’ (Head et al., 2014: 6). Lai et al. (2011) alluded
to higher camouflaging in women than men with autism
based on the observation that, given similar levels of
childhood autistic symptoms measured by the Autism
Diagnostic Interview–Revised (ADI-R), women with
autism tended to show less pronounced autistic features in
interpersonal–social contexts as measured by the Autism
Diagnostic Observation Schedule (ADOS).
By analysing behaviour from the demonstration
activities in the ADOS-2, Rynkiewicz et al. (2016) found
girls with autism used gestures more vividly than boys
with autism and suspected this to be one component of
enhanced camouflaging in females which ‘may pose risk
of underdiagnosis or not receiving the appropriate diagno-
sis for this population’ (p. 6). Finally, Lehnhardt and col-
leagues studied cognitive profiles in late-diagnosed men
and women with autism, both presumably missing early
diagnosis partly due to camouflaging. They found that
men with autism had higher verbal abilities than women
with autism, whereas women with autism had higher pro-
cessing speed and better executive function than men with
autism. They proposed that this sex/gender-differential
finding might indicate ‘different strategies being applied
to camouflage the autistic background during childhood
and adolescence’ (Lehnhardt et al., 2016: 150).
These pioneering studies indicate that camouflaging can
be conceptualized as using learned social communicative
behaviours (e.g. imitation, gestures and conversation skills)
to mask underlying difficulties related to autism. Females
with autism tend to employ more camouflaging than males
with autism. This skill is probably supported by verbal
ability and executive function. Since camouflaging involves
real-time monitoring of the environment to make careful
and appropriate responses, we hypothesize that in order to
successfully camouflage, at the cognitive level one requires
sensitive monitoring of the environment (i.e. being highly
vigilant) and/or a more conservative response strategy (i.e.
being highly cautious). Nevertheless, camouflaging may
also be supported by other skills deemed to be relevant by
its definition, such as social imitation ability, and/or other
as yet unrecognized cognitive or behavioural abilities.
None of the above studies have operationalized and
quantified camouflaging in autism or clarified its neuro-
cognitive correlates. We consider there are at least
two complementary approaches needed to advance our
understanding. The first is to take a grounded theory
approach (Glaser and Strauss, 2009), starting from
research questions such as ‘what is camouflaging in
autism?’ and ‘what are the required abilities and skills for
camouflaging?’ and to collect qualitative data to inform
concept formation. The second is to take a positivist
approach and operationalize camouflaging using exist-
ing, standardized measures and to test for relevant
hypotheses derived from the observations and findings
summarized above. Here we take the latter approach,
using existing standardized and validated measures to
(1) derive an estimate of camouflaging in adults with
autism, (2) compare camouflaging between males and
females, (3) test whether more camouflaging is associ-
ated with more severe anxiety and depression and (4) test
whether more camouflaging is associated with better
verbal ability, better signal detection from background
events and more conservative responses. In addition, in a
hypothesis-free manner, we explore the neuroanatomical
correlates of individual differences in camouflaging and
then use the ‘big data’ from the neuroscience literature to
draw a ‘reverse inference’ to identify other potentially
associated cognitive correlates of camouflaging.
Participants comprised 30 adult females and 30 adult males
with autism (none with intellectual disability) matched for
age (18–49 years), verbal IQ (VIQ), performance IQ (PIQ)
and full-scale IQ (FIQ). All participants had a formal clini-
cal diagnosis of International Classification of Diseases,
Tenth Edition (ICD-10; World Health Organization (WHO),
1992) childhood autism or Asperger’s syndrome and/or
Diagnostic and Statistical Manual of Mental Disorders
(4th ed., text rev.; DSM-IV-TR; American Psychiatric
Association (APA), 2000) autistic disorder or Asperger’s
disorder, as assessed by a psychiatrist or clinical psycholo-
gist in the National Health Service, United Kingdom.
Additionally, all but two participants reached the diagnostic
algorithm cut-offs on the ADI-R (Lord et al., 1994). The
two exceptions were female participants where ADI-R was
unavailable due to childhood caregivers being unable to be
interviewed. One of these individuals scored above the cut-
off for ‘autism spectrum’ on the ADOS (Lord et al., 2000)
and the other was positive for a diagnosis on the Adult
Asperger Assessment (AAA) which incorporates caregiver
reports of childhood behaviours and developmental his-
tory (Baron-Cohen et al., 2005). Scoring one point below in
only one of the three core symptom domains of ADI-R was
permitted, to allow for possible underestimation of early
developmentally atypical behaviours in the recall of car-
egivers whose children are now adults. ADOS module 4
was performed, but the score was not used as an inclusion
at University of Cyprus on December 1, 2016aut.sagepub.comDownloaded from
4 Autism
Behavioural, cognitive and neuroanatomical characteri-
zations of this cohort have been reported previously (Ecker
et al., 2012, 2013; Lai et al., 2011, 2012, 2013; Wilson
et al., 2014), along with detailed project and recruitment
information. The sample included in this study comprises
the autism groups reported in a previous neuroimaging
study (Lai et al., 2013).
Behavioural and cognitive measures
All participants were assessed using the Wechsler
Abbreviated Scale of Intelligence (Wechsler, 1999) for
estimation of VIQ, PIQ and FIQ. ADI-R was conducted to
assess childhood autism characteristics (Lord et al., 1994).
Module 4 of the ADOS (Lord et al., 2000) was used
to quantify current, adult (‘external’) behavioural charac-
teristics related to autism. The ADOS is a standardized
activity- and interview-based semi-structured assessment
for current behavioural presentation in a quasi-natural,
interpersonal context. Behaviours of the participant were
coded immediately after the assessment session into
31 items, of which 16 were entered into the ‘diagnostic
algorithm’. The diagnostic algorithm score quantifies the
adult’s cardinal social interactive and communicative
behaviours associated with autism.
Among a battery of self-report questionnaires obtained
in this project (Lai et al., 2011), the Autism Spectrum
Quotient (AQ) (Baron-Cohen et al., 2001b) was selected
to measure participants’ self-reflection (‘internal’ percep-
tion) of their personal characteristics related to autism.
The AQ contains 50 items measuring autistic-like traits in
terms of social skills, attention switching, attention to
detail, communication and imagination. Among a battery
of cognitive tests in this project, the ‘Reading the Mind in
the Eyes’ Test (RMET) (Baron-Cohen et al., 2001a, 2015)
was selected to measure participants’ actual (‘internal’)
capability in advanced mentalizing and complex emotion
recognition. The 36-item RMET requires participants to
infer mental status solely from photos of a person’s eyes
and immediate surrounding areas of the face.
Based on our hypotheses, to test for the clinical and
cognitive correlates of camouflaging, we selected the
21-item Beck Anxiety Inventory (BAI; Beck et al., 1988)
to measure anxiety symptoms and the 21-item Beck
Depression Inventory (BDI; Beck et al., 1961) to measure
symptoms of depression. VIQ was selected to indicate the
verbal ability of an individual. From our available cogni-
tive measures, performance on executive function tasks
most closely reflects our construct of interest (i.e. monitor-
ing the environment and patterns of behavioural response).
We therefore selected an online version of the Go/No-Go
task and derived performance measures using the signal-
detection theory (SDT) framework (Green and Swets,
1966), namely, sensitivity (d′ = ZHitZFA, where ZHit is the
corresponding Z value in the normal distribution for the
probability of Hit, that is, signal present and the response
is ‘present’, and ZFA is the same for False Alarm, that
is, signal absent but the response is ‘present’) and crite-
rion (response bias) (C = −0.5 × (ZHit + ZFA)). Sensitivity d′
indicates the participant’s ability to discriminate signal
from background noise, and criterion C quantifies how
liberal (i.e. C < 0) or conservative (i.e. C > 0) the response
strategy (bias) is. Additional details of the implementation
of cognitive tasks were reported earlier (Lai et al., 2012).
Operationalizing camouflaging using
standardized measures
As camouflaging could be defined as (consciously or
unconsciously) compensating for and/or masking difficul-
ties in social and interpersonal situations, we operational-
ized camouflaging as the discrepancy between the person’s
‘external’ behavioural presentation in social–interpersonal
contexts and the person’s ‘internal’ status (i.e. dispositional
traits and/or social cognitive capability).
We used the ADOS diagnostic algorithm score as
reflecting external presentation and both the AQ score
and RMET correct score as reflecting internal status
(self-rated dispositional traits and performance-based
socio-cognitive capability, respectively). We used the
ADOS diagnostic algorithm score to characterize one’s
behavioural presentation because it is the only psycho-
metrically tested, reliable measure of social communica-
tion behaviours in interpersonal contexts for individuals
with autism. Although we have previously questioned the
validity of using ADOS module 4 diagnostic algorithm
cut-off for making diagnostic judgement for autism in
adults without intellectual disability, particularly in females,
we do not question the validity of ADOS in quantitatively
measuring and describing cardinal social communication
behaviours (Lai et al., 2011). The internal (i.e. latent) status
of autism ideally is captured by a wide array of measures
on the cognitive and psychological characteristics of the
individual. Relying on one particular measure risks biases
resulting from the measure’s inherent limitation (e.g. a
self-report measure is dependent on one’s perception of
their own behavioural/cognitive styles). We were confined
by available data but were able to capture two key aspects,
namely, one’s perception of their personal characteristics
associated with autism and one’s cognitive performance
on a mentalizing and emotion recognition task; both are
integral parts in the assessment and understanding of indi-
vidual characteristics of autistic people.
