1146477

research-article2023
AUT0010.1177/13623613221146477AutismHorwitz et al.

Original Article

Autism

Sex differences in the course of autistic and 2023, Vol. 27(6) 1716­–1729
© The Author(s) 2023

co-occurring psychopathological symptoms Article reuse guidelines:

in adolescents with and without autism sagepub.com/journals-permissions

DOI: 10.1177/13623613221146477
https://doi.org/10.1177/13623613221146477
journals.sagepub.com/home/aut
spectrum disorder

Ernst Horwitz1 , Melissa Vos2, Annelies De Bildt2,

Kirstin Greaves-Lord3,4, Nanda Rommelse5, Robert Schoevers2
and Catharina Hartman2

Abstract
There is an ongoing debate as to whether autism spectrum disorder (ASD) is expressed differently in women than
men. It is unclear whether differences found are specific to autism or merely reflecting normative development. In
this study, we compared sex differences in developmental trajectories of autistic and co-occurring psychopathological
symptoms in adolescents with milder forms of autism to those in a normative group matched for intelligence quotient
(IQ) and socioeconomic status. Data of five assessment waves from ages 11 to 22 years were analyzed using linear
mixed modeling. We found that in adolescence, sex differences in developmental trajectories of psychopathological
symptoms specific for autism are confined to the repetitive stereotyped domain (males had higher scores on the
sensory/stereotypic and resistance to change domains, the latter difference disappeared during adolescence due to an
increase of these problems in females with ASD). Other sex differences, among which an increase over time in affective
and anxiety problems in females was the most outstanding, were also observed in typically developing females. These
sex-specific differences have relevance in the clinical care of men and women with autism, although they are subtle
compared to differences between individuals with and without autism, which are broadly present in internalizing and
externalizing problem domains.

Lay abstract
There is an ongoing debate as to whether autism spectrum disorder (ASD) is expressed differently in women than
men. Studies on sex differences in autistic symptoms and symptoms of other psychiatric problems present in individuals
with autism generally do not include a general population comparison group, making it unclear whether differences are
specific to autism or merely reflecting development in the general population. In this study, we compared sex differences
in the course of autistic and at the same time present symptoms of other psychiatric problems in adolescents with milder
forms of ASD to those in a group of the general population with an equal intelligence quotient (IQ) and socioeconomic
status. Data of five assessment moments from ages 11 to 22 years were analyzed using a statistic procedure that allowed
us to determine which factors affect the course of symptoms over time. We found that in adolescence, sex differences
in the course of psychopathological symptoms specific for autism are confined to the repetitive stereotyped domains.
Males had higher scores on the sensory/stereotypic and resistance to change domains, the latter difference disappeared
during the course of adolescence due to an increase of these problems in autistic females. Other sex differences, among
which an increase over time in mood and anxiety problems in females was the most outstanding, were also observed
in females without autism. These sex-specific differences have relevance in the clinical care of autistic men and women,
although they are subtle compared to differences between individuals with and without autism.

5
1
GGZ Friesland, The Netherlands Karakter Child and Adolescent Psychiatry University Centre, The
2
University Medical Center Groningen, The Netherlands Netherlands
3
Erasmus Medical Center Rotterdam, The Netherlands Corresponding author:
4
Sophia Children’s Hospital, The Netherlands Ernst Horwitz, GGZ Friesland, Leeuwarden, The Netherlands.
Email: [email protected]
Horwitz et al. 1717

Keywords
adolescents, autism spectrum disorders, course, psychiatric comorbidity, sex differences