The three scores were first standardized (termed
as SADOS, SAQ and SRMET) by mean-centring (to the
whole autism sample in this study, N = 60) and scaling
(i.e. divided by the maximum possible score of each) to
generate uniformly scaled measures that can be arithmeti-
cally manipulated (i.e. added to or subtracted from each
other); the uniformly scaled measures were derived in this
at University of Cyprus on December 1, 2016aut.sagepub.comDownloaded from
Lai et al. 5
way rather than using z-scores because z-standardization
was problematic for the ADOS score, as (1) there was no
available autistic population mean and standard deviation,
and (2) sample mean and standard deviation were not
valid substitutes here as the distribution was skewed.
A first measure of camouflaging was quantified as the
difference between self-rated autistic-like traits and
external behaviours (CF1 = SAQSADOS), and a second
measure between mentalizing ability and external behav-
iours (CF2 = −SRMETSADOS); higher scores on CF1 and
CF2 indicate more camouflaging. Finally, using principal
component analysis, the first principal component score
of CF1 and CF2 was taken as a single, parsimonious
measure (the ‘camouflaging score’, CF) that incorporates
information from all relevant measures for further
Neuroimaging measures
Participants were scanned using a contemporary 3 T
MRI scanner (GE Medical Systems HDx) fitted with an
8-channel receive-only RT head-coil using Driven
Equilibrium Single Pulse Observation of T1 (DESPOT1)
(Deoni et al., 2008). Simulated T1-weighted inversion
recovery (IR) images were created via ImageJ. Pre-
processing was conducted using the SPM12 software
(Wellcome Trust Centre for Neuroimaging): tissue seg-
mentation was done by Segment (previously New
Segment), and the segmented grey matter (GM) images of
the 60 participants were non-linearly normalized (with
modulation) to the standard Montreal Neurological
Institute (MNI) space using DARTEL (Ashburner, 2007)
and smoothed with a 4-mm full-width at half maximum
Gaussian kernel. Additional information about image
acquisition has been reported previously (Lai et al., 2013).
Statistical analysis
Statistical distribution of CF was examined for skewness
and kurtosis. Independent-samples t-tests were used
for comparisons between men and women with autism.
Pearson’s correlation was used to examine the correlation
patterns between CF and clinical/cognitive measures. Sex/
gender difference in correlation patterns was tested by
constructing a multiple regression model with clinical/
cognitive measures as the dependent variable and CF, sex/
gender and CF-by-sex/gender interaction as predictors,
and then examining the significance of β on the interaction
term: a significant interaction suggests that the correlation
between CF and the dependent variable is dependent on
sex/gender. Critical level for statistical significance was
set at α = 0.05. These analyses were implemented using the
IBM SPSS Statistics version 22.
Hypothesis-free exploration of the neuroanatomical
correlates of CF was performed using mass-univariate
tests with SPM12. Prior to statistical modelling, each
modulated GM map was rescaled by individual total GM
volume (i.e. voxel value divided by individual total GM
volume) to derive relative regional GM volume estimates.
First, for all participants, we fit a parsimonious general
linear model (GLM) at each voxel, with sex/gender and
CF as fixed factors, along with their interaction term, and
age as a nuisance covariate. When significant CF-by-sex/
gender interaction was found, we subsequently performed
sex/gender-stratified whole-GM mass-univariate tests
again to identify CF–GM volume association separately
for the male and female groups; here, CF was the fixed
factor and age was a nuisance covariate. All whole-GM
voxel-level tests were restricted to voxels with a partial
volume estimate >0.25. Multiple comparison correction
was performed at the cluster level by controlling topologi-
cal false discovery rate (FDR) calculated under Gaussian
Random Field Theory (Chumbley and Friston, 2009),
using a cluster-forming voxel-level height threshold of
p < 0.01 for each contrast and a spatial extent threshold
(corrected for non-stationarity (Hayasaka et al., 2004))
that ensures a cluster-wise FDR at q < 0.05.
Statistical characteristics and sex/gender
differences in the CF measure
Sample characteristics are given in Table 1. In the present
sample, CF did not significantly deviate from the normal
distribution in either males with autism (skewness
z-score = −1.351, kurtosis z-score = −0.363, Shapiro–Wilk
test p = 0.288) or females with autism (skewness z-score
= −1.710, kurtosis z-score = 1.013, Shapiro–Wilk test p =
0.213). Across the whole sample, CF was not significantly
correlated with age (Pearson’s r = 0.188, p = 0.151), VIQ
(r = 0.180, p = 0.168), PIQ (r = 0.053, p = 0.685), or FIQ
(r = 0.136, p = 0.301); this was also the case when data
were split into males only (age r = 0.303, p = 0.103; VIQ
r = 0.095, p = 0.617; PIQ r = 0.099, p = 0.604; FIQ r = 0.120,
p = 0.527) or females only (age r = 0.053, p = 0.783; VIQ
r = 0.277, p = 0.138; PIQ r = 0.116, p = 0.540; FIQ r = 0.215,
p = 0.253). Women with autism on average scored signifi-
cantly higher on CF than men with autism (p < 0.001),
with an effect size showing almost 1 standard deviation of
difference (Cohen’s d = 0.98). However substantial varia-
bility on this measure was present in both males and
females (see Figure 1(a)).
Testing correlations between CF and
Across the whole sample, CF was positively correlated
with the total score on the BDI (r = 0.301, p = 0.019) but
not the BAI (r = 0.202, p = 0.121). When sex/gender-
stratified, the CF-BDI correlation was significant in males
(r = 0.533, p = 0.002) but not females (r = 0.030, p = 0.876).
at University of Cyprus on December 1, 2016aut.sagepub.comDownloaded from
6 Autism
Table 1. Characteristics of the sample: men and women with autism.
Mean (SD) [range]aMen (M) (n = 30) Women (F) (n = 30) StatisticsbEffect size (d)
Age (years) 27.2 (7.3) 27.8 (7.6) ns (p = 0.761) 0.08
Verbal IQ 114.3 (12.9) 115.8 (13.1) ns (p = 0.656) 0.12
Performance IQ 113.3 (15.0) 110.4 (16.7) ns (p = 0.492) 0.18
Full-scale IQ 115.4 (14.1) 114.9 (13.8) ns (p = 0.904) 0.04
Reciprocal social 18.0 (5.1) [10–27] 16.4 (4.3) [11–26] ns (p = 0.215) 0.34
Communication 15.3 (3.5) [8–22] 13.1 (3.9) [8–22] M > F (p = 0.029) 0.59
RRSB 5.6 (2.5) [2–10] 4.3 (1.7) [2–8] M > F (p = 0.023) 0.60
Social communication 8.5 (5.0) [1–17] 4.3 (3.6) [0–13] M > F (p < 0.001) 1.04
RRSB 1.0 (1.0) [0–4] 0.1 (0.3) [0–1] M > F (p < 0.001) 1.25
Autism Spectrum Quotient 32.7 (7.3) 37.5 (6.7) F > M (p = 0.010) 0.69
RMET correct score 22.8 (5.8) 23.4 (6.2) ns (p = 0.700) 0.10
CF −0.168 (0.388) 0.168 (0.294) F > M (p < 0.001) 0.98
Go/No-Go taskd
Sensitivity, d3.60 (0.67) 3.71 (1.16) ns (p = 0.660) 0.12
Response bias, C−0.027 (0.198) 0.029 (0.232) ns (p = 0.327) 0.26
Beck Depression Inventory 14.5 (10.3) 14.6 (9.0) ns (p = 0.958) 0.01
Beck Anxiety Inventory 14.1 (9.9) 15.5 (10.1) ns (p = 0.580) 0.14
SD: standard deviation; ns: non-significant (p > 0.05, two-tailed, not corrected for multiple comparisons); ADI-R: Autism Diagnostic Interview–Revised;
RRSB: repetitive, restrictive and stereotyped behaviour; ADOS: Autism Diagnostic Observation Schedule.
aFor ADI-R and ADOS scores.
bIndependent-samples t-tests, except Mann–Whitney tests for ADOS scores (distribution significantly deviant from normal).
cn = 30 for men, n = 28 for women (n = 2 data missing due to childhood caregiver unavailability).
dn = 29 for men (n = 1 data missing due to technical failure), n = 30 for women.
Figure 1. Sex/gender differences in camouflaging and its association with depressive symptoms and signal-detection sensitivity: (a)
a dot and box-and-whisker plot showing the distribution of camouflaging (quantified by the measure CF) in men and women with
autism; (b) CF-BDI score correlations stratified by sex/gender; (c) CF-sensitivity correlations stratified by sex/gender.
at University of Cyprus on December 1, 2016aut.sagepub.comDownloaded from
Lai et al. 7
Male–female difference in correlation patterns approached
significance (p = 0.068), testing the significance of CF-
by-sex/gender interaction in a multiple regression model
(see Figure 1(b)). Partialling out the effect of either PIQ or
FIQ did not change any of these correlation patterns and
Testing correlations between CF and VIQ, signal
detection and response bias
As reported above, across the whole sample, CF was not
significantly correlated with VIQ (r = 0.180, p = 0.168);
this was also true when examining only males (r = 0.095,
p = 0.617) or only females (r = 0.277, p = 0.138). The Go/
No-Go data for one male were missing due to technical
failure, so data from only 29 men were included in the
analysis. Across the whole sample, CF was positively
correlated with Go/No-Go sensitivity d′ (r = 0.311,
p = 0.017) but not response bias C (r = 0.108, p = 0.416).
When sex/gender-stratified, the CF-d′ correlation was
significant in females (r = 0.432, p = 0.017) but not males
(r = 0.233, p = 0.223). Male–female difference in correla-
tion patterns approached significance (p = 0.072) (see
Figure 1(c)). Partialling out the effect of either PIQ or
VIQ did not change any of these correlation patterns and
Exploring the neuroanatomical correlates of
CF and using reverse inference to identify
associated cognitive terms
The hypothesis-free, whole-GM voxel-based GLM identi-
fied no regions showing significant main effects of CF, but
there were two significant clusters showing significant
CF-by-sex/gender interaction, indicating sex/gender-
dependent correlation patterns between CF and regional
GM volume, at left medial temporal lobe (cluster size
ke = 4248, cluster-level FDR-corrected q < 0.001, peak-
coordinate MNI (−26, −11, −28), Z = 4.04) and cerebellum
(ke = 3638, q < 0.001, peak-coordinate MNI (20, −70,
−14), Z = 3.96), where increased CF was associated with
decreased volume in females yet to a significantly dif-
ferent extent in males. To further identify anatomical cor-
relates specific for females and males, we ran whole-GM
voxel-based GLM again but separately by sex/gender.
In females, there were significant negative correlations
between CF and GM volume at bilateral cerebellum,
occipital and medial temporal structures which substan-
tially overlapped with the above-mentioned regions show-
ing CF-by-sex/gender interaction (right-lateralized cluster
ke = 3605, q = 0.001, peak-coordinate MNI (15, −65, −8),
Z = 4.04; left-lateralized cluster ke = 1922, q = 0.027,
peak-coordinate MNI (−15, −53, −25), Z = 3.81); no
regions showed significant positive correlation with CF
(see Figure 2). In males, there were no regions positively
or negatively correlated with CF that survived multiple
comparison corrections.
In order to identify the terms in the scientific literature
that are mostly associated with the identified voxels show-
ing significant association with camouflaging in women
with autism (i.e. making reasonable ‘reverse inference’
(Poldrack, 2006)), we submitted the statistical map to
the Neurosynth Image Decoder (
decode/; Gorgolewski et al., 2016) and visualized the top
60 terms showing highest correlation (r = 0.07–0.17) by
a word-cloud (produced using R and the ‘wordcloud’
library; see Figure 3(a)). This qualitatively shows terms
of anatomical regions (e.g. cerebellum, medial temporal
lobe, para/hippocampus, amygdala) and terms about emo-
tion and memory – see Figure 3(b) for the top 30 terms
(r = 0.07–0.10) after filtering out anatomical terms.
In this exploratory study, we used an existing, well-
characterized dataset in which standardized and widely
used measures of behavioural characteristics, self-rated
traits and ability to infer others’ mental states were avail-
able. We attempted to operationalize and quantify cam-
ouflaging in adults with autism, for the first time in
the scientific literature, as the quantitative discrepancy
between the person’s ‘external’ behavioural presentation
in social–interpersonal contexts (measured by the ADOS)
and the person’s ‘internal’ status (dispositional traits
measured by the AQ and social cognitive capability
measured by the RMET).
We found that the operationalized camouflaging meas-
ure was not significantly correlated with age, VIQ, PIQ
or FIQ. On average, women with autism showed more
camouflaging than men with autism, but there was sub-
stantial variability in both groups. Greater camouflaging
was associated with more depressive symptoms in men
with autism and better signal-detection sensitivity in
women with autism. The brain volumetric associations
with camouflaging were largely sex/gender-dependent.
Following a positivist approach, our first aim was to
quantitatively describe camouflaging in autism through an
operationalization using existing measures. We acknowl-
edge that our operationalization is not a direct measure-
ment of camouflaging, which, in our view, could only be
created in a valid way after careful concept formation via
a qualitative (e.g. grounded theory) research approach
(Glaser and Strauss, 2009). Nevertheless, our operationali-
zation (i.e. the external–internal discrepancy) is a first step
and may provide a reasonable proxy.
Using this proxy measure, which follows a normal dis-
tribution in our sample, we observed that individual differ-
ences in the degree of camouflaging were independent of
age and IQ, in men and women with autism without intel-
lectual disability. This suggests that the extent of camou-
flaging in adults with autism does not merely mirror degree
at University of Cyprus on December 1, 2016aut.sagepub.comDownloaded from
8 Autism
Figure 3. Word-clouds showing (a) the top 60 terms correlated with brain regions associated with camouflaging in women with
autism, based on reverse inference using the Neurosynth Image Decoder, and (b) the top 30 terms after excluding anatomical terms.
Figure 2. Sagittal slices illustrating grey matter regions showing sex/gender-differential associations between CF and regional
volume (in red, involving left medial temporal lobe and cerebellum), overlaid with regions showing negative correlations between
CF and regional volume in women with autism (in blue, involving cerebellum, occipital and medial temporal structures); threshold
for visualization follows that described in the ‘Methods’ section.
at University of Cyprus on December 1, 2016aut.sagepub.comDownloaded from
Lai et al. 9
of experience (reflected in chronological age). If camou-
flaging is partly learnt, one might expect a correlation
with age/experience at younger ages. Recent studies have
alluded to the presence of camouflaging in teenage (Head
et al., 2014; Tierney et al., 2016) or even school age years
(Hiller et al., 2014; Rynkiewicz et al., 2016). The develop-
mental course of camouflaging remains to be revealed
by longitudinal studies. In particular, how it affects one’s
clinical experience (e.g. getting a timely autism diagnosis,
developing mental health challenges) should be a focus of
investigation. We also surprisingly found no significant
association between camouflaging and VIQ, PIQ and FIQ
in either men or women with autism. This suggests in this
population, the extent of camouflaging does not merely
reflect general reasoning ability or speed of processing.
Instead, it may be more specifically associated with par-
ticular aspects of cognitive ability (discussed below) or
personality, motivational or contextual factors.
We observed an on-average higher extent of camou-
flaging in women than men with autism of rather large
effect size (Cohen’s d = 0.98). This fits well with previous
findings from contrasting current ADOS and childhood
ADI-R scores in men and women separately (Lai et al.,
2011). It also corresponds well with the reports from
women with autism and their parents as well as expert
clinician’s observations (summarized in ‘Introduction’
section). Here, we are not able to delineate what contrib-
utes to this difference between men and women, but we
suspect that socio-cultural factors, in particular gender-
based expectations and gender socialization across devel-
opment, may be key players (Kreiser and White, 2014).
For example, protective same-gender friendship (i.e. being
‘mothered’) may conceal a girl or woman’s social diffi-
culties; gender-based expectations may prompt a girl or
woman with autism to ‘act like a girl/woman’ and ‘be more
social’, and she may therefore develop higher censuring of
own behaviours and more imitation or emulation of gen-
der-normative social behaviours. Behavioural compo-
nents contributing to the presentation and developmental
course of camouflaging should be explored via qualitative
research (Tierney et al., 2016) as an immediate future
research direction (Lai et al., 2015).
It is important to note from our data that although cam-
ouflaging might have been portrayed as an integral part
of the ‘female-phenotype of autism’, it is not specific to
females. Even with sex/gender differences of large effect,
the distributions of camouflaging score overlapped sub-
stantially between men and women with autism: there
were women who showed little camouflaging and men
who presented marked camouflaging (Figure 1(a)). On
average, a sex/gender difference in camouflaging is
evident, but it should be viewed as a phenomenon reflect-
ing individual differences in social coping, rather than a
diagnostic behavioural pattern distinguishing females
versus males with autism at an individual level.
We tested the hypotheses that higher camouflaging is
associated with higher anxiety and depression, and cogni-
tively with better verbal ability, better signal detection
from background events and more conservative responses.
Findings confirmed some of the hypotheses, yet in a
potentially sex/gender-dependent manner. When taking
sex/gender into consideration, trend-level significant sex/
gender-differential correlation patterns were observed
between camouflaging and the clinical and cognitive
We predicted that camouflaging is exhausting and
brings excessive stress and, therefore, may be associated
with anxiety and depressive symptoms. Based on the back-
ground that men and women with autism in this study
showed no differences in either symptom scores (but both
had elevated scores approaching clinical range; Table 1),
we found a pattern in support of the prediction for depres-
sive symptoms in men (r = 0.533, p = 0.002) but not women
with autism (r = 0.030, p = 0.876); we found no significant
relationship between camouflaging and anxiety symptoms
in either sex or gender. The nature of this study does not
allow for testing causal relationships, yet based on the
cross-sectional correlational patterns we suspect that the
lack of association with anxiety might indicate that camou-
flaging in adults is an already adapted behavioural pattern.
Investigation into the child and youth autistic population is
necessary to address any potential associations between
camouflaging and anxiety at younger ages. The positive
association with depressive symptoms (as predicted) in
men raises the possibility that they are more susceptible to
the burden of camouflaging than women with autism are
– perhaps women would have had more practice with, and
might be better adapted to, implementing camouflaging
due to gender-related social experience and demands.
These ideas await rigorous investigation of the causal
relationships between stress, anxiety/depression, cognitive
features, camouflaging and social adaptation, using struc-
tural equation modelling (for cross-sectional data) or lon-
gitudinal designs.
The association between camouflaging and cognitive
performance may shed light on potential cognitive under-
pinnings of camouflaging. Lehnhardt et al. (2016) sus-
pected that verbal abilities might serve an important role
for males with autism when it comes to camouflaging.
Contrary to this prediction, we did not find a significant
correlation between verbal ability and camouflaging in
either men or women with autism. This suggests that the
extent of camouflaging does not merely reflect verbal
knowledge or reasoning; rather, it might be associated
with verbal skills beyond these or might be underpinned
by other cognitive capabilities.
We then examined performance on a response inhibition
Go/No-Go task because among our available measures in
this project, the SDT parameters from the Go/No-Go task
most closely reflect the theoretical constructs of interest
at University of Cyprus on December 1, 2016aut.sagepub.comDownloaded from
10 Autism
(i.e. one requires sensitive real-time monitoring of the
environment, and/or a cautious, conservative response
strategy, to successfully camouflage). The background is
that men and women with autism in this study equally
showed on-average poorer sensitivity in detecting signal
from background compared to neurotypical controls, but
were no more liberal or conservative in response strategy
(see a previous study on a slightly larger but highly over-
lapping sample (Lai et al., 2012)). In this context, we again
found a potentially sex/gender-differential pattern. Women
(but not men) with higher camouflaging showed better sig-
nal-detection sensitivity, whereas there was no significant
association between camouflaging and response strategy
in either sex or gender. Lehnhardt et al. (2016) alluded to
possible sex/gender-differential cognitive underpinnings of
camouflaging by the indirect evidence that women with
autism show higher processing speed and better executive
functions (mainly in trail-making and verbal fluency tests)
than men with autism. Our findings echo this suggested
association between executive functions and camouflaging
in women with autism by directly showing a predicted rela-
tionship. Again, causal inference cannot be made: it could
be the case that better signal-detection supports and prompts
more camouflaging or that more frequent camouflaging
enhances cognitive control and signal-detection sensitivity.
In sum, the converging message points to the need for stud-
ying the relationships between camouflaging and executive
functions, particularly in females. Whether this is equally
important in males with autism is unclear. Direct hypothe-
sis testing concerning the cognitive bases of camouflaging
(and the examination of sex/gender-differential relation-
ships) is much needed.
Our last aim was purely exploratory and hypothesis gen-
erating. When testing for association between regional GM
volume and camouflaging, we found statistically signifi-
cant sex/gender-dependent association patterns while not
finding any region that showed a significant overall corre-
lation with camouflaging across sex/gender at the same sta-
tistical threshold. This indicates that the neuroanatomical
association of camouflaging in autism may be largely sex/
gender-dependent, particularly around the medial temporal
and cerebellar structures. When dissecting the sex/gender-
differential pattern, we found a lack of association in males
but a significant negative correlation in females (i.e. the
higher camouflaging, the smaller regional volume).
When using the Neurosynth Image Decoder for reverse
inference (Gorgolewski et al., 2016; Poldrack, 2006), that
is, to identify scientific terms in the ‘big data’ of the neuro-
science literature mostly associated with the voxels show-
ing significant correlation with camouflaging in women
with autism, we found anatomical terms (e.g. cerebellum,
medial temporal lobe, para/hippocampus, amygdala) as
well as cognitive terms about emotion and memory. This
exploratory, hypothesis-generating approach, in association
with the cognitive findings regarding executive functions
(for which the cerebellum is closely involved), gives
candidate neurocognitive components for future hypothe-
sis testing to uncover the bases of camouflaging, particu-
larly for females.
As the first study operationalizing and quantifying cam-
ouflaging, the findings should be considered exploratory
and have to be interpreted with caution, keeping in mind
the following limitations. First, camouflaging was quanti-
fied by a mathematical manipulation of available measures
based on our operationalization as the discrepancy between
the person’s ‘external’ behavioural presentation in social–
interpersonal contexts and the person’s ‘internal’ status.
Factors potentially affecting scoring of these contributing
measures will have impacts on the derived camouflaging
measure, and therefore, findings need to be interpreted
considering these potentially confounding factors. For
example, social communication behaviours measured by
the ADOS may be affected by one’s anxiety level during
the assessment, and gender stereotype of the examiner
may affect how behaviours are scored; dispositional traits
measured by the AQ may be influenced by one’s self-
referential ability and even intuitive/automatic masking
of difficulties; social cognitive ability measured by the
RMET may be affected by one’s lexicon and verbal abili-
ties. Additionally, although the content validity of the
camouflaging measure is ensured based on the concept of
external–internal discrepancy, whether this measure
shows satisfactory concurrent validity awaits comparisons
with future studies that also quantify camouflaging. For
example, when a substantially large sample is available,
quantifying camouflaging by regression methods (e.g. the
residuals after regressing out the variances of ‘internal/
actual’ characteristics from ‘external’ behavioural mani-
festations) could provide another metric to be examined.
In addition, instead of operationalizing camouflaging as
external–internal discrepancy, one could also operational-
ize it by social imitation capacities based on the hypothesis
that camouflaging heavily involves social imitation and
adaptation; these capacities could be measured by compo-
nents of well-established instruments such as the Self-
Monitoring Scale (Snyder, 1974), its revision (Lennox and
Wolfe, 1984) and the Multidimensional Iowa Suggestibility
Scale (Kotov et al., 2004).
Second, all analyses of the relationships between cam-
ouflaging and clinical symptoms, cognitive abilities and
regional brain volume were correlational in nature, and no
causal relationships could be inferred. Mechanisms dis-
cussed above are speculations and have to be tested using
longitudinal or intervention designs or hypothesis-based
modelling with more comprehensive data collection rele-
vant to this topic.
Third, due to the limitation of the dataset (i.e. we did
not perform ADOS for the control sample in the cohort),
we were not able to compare sex/gender difference in cam-
ouflaging in autism in the context of probable neurotypical
sex/gender difference. If there is an underlying neurotypi-
cal sex/gender difference, then the findings need to be
at University of Cyprus on December 1, 2016aut.sagepub.comDownloaded from
Lai et al. 11
interpreted accordingly (Lai et al., 2015). This further
examination is crucial but will be available only with stud-
ies capable of quantifying camouflaging in neurotypical
individuals, based on the same operationalization (i.e. the
external–internal discrepancy) but using measures other
than the ADOS that can quantify social communicative
behaviour and are sensitive enough to pick up individual
differences in the neurotypical population.
Fourth, the findings are derived from a moderate-sized
adult sample with autism and without intellectual disabil-
ity. The extent to which the findings generalize to the full
autism spectrum has to be further examined. We suspect
that beyond individual general and specific cognitive
factors, one’s personality, social experience and develop-
mental stage (which are associated with age-relevant
social demands), as well as the socio-cultural context, will
all have particular influences.
In conclusion, this study provides a first attempt to oper-
ationalize and quantify camouflaging in men and women
with autism, showing substantial inter-individual variabil-
ity but on-average higher levels in women than men, and
demonstrates potentially sex/gender-dependent associations
with depressive symptoms, signal-detection sensitivity and
regional brain volume. We urge more investigations into
this clinically important phenomenon to better delineate
the construct. Ideally, this should include (1) qualitative (or
mixed-design) approaches to reveal first-person account
and second/third-person observation about what triggers
(e.g. when and why one camouflages) and constitutes cam-
ouflaging (e.g. what the behavioural components are, and
which of them are automatic/intuitive vs requiring one to
act/perform with effort, and which of them are simply
masking vs compensating); (2) psychological studies to
understand the personality, cognitive and contextual bases
of camouflaging; and (3) clinical studies to assess the posi-
tive and negative consequences of camouflaging, as well as
how camouflaging has an impact on the diagnosis of autism
(e.g. whether higher levels of camouflaging result in delayed
or missed diagnosis) and the identification of relevant
clinical issues. These studies will benefit from examining
camouflaging-related factors not only in individuals cur-
rently having a clinical diagnosis of autism but also those
who may be on the spectrum (e.g. those having high-level
autistic-like traits and/or social adaptation difficulties yet
who have failed or have not yet to be diagnosed with autism),
in order to inform how camouflaging may have a real-world
healthcare impact. A thorough understanding of camouflag-
ing in autism may improve the diagnosis of autism across
sex/gender, the identification of needs and assets for each
person and the tailored individualized supports.
The Medical Research Council Autism Imaging Multicentre
Study Consortium (MRC AIMS Consortium) is a UK collabora-
tion between the Institute of Psychiatry, Psychology &
Neuroscience (IoPPN) at King’s College, London, the Autism
Research Centre, University of Cambridge, and the Autism
Research Group, University of Oxford. The Consortium mem-
bers are in alphabetical order: Anthony J. Bailey (Oxford),
Simon Baron-Cohen (Cambridge), Patrick F. Bolton (IoPPN),
Edward T. Bullmore (Cambridge), Sarah Carrington (Oxford),
Marco Catani (IoPPN), Bhismadev Chakrabarti (Cambridge),
Michael C. Craig (IoPPN), Eileen M. Daly (IoPPN), Sean C. L.
Deoni (IoPPN), Christine Ecker (IoPPN), Francesca Happé
(IoPPN), Julian Henty (Cambridge), Peter Jezzard (Oxford),
Patrick Johnston (IoPPN), Derek K. Jones (IoPPN), Meng-
Chuan Lai (Cambridge), Michael V. Lombardo (Cambridge),
Anya Madden (IoPPN), Diane Mullins (IoPPN), Clodagh M.
Murphy (IoPPN), Declan G. M. Murphy (IoPPN), Greg Pasco
(Cambridge), Amber N.V. Ruigrok (Cambridge), Susan A.
Sadek (Cambridge), Debbie Spain (IoPPN), Rose Stewart
(Oxford), John Suckling (Cambridge), Sally J. Wheelwright
(Cambridge) and Steven C. Williams (IoPPN).
Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with
respect to the research, authorship and/or publication of this
The study was supported by the UK Medical Research Council
(grant number GO 400061) and European Autism Interventions
– a Multicentre Study for Developing New Medications (EU-
AIMS); EU-AIMS has received support from the Innovative
Medicines Initiative Joint Undertaking under grant agreement no
115300, resources of which are composed of financial contribu-
tion from the European Union’s Seventh Framework Programme
(FP7/2007–2013) and EFPIA companies’ in-kind contribution.
The study was conducted in association with the National
Institute for Health Research Collaborations for Leadership in
Applied Health Research and Care (NIHR CLAHRC) East of
England (EoE). During the period of the study, Dr Lai was sup-
ported by the William Binks Autism Neuroscience Fellowship
and Wolfson College, Cambridge, UK, and the O’Brien Scholars
Program within the Child and Youth Mental Health Collaborative
at the Centre for Addiction and Mental Health and The Hospital
for Sick Children, Toronto, Canada. Drs Baron-Cohen, Lombardo,
Ruigrok, Chakrabarti, and Auyeung were supported by the
Autism Research Trust during the period of this work.
American Psychiatric Association (APA) (2000) Diagnostic
and Statistical Manual of Mental Disorders (DSM-IV-TR).
4th ed. Text revision. Washington, DC: APA.
Ashburner J (2007) A fast diffeomorphic image registration
algorithm. Neuroimage 38(1): 95–113.
Attwood T (2006) The pattern of abilities and development of
girls with Asperger’s syndrome. In: Attwood T, Bolick T,
Faherty C, et al. (eds) Asperger’s and Girls. Arlington, TX:
Future Horisons, Inc., pp. 1–8.
Attwood T (2007) The Complete Guide to Asperger’s Syndrome.
London: Jessica Kingsley Publishers.
Baron-Cohen S, Bowen DC, Holt RJ, et al. (2015) The ‘Reading
the Mind in the Eyes’ test: complete absence of typical sex
difference in ~400 men and women with autism. PLoS ONE
10(8): e0136521.
at University of Cyprus on December 1, 2016aut.sagepub.comDownloaded from
12 Autism
Baron-Cohen S, Wheelwright S, Hill J, et al. (2001a) The
‘Reading the Mind in the Eyes’ test revised version: a study
with normal adults, and adults with Asperger syndrome or
high-functioning autism. Journal of Child Psychology and
Psychiatry and Allied Disciplines 42(2): 241–251.
Baron-Cohen S, Wheelwright S, Robinson J, et al. (2005) The
Adult Asperger Assessment (AAA): a diagnostic method.
Journal of Autism and Developmental Disorders 35(6):
Baron-Cohen S, Wheelwright S, Skinner R, et al. (2001b) The
autism spectrum quotient (AQ): evidence from Asperger
syndrome/high-functioning autism, males and females,
scientists and mathematicians. Journal of Autism and
Developmental Disorders 31(1): 5–17.
Baxter AJ, Brugha TS, Erskine HE, et al. (2015) The epide-
miology and global burden of autism spectrum disorders.
Psychological Medicine 45(3): 601–613.
Beck AT, Epstein N, Brown G, et al. (1988) An inventory
for measuring clinical anxiety: psychometric properties.
Journal of Consulting and Clinical Psychology 56(6):
Beck AT, Ward CH, Mendelson ML, et al. (1961) An inventory
for measuring depression. Archives of General Psychiatry
4(6): 561.
Begeer S, Mandell D, Wijnker-Holmes B, et al. (2013) Sex dif-
ferences in the timing of identification among children and
adults with autism spectrum disorders. Journal of Autism
and Developmental Disorders 43(5): 1151–1156.
Boyd K, Woodbury-Smith M and Szatmari P (2011) Managing
anxiety and depressive symptoms in adults with autism-
spectrum disorders. Journal of Psychiatry & Neuroscience
36(4): E35–E36.
Chumbley JR and Friston KJ (2009) False discovery rate revis-
ited: FDR and topological inference using Gaussian random
fields. Neuroimage 44(1): 62–70.
Deoni SC, Williams SC, Jezzard P, et al. (2008) Standardized
structural magnetic resonance imaging in multicentre
studies using quantitative T1 and T2 imaging at 1.5 T.
Neuroimage 40(2): 662–671.
Dworzynski K, Ronald A, Bolton P, et al. (2012) How different
are girls and boys above and below the diagnostic thresh-
old for autism spectrum disorders? Journal of the American
Academy of Child and Adolescent Psychiatry 51(8): 788–797.
Ecker C, Ginestet C, Feng Y, et al. (2013) Brain surface anatomy
in adults with autism: the relationship between surface area,
cortical thickness, and autistic symptoms. JAMA Psychiatry
70(1): 59–70.
Ecker C, Suckling J, Deoni SC, et al. (2012) Brain anatomy and
its relationship to behavior in adults with autism spectrum
disorder: a multicenter magnetic resonance imaging study.
Archives of General Psychiatry 69(2): 195–209.
Ernsperger L and Wendel D (2007) Girls under the Umbrella
of Autism Spectrum Disorders: Practical Solutions for
Addressing Everyday Challenges. Shawnee Mission, KS:
Autism Asperger Publishing Company.
Giarelli E, Wiggins LD, Rice CE, et al. (2010) Sex differences
in the evaluation and diagnosis of autism spectrum disorders
among children. Disability and Health Journal 3(2): 107–116.
Gillberg C (1991) Clinical and neurobiological aspects of
Asperger syndrome in six families studied. In: Frith U (ed.)
Autism and Asperger Syndrome. Cambridge: Cambridge
University Press, pp. 122–146.
Glaser BG and Strauss AL (2009) The Discovery of Grounded
Theory: Strategies for Qualitative Research. London:
Transaction Publishers.
Gorgolewski KJ, Varoquaux G, Rivera G, et al. (2016) a repository for sharing unthresholded
statistical maps, parcellations, and atlases of the human
brain. Neuroimage 124(Pt B): 1242–1244.
Gould J and Ashton-Smith J (2011) Missed diagnosis or mis-
diagnosis: girls and women on the autism spectrum. Good
Autism Practice 12(1): 34–41.
Green DM and Swets JA (1966) Signal Detection Theory and
Psychophysics. New York: Wiley.
Hayasaka S, Phan KL, Liberzon I, et al. (2004) Nonstationary
cluster-size inference with random field and permutation
methods. Neuroimage 22(2): 676–687.
Head AM, McGillivray JA and Stokes MA (2014) Gender differ-
ences in emotionality and sociability in children with autism
spectrum disorders. Molecular Autism 5(1): 19.
Hendrickx S (2015) Women and Girls with Autism Spectrum
Disorder: Understanding Life Experiences from Early
Childhood to Old Age. London: Jessica Kingsley
Hiller RM, Young RL and Weber N (2014) Sex differences in
autism spectrum disorder based on DSM-5 criteria: evidence
from clinician and teacher reporting. Journal of Abnormal
Child Psychology 42(8): 1381–1393.
Howlin P and Moss P (2012) Adults with autism spectrum dis-
orders. Canadian Journal of Psychiatry 57(5): 275–283.
Howlin P, Moss P, Savage S, et al. (2013) Social outcomes in
mid- to later adulthood among individuals diagnosed with
autism and average nonverbal IQ as children. Journal of
the American Academy of Child and Adolescent Psychiatry
52(6): 572.e1–581.e1.
Idring S, Rai D, Dal H, et al. (2012) Autism spectrum disorders
in the Stockholm Youth Cohort: design, prevalence and
validity. PLoS ONE 7(7): e41280.
Jensen CM, Steinhausen HC and Lauritsen MB (2014) Time
trends over 16 years in incidence-rates of autism spec-
trum disorders across the lifespan based on nationwide
Danish register data. Journal of Autism and Developmental
Disorders 44(8): 1808–1818.
Kim YS, Leventhal BL, Koh YJ, et al. (2011) Prevalence of
autism spectrum disorders in a total population sample.
American Journal of Psychiatry 168(9): 904–912.
Kocovska E, Biskupsto R, Carina Gillberg I, et al. (2012) The
rising prevalence of autism: a prospective longitudinal study
in the Faroe Islands. Journal of Autism and Developmental
Disorders 42(9): 1959–1966.
Kopp S and Gillberg C (1992) Girls with social deficits and
learning problems: autism, atypical Asperger syndrome or
a variant of these conditions. European Child & Adolescent
Psychiatry 1(2): 89–99.
Kotov R, Bellman S and Watson D (2004) Multidimensional
Iowa Suggestibility Scale (MISS). Stony Brook, NY: Stony
Brook University.
Kreiser NL and White SW (2014) ASD in females: are we over-
stating the gender difference in diagnosis? Clinical Child
and Family Psychology Review 17(1): 67–84.
at University of Cyprus on December 1, 2016aut.sagepub.comDownloaded from
Lai et al. 13
Lai MC and Baron-Cohen S (2015) Identifying the lost gen-
eration of adults with autism spectrum conditions. Lancet
Psychiatry 2(11): 1013–1027.
Lai MC, Lombardo MV, Auyeung B, et al. (2015) Sex/gender
differences and autism: setting the scene for future research.
Journal of the American Academy of Child and Adolescent
Psychiatry 54(1): 11–24.
Lai MC, Lombardo MV, Pasco G, et al. (2011) A behavioral
comparison of male and female adults with high function-
ing autism spectrum conditions. PLoS ONE 6(6): e20835.
Lai MC, Lombardo MV, Ruigrok AN, et al. (2012) Cognition in
males and females with autism: similarities and differences.
PLoS ONE 7(10): e47198.
Lai MC, Lombardo MV, Suckling J, et al. (2013) Biological
sex affects the neurobiology of autism. Brain 136(Pt 9):
Lawson W (1998) Life behind Glass: A Personal Account of
Autism Spectrum Disorder. London: Jessica Kingsley
Lehnhardt FG, Falter CM, Gawronski A, et al. (2016) Sex-related
cognitive profile in autism spectrum disorders diagnosed
late in life: implications for the female autistic phenotype.
Journal of Autism and Developmental Disorders 46(1):
Lennox RD and Wolfe RN (1984) Revision of the self-monitor-
ing scale. Journal of Personality and Social Psychology 46:
Lord C, Risi S, Lambrecht L, et al. (2000) The autism diagnostic
observation schedule-generic: a standard measure of social
and communication deficits associated with the spectrum
of autism. Journal of Autism and Developmental Disorders
30(3): 205–223.
Lord C, Rutter M and Le Couteur A (1994) Autism Diagnostic
Interview-Revised: a revised version of a diagnostic inter-
view for caregivers of individuals with possible perva-
sive developmental disorders. Journal of Autism and
Developmental Disorders 24(5): 659–685.
Marshall TA (2015) I am AspienWoman: The Unique
Characteristics, Traits, and Gifts of Adult Females on
the Autism Spectrum. Sunshine Coast, QLD, Australia:
Mattila ML, Kielinen M, Linna SL, et al. (2011) Autism spectrum
disorders according to DSM-IV-TR and comparison with
DSM-5 draft criteria: an epidemiological study. Journal of
the American Academy of Child and Adolescent Psychiatry
50(6): 583.e11–592.e11.
Miller JK (2003) Women from Another Planet? Our Lives in the
Universe of Autism. Bloomington, IN: 1st Books Library.
Poldrack RA (2006) Can cognitive processes be inferred from
neuroimaging data? Trends in Cognitive Sciences 10(2):
Russell G, Steer C and Golding J (2011) Social and demographic
factors that influence the diagnosis of autistic spectrum
disorders. Social Psychiatry and Psychiatric Epidemiology
46(12): 1283–1293.
Rutherford M, McKenzie K, Johnson T, et al. (2016) Gender
ratio in a clinical population sample, age of diagnosis and
duration of assessment in children and adults with autism
spectrum disorder. Autism 20: 628–634.
Rynkiewicz A, Schuller B, Marchi E, et al. (2016) An investi-
gation of the ‘female camouflage effect’ in autism using a
computerized ADOS-2 and a test of sex/gender differences.
Molecular Autism 7: 10.
Shattuck PT, Durkin M, Maenner M, et al. (2009) Timing of
identification among children with an autism spectrum dis-
order: findings from a population-based surveillance study.
Journal of the American Academy of Child and Adolescent
Psychiatry 48(5): 474–483.
Simone R (2010) Aspergirls: Empowering Females with
Asperger Syndrome. London: Jessica Kingsley Publishers.
Snyder M (1974) Self-monitoring of expressive behavior. Journal
of Personality and Social Psychology 30(4): 526–537.
Tierney S, Burns J and Kilbey E (2016) Looking behind the
mask: social coping strategies of girls on the autistic spec-
trum. Research in Autism Spectrum Disorders 23: 73–83.
Wechsler D (1999) Wechsler Abbreviated Scale of Intelligence
(WASI). New York: The Psychological Corporation.
Willey LH (1999) Pretending to Be Normal: Living with
Asperger’s Syndrome. London: Jessica Kingsley Publishers.
Williams D (1992) Nobody Nowhere: The Remarkable
Autobiography of an Autistic Girl. London: Jessica Kingsley
Wilson CE, Happe F, Wheelwright SJ, et al. (2014) The neu-
ropsychology of male adults with high-functioning autism
or Asperger syndrome. Autism Research 7(5): 568–581.
Wilson CE, Murphy CM, McAlonan G, et al. (2016) Does sex
influence the diagnostic evaluation of autism spectrum dis-
order in adults? Autism 20: 808–819.
Wing L (1981) Sex ratios in early childhood autism and related
conditions. Psychiatry Research 5(2): 129–137.
World Health Organization (WHO) (1992) The ICD-10
Classification of Mental and Behavioural Disorders:
Clinical Descriptions and Diagnostic Guidelines. Geneva:
Zablotsky B, Black LI, Maenner MJ, et al. (2015) Estimated
prevalence of autism and other developmental disabili-
ties following questionnaire changes in the 2014 National
Health Interview Survey. National Health Statistics Reports
87: 1–20.
Zaks Z (2006) Life and Love: Positive Strategies for Autistic
Adults. Shawnee Mission, KS: Autism Asperger Publishing
at University of Cyprus on December 1, 2016aut.sagepub.comDownloaded from
... Pour dépasser ces contradictions et incertitudes, et expliquer la disparité entre les genres en matière de diagnostic et les difficultés pour conceptualiser les présentations cliniques des femmes autistes, une hypothèse connaît un essor indéniable dans la littérature et les discussions scientifiques dans ce champ [100], il s'agit de l'hypothèse du « camouflage » [76], suivant laquelle les personnes autistes, et peut-être plus vraisemblablement encore les femmes [101], parviennent consciemment ou inconsciemment, à cacher les caractéristiques autistiques et/ou minimiser la visibilité des difficultés sociales en employant des stratégies compensatoires. C'est une hypothèse qui prend de l'ampleur, de sorte que même le DSM-5 [3] fait mention, dans son critère C, de ce masquage et reconnaît que les symptômes « peuvent être masqués par des stratégies apprises au cours de la vie ultérieure ». ...
... Les discussions terminologiques sont d'ailleurs toujours ouvertes dans ce champ, notamment le terme de camouflage est critiqué pour ce qu'il sous-entend de la tromperie volontaire, tandis que certains auteurs soulignent davantage qu'il puisse s'agir d'une réponse automatique et orientée vers le maintien de sa propre sécurité [110]. C'est cependant le terme qui semble le plus unanimement utilisé dans ce champ de recherche et par la communauté autistique elle-même [34,101,108,110]. ...
... Il entraînerait, alors, selon certaines sources contrastées, des manifestations de tristesse et de stress, voire plus franchement des symptômes anxieux et dépressifs. Il pourrait même constituer un marqueur de risque plus élevé pour la suicidalité chez les adultes autistes [30,32,36,61,62,66,101,105,113,114]. ...
Résumé Objectifs La littérature scientifique suggère que le Trouble du spectre autistique (TSA), chez les femmes, peut se manifester, de façon différente, en raison de nombreux facteurs, impactant, de ce fait, l’obtention de ce diagnostic, en particulier à l’âge adulte. D’après des données récentes, la prévalence du diagnostic de TSA évoluerait vers un rapport de trois hommes pour une femme. Plusieurs auteurs ont tenté d’expliquer ce ratio en évoquant, notamment, de possibles ressources subjectives amenant les femmes à ne pas présenter les manifestations autistiques attendues par les outils standards. En particulier, certaines « stratégies de camouflage » ne seraient pas détectées lors des évaluations habituelles. Or, ces stratégies comportent certains désavantages, notamment sur le registre de la construction identitaire. Méthode Nous avons procédé à une revue de littérature sur les caractéristiques et expressions du phénotype de l’autisme féminin et sur les stratégies dites de « camouflage ». Dans ce cadre, nous avons traduit en français le Camouflaging Autistic Traits Questionnaire (CAT-Q), développé et validé par Hull et ses collègues (2019), afin d’encourager son évaluation scientifique dans le cadre de recherches sur le diagnostic de TSA pour des femmes. Résultats Nous avons alors repéré plusieurs raisons aux origines des problématiques diagnostiques rencontrées par des femmes autistes : l’aspect androcentrique de la recherche sur l’autisme ; les biais sexuels ; la complexité diagnostique en lien avec les nombreuses comorbidités ; ainsi que l’inadéquation des instruments et dispositifs diagnostiques. L’évolution de l’autisme au féminin suppose aussi que la construction de la personnalité vient moduler l’expression de ces troubles. Discussion Ce travail ouvre, à la fois, à une réflexion pour l’amélioration des procédures et outils diagnostiques pour tous les sujets autistes et à un intérêt clinique pour leurs ressources psychiques. Nous suggérons aussi que l’emploi d’une épreuve projective pourrait permettre de contourner certains des biais ainsi manifestés par les instruments usuels. Conclusion Ces discussions sont à poursuivre pour accompagner la prise de conscience des rapports entre l’autisme et le genre et avancer sur la modélisation des « stratégies de camouflage », cela afin de proposer des adaptations dans les pratiques cliniques et diagnostiques, répondre aux enjeux importants autour de la population de femmes autistes, et peut-être plus globalement participer à repenser l’hétérogénéité et les définitions de l’autisme.
... These methods broadly fall under two categories: internal-external discrepancy and selfreport approaches (Hull et al., 2019). Internal-external discrepancy approaches focus on quantifying the degree to which an individual's autistic social difficulties or differences are camouflaged during an interaction; that is, quantifying the difference between an individual's 'true' autistic state and their observable behavioural presentation (e.g., Lai et al., 2017. In contrast, self-report approaches focus on quantifying an individual's use of specific camouflaging strategies or behaviour via self-reflection (e.g., Hull et al., 2019;Livingston et al., 2020). ...
... Consequently, several new theories and concepts related to social models of autism have emerged (e.g., neurodiversity, the double empathy problem, stigma, and autistic perspectives). A line of research that has recently emerged against this backdrop examines ways in which autistic people cope, adapt to, and influence the predominately neurotypical social world by camouflaging (Dean et al., 2017;Hull et al., 2017;Lai et al., 2017;Schuck et al., 2019). ...
... Table 3 provides an overview of participant characteristics for participants with autism diagnoses or high autistic traits. There were four sets of studies in which samples were partially or fully duplicated (Hull et al., 2019;Hull, Lai, et al., 2020;Hull, Levy, et al., 2021;Jorgenson et al., 2020;Bernardin et al., 2021;Lai et al., 2017Livingston, Shah, & Happé, 2019;Livingston et al., 2020). In these instances, only information from one study (the study with the largest N) was counted when calculating aggregated participant characteristics. ...
Some autistic individuals modify their innate autistic social behaviour in order to adapt to, cope within, and/or influence the predominately non-autistic social environment; a phenomenon often termed ‘camouflaging’ (Attwood, 2007; Dean et al., 2017; Hull et al., 2017; Lai et al., 2017; Schuck et al., 2019). Camouflaging is one social coping strategy used by autistic people attempting to overcome social challenges within cross-neurotype social interactions and secure employment, develop friendships and romantic relationships, and avoid stigmatisation (Cage & Troxell-Whitman, 2019; Hull et al., 2017). Yet the act of camouflaging is thought to be cognitively effortful and taxing; prone to breakdown under increased social demands and complexity and/or psychological distress; and associated with increased mental health difficulties, misdiagnosis, and identity confusion (e.g., Beck et al., 2020; Cage & Troxell-Whitman, 2019; Cassidy et al., 2018; Hull et al., 2021; Lai et al., 2017; Livingston, Colvert, et al., 2019). Camouflaging research is in infancy; conceptualisations, definitions and measures of camouflaging are still emerging, and much is unknown about relationships between camouflaging and various constructs such as mental health, wellbeing, and the achievement of important social and employment outcomes. This thesis presents a combination of qualitative and quantitative methods to further current understanding of social coping in autistic people by furthering the current conceptualisation of camouflaging including camouflaging behaviours and processes; examining the relationships between camouflaging and social, employment, and mental health outcomes; and exploring social experiences that contrast with camouflaging. The first chapter provides a general introduction to, and overview of, the relevant background research and provides a rationale for the work presented in the thesis. Chapter 2 involves a discussion of methodological considerations involved in the design and analysis of research presented in the thesis. Chapter 3, a systematic review, provides a comprehensive and critical evaluation of the current quantitative camouflaging research base; identifying consistencies in the current evidence as well as issues that require further research. Chapters 4 and 5 describe an interpersonal recall study, using thematic analysis to detail the development, process, and consequences of camouflaging (Chapter 4) and content analysis to describe the behaviours exhibited, altered, or avoided by autistic adults when camouflaging (Chapter 5). Chapter 6, a quantitative cross-sectional study, details associations between camouflaging and social and employment outcomes and indicators of mental health difficulties/psychological distress. Chapter 7 involves a qualitative survey and uses thematic analysis to explore an alternative to camouflaging, specifically autistic adults’ experiences of socialising in ways that feel authentic to them. The final chapter (Chapter 8) provides an overarching discussion of the findings and implications of the thesis with consideration to strengths and limitations.
... These same executive functions have been found to be impaired in people with autism [29,30], with the most severe impairments being reported in diagnosed females [31]. However, ASD females tend to camouflage many of these symptoms either through internalization or by masking, in which observed social behaviours are mirrored [32][33][34]. Due to this camouflaging effect, females often take longer to be diagnosed than males or they may go their entire lives without being diagnosed [33,34]. The internalization of symptoms often leads to more overall distress and a higher likelihood of comorbidities such as depression or anxiety among females with ASD [32]. ...
... Due to this camouflaging effect, females often take longer to be diagnosed than males or they may go their entire lives without being diagnosed [33,34]. The internalization of symptoms often leads to more overall distress and a higher likelihood of comorbidities such as depression or anxiety among females with ASD [32]. ...
Oxytocin is an endogenous neuropeptide hormone that influences social behaviour and bonding in mammals. Variations in oxytocin receptor (OXTR) expression may play a role in the social deficits seen in autism spectrum disorder. Previous studies from our laboratory found a dense population of OXTR in the human substantia nigra (SN), a basal ganglia structure in the midbrain that is important in both movement and reward pathways. Here, we explore whether differences in OXTR can be identified in the dopaminergic SN pars compacta of individuals with autism. Postmortem human brain tissue specimens were processed for OXTR autoradiography from four groups: males with autism, females with autism, typically developing (TD) males and TD females. We found that females with autism had significantly lower levels of OXTR than the other groups. To examine potential gene expression differences, we performed in situ hybridization in adjacent slides to visualize and quantify OXTR mRNA as well as mRNA for tyrosine hydroxylase. We found no differences in mRNA levels for either gene across the four groups. These results suggest that a dysregulation in local OXTR protein translation or increased OXTR internalization/recycling may contribute to the differences in social symptoms seen in females with autism. This article is part of the theme issue ‘Interplays between oxytocin and other neuromodulators in shaping complex social behaviours’.
An autism diagnosis in adulthood can increase women’s self-compassion and inform appropriate supports. This study explored what helped and hindered autistic women when accessing an adulthood autism diagnosis in Australia. Ten autistic women diagnosed as adults within the past 5 years participated in a semi-structured interview about their experiences of accessing a diagnosis. Framework analysis was used to identify barriers and facilitators on person, provider and system levels. Person-level factors included women’s recognition of their autism, motivation, preparation, social support and approach during the assessment. Provider-level factors related to providers’ level of knowledge and skill in working with autistic women, as well as the women’s experience of being dismissed or misdiagnosed. Interactions between person- and provider-level factors highlighted that the diagnostic process is relational. System-level factors included the requirements of the diagnostic process (time, financial costs, and technology) and the nature of diagnostic criteria and assessment tools used. These factors provided the context in which person-level and provider-level factors operated. The experiences of participants highlight improvements that could be made to accessing an adulthood autism diagnosis for women in Australia, including provider knowledge of the heterogeneity of autism and the development of resources to help autistic women prepare for their diagnostic assessment. Lay abstract An autism diagnosis can have a big impact on women and make it possible to access support. This study explored women’s experiences of being diagnosed with autism as an adult in Australia, to try to understand what was helpful (facilitators) and unhelpful (barriers) for them during this process. We interviewed 10 autistic women who had been diagnosed in the last 5 years. Framework analysis was used to understand the data. We wanted to understand barriers and facilitators relating to the individual participants, the professionals they saw and the system they went through for their diagnostic assessment. Women reported that being able to recognise they were autistic, being motivated, preparing for the assessment, having social support and unmasking to be themselves were helpful during the diagnostic process. They reported that having a knowledgeable diagnostician who made accommodations for their needs assisted them during the assessment process. When providers dismissed the participants when they first raised the possibility they were autistic, it delayed them in seeking an assessment. At the system level, the women in this study found some aspects of the healthcare system difficult to navigate, particularly costs and long waitlists. Some found the assessment tools used were not well suited to them. The experiences of the women in this study highlight improvements that could be made to accessing an adulthood autism diagnosis in Australia. These include improving provider knowledge of the varied presentation of autism and the development of resources to help autistic women prepare for their diagnostic assessment.
Full-text available
Surfacing the Perspective of Autistic Girls Aged Between Thirteen and Eighteen Within a Complex Social Discourse on Autism: A Qualitative Inquiry
Among individuals with eating disorders (ED), those with co-occurring autism are often considered to have more severe presentations and poorer prognosis. However, previous findings have been contradictory and limited by small sample size and/or cross-sectional assessment of autistic traits. We examine the hypothesis that autism diagnosis and autism polygenic score (PGS) are associated with increased ED severity in a large ED cohort using a broad range of ED severity indicators. Our cohort included 3189 individuals (64 males) born 1977-2000 with current or previous anorexia nervosa who participated in the Anorexia Nervosa Genetics Initiative-Sweden (ANGI-SE) and for whom genotypes and linkage to national registers were available. We identified 134 (4.2%) individuals with registered autism diagnoses. Individuals with confirmed autism diagnosis had significantly more severe ED across three sets of severity indicators. Some of the largest effects were found for the proportion of individuals who attempted suicide and who received tube feeding (higher in autism), and for the time spent in inpatient care (longer in autism). Results for autism PGS were not statistically significant. Adapting ED treatment to the needs of individuals with co-occurring autism is an important research direction to improve treatment outcome in this group.
Full-text available
Girls and boys might differ in autistic symptoms and associated cognitive difficulties such as executive function (EF). We investigated sex differences in the relationship between parent rated EF and autistic symptoms in 116 children and adolescents (25 girls) aged 5–19 years with an intelligence quotient above 70 and an autism spectrum disorder (ASD) diagnosis. They were rated with the behavior rating inventory of executive function (BRIEF) and the autism diagnostic interview revised (ADI-R). We found a positive association between EF and the ADI-R domains of reciprocal social interaction ( p < 0.001) and communication ( p = 0.001) in girls, while these relationships were small and non-significant in boys. Our results provide a greater understanding of the sex-specific characteristics of children and adolescents with ASD.
Full-text available
In this paper, we report on a participatory oral history study documenting the lives of late-diagnosed autistic adults in Australia. We interviewed 26 autistic adults about their life history and the impact of late diagnosis. All were diagnosed after the age of 35, growing up in an era when autism was not well known. Using reflexive thematic analysis, we uncovered a rich body of reflections on shared Autistic identity and identified three major themes within that data set: ‘conceptualising the Autistic family’, ‘creating Autistic community’, and ‘contesting Autistic identity’. Overall, the study provides insights into the active creation of shared Autistic identity and the importance of Autistic community to these late-diagnosed autistic adults.
‘Neurodiversity’ is a term used to describe people with neurological variation; this includes people with autism, dyspraxia, dyslexia and attention deficit hyperactivity disorder. Studies have shown an increased prevalence of malocclusion and dental trauma in people with these conditions. Unfortunately, a lack of understanding around such neurological differences can create barriers when accessing orthodontic care. The aim of this article is to raise awareness around the subject and to suggest reasonable adjustments to practice which may subsequently be used by the orthodontic team to improve the patient experience. CPD/Clinical Relevance: Recognition of neurodiversity and the need to make reasonable adjustments to orthodontic care may aid compliance and improve outcomes in this group of patients.
Full-text available
It is unknown whether sex influences the diagnostic evaluation of autism spectrum disorder, or whether male and female adults within the spectrum have different symptom profiles. This study reports sex differences in clinical outcomes for 1244 adults (935 males and 309 females) referred for autism spectrum disorder assessment. Significantly, more males (72%) than females (66%) were diagnosed with an autism spectrum disorder of any subtype (x2 = 4.09; p = 0.04). In high-functioning autism spectrum disorder adults (IQ > 70; N = 827), there were no significant sex differences in severity of socio-communicative domain symptoms. Males had significantly more repetitive behaviours/restricted interests than females (p = 0.001, d = 0.3). A multivariate analysis of variance indicated a significant interaction between autism spectrum disorder subtype (full-autism spectrum disorder/partial-autism spectrum disorder) and sex: in full-autism spectrum disorder, males had more severe socio-communicative symptoms than females; for partial-autism spectrum disorder, the reverse was true. There were no sex differences in prevalence of co-morbid psychopathologies. Sex influenced diagnostic evaluation in a clinical sample of adults with suspected autism spectrum disorder. The sexes may present with different manifestations of the autism spectrum disorder phenotype and differences vary by diagnostic subtype. Understanding and awareness of adult female repetitive behaviours/restricted interests warrant attention and sex-specific diagnostic assessment tools may need to be considered.
Full-text available
Background: Autism spectrum conditions (autism) are diagnosed more frequently in boys than in girls. Females with autism may have been under-identified due to not only a male-biased understanding of autism but also females' camouflaging. The study describes a new technique that allows automated coding of non-verbal mode of communication (gestures) and offers the possibility of objective, evaluation of gestures, independent of human judgment. The EyesWeb software platform and the Kinect sensor during two demonstration activities of ADOS-2 (Autism Diagnostic Observation Schedule, Second Edition) were used. Methods: The study group consisted of 33 high-functioning Polish girls and boys with formal diagnosis of autism or Asperger syndrome aged 5-10, with fluent speech, IQ average and above and their parents (girls with autism, n = 16; boys with autism, n = 17). All children were assessed during two demonstration activities of Module 3 of ADOS-2, administered in Polish, and coded using Polish codes. Children were also assessed with Polish versions of the Eyes and Faces Tests. Parents provided information on the author-reviewed Polish research translation of SCQ (Social Communication Questionnaire, Current and Lifetime) and Polish version of AQ Child (Autism Spectrum Quotient, Child). Results: Girls with autism tended to use gestures more vividly as compared to boys with autism during two demonstration activities of ADOS-2. Girls with autism made significantly more mistakes than boys with autism on the Faces Test. All children with autism had high scores in AQ Child, which confirmed the presence of autistic traits in this group. The current communication skills of boys with autism reported by parents in SCQ were significantly better than those of girls with autism. However, both girls with autism and boys with autism improved in the social and communication abilities over the lifetime. The number of stereotypic behaviours in boys significantly decreased over life whereas it remained at a comparable level in girls with autism. Conclusions: High-functioning females with autism might present better on non-verbal (gestures) mode of communication than boys with autism. It may camouflage other diagnostic features. It poses risk of under-diagnosis or not receiving the appropriate diagnosis for this population. Further research is required to examine this phenomenon so appropriate gender revisions to the diagnostic assessments might be implemented.
This article reports on gender ratio, age of diagnosis and the duration of assessment procedures in autism spectrum disorder diagnosis in a national study which included all types of clinical services for children and adults. Findings are reported from a retrospective case note analysis undertaken with a representative sample of 150 Scottish children and adults recently diagnosed with autism spectrum disorder. The study reports key findings that the gender ratio in this consecutively referred cohort is lower than anticipated in some age groups and reduces with increasing age. The gender ratio in children, together with the significant difference in the mean age of referral and diagnosis for girls compared to boys, adds evidence of delayed recognition of autism spectrum disorder in younger girls. There was no significant difference in duration of assessment for males and females suggesting that delays in diagnosis of females occur prior to referral for assessment. Implications for practice and research are considered.
The socio-communication deficits of individuals on the autistic spectrum are well documented. However, this has largely been based on the male population and less is known about how females with autism manage social relationships across the developmental period. Ten adolescent female participants with a diagnosis of autism were interviewed and Interpretative Phenomenological Analysis used to explore their experiences of managing their social relationships. The results showed that participants were motivated to develop and maintain friendships, but during adolescence this became increasingly difficult. Consequently, they developed explicit strategies to manage these relationships, including masking and imitation. The use of such strategies was both advantageous and disadvantageous, and such findings should be considered in developing gender sensitive assessment and developmentally appropriate support.