Introduction Consequently, to increase our understanding of sex dif-

ferences in autism, it is important to have detailed infor-
Like in most neurodevelopmental disorder, there is a male mation on these differences in prevalence and course of
preponderance in autism spectrum disorder (ASD). A autistic and comorbid psychopathological symptomatol-
recent systematic review and meta-analysis by Loomes ogy (i.e. co-occurring internalizing and externalizing
and colleagues (2017) summarized the prevalence studies symptoms) in autistic persons, while at present, only
on autism and reported a large variability across studies coarse information is available. Autism has a multifaceted
with an overall male-to-female ratio of 4.2:1 in children presentation, and it is plausible that sex differences differ
aged 0–18 years. Intelligence quotient (IQ) has most con- along these problem domains within autism; therefore, an
sistently been associated with the sex difference in autism analysis of multidomain sex differences is relevant. In
(Rivet & Matson, 2011), with the proportion of males addition, a reference sample of same-aged individuals
increasing among participants with higher IQ (Volkmar & without autism is needed in order to determine if potential
Szatmari, 1993). Another factor is the age of participants, sex differences in psychopathology in persons with autism
in the meta-analysis of Loomes et al. (2017), the male-to- reflect normative (i.e. similar sex differences as in typi-
female ratio was higher in the age group over 6 years than cally developing peers) or autism-specific developmental
under 6 years. Researchers have also related the disparity patterns (i.e. different sex differences compared to typi-
to differences in genetic background, hormonal status, and cally developing peers). A recent study by Demetriou and
neurocognitive development, but with mixed results (May colleagues (2021), in which sex differences in cognitive
et al., 2019; Rivet & Matson, 2011). Recently, social and performance shown not to be specific to the autistic group,
cultural factors that might contribute to underidentification underlines the importance of the inclusion of a normative
of autism in females have been put forward. More specifi- group.
cally, a steep increase is seen in the number of publications In this study, we will focus on sex differences in the
on the hypothesis of a “female autistic phenotype” and development of autism, internalizing (anxiety and depres-
“camouflaged” autism symptoms in females diagnosed sion), attention-deficit/hyperactivity disorder (ADHD)
later in life (adolescence or afterwards) (e.g. Bargiela and externalizing (aggression) symptoms in persons with
et al., 2016; Hull et al., 2020). Central to this idea is that and without a clinical diagnosis of autism from late child-
that there is a female-specific manifestation of autism hood to early adulthood. In the next paragraph, we will
symptoms and co-occurring traits. Compared to boys/men, summarize what is already known about sex differences in
girls/women with autism are believed to show relatively the developmental course of autism, internalizing, ADHD,
higher social motivation, a greater capacity for traditional and externalizing symptomatology in autistic persons, as
friendships, fewer repetitive and stereotyped behaviors, well as in the general population.
less-externalizing behaviors (hyperactivity/impulsivity
and conduct problems), yet more internalizing problems
(anxiety, depression, and eating disorders) (Bargiela et al., The course of autistic symptoms
2016; Hull et al., 2020). Furthermore, girls/women are during adolescence in males and
believed to be inclined, more than boys/men, to hide their
females
autistic behavior by masking and using compensatory
strategies (“camouflaging”) to navigate the social world Studies on the developmental course of autistic traits in
(Lai et al., 2020). As a consequence, girls/females with adolescents in the general population are confined to the
autism may more often be misdiagnosed (foremost with social-communicative problem domain of autism. Overall,
internalizing and/or personality disorders) or late-diag- the literature suggests that boys score higher than girls in
nosed compared to boys/males with autism. This seems to early adolescence (Mandy et al., 2018; Robinson et al.,
be especially the case in girls/females at the milder end of 2011); however, the study by Mandy et al. showed an
the autism spectrum (Lai et al., 2017), in which group, the increase in scores in girls from 10 to 16 years that was
autistic symptoms are harder to discern from other forms steeper than in boys. As a result, the sex difference in
of psychopathology or extremes of normotypic behavior. scores at age 10 disappeared during adolescence. Results
Despite the apparent clinical validity of this theory, studies on sex differences in autistic symptoms in persons with
on this topic have been mainly conducted with small con- autism come from cross-sectional studies and are some-
venience samples that lack a proper longitudinal case–con- what inconsistent due to a large variability in IQ, age, and
trol design (Fombonne, 2020). One of the alternative autism severity (Fountain et al., 2012; Louwerse et al.
explanations for the reported subjective sensations of “pre- 2015; Simonoff et al., 2019; Woodman et al., 2016). In a
tending” in late-diagnosed females with autism may be systematic review and meta-analysis of cross-sectional
social anxiety (Fombonne, 2020). studies that reported on sex and age differences in
1718 Autism 27(6)

symptom severity in the core triad of autism problem exhibited a slight decrease in symptoms and males exhib-
domains as measured by—mainly parent report—stand- ited a stable course from mid-to-late adolescence
ardized instruments (in toddlers and pre-schoolers, chil- (Legerstee et al., 2013; Ohannessian et al., 2017; Van Oort
dren, adolescents and adults), Van Wijngaarden-Cremers et al., 2009). A sex difference in the course of depressive
et al. (2014) reported no sex differences in social behavior symptoms emerged early in adolescence when girls’ symp-
and communication problems, but more repetitive and ste- toms accelerated, whereas boys’ symptoms accelerated
reotyped behaviors in boys than girls aged 6–18. Kaat later in adolescence, but did not reach female levels of
et al. (2021) combined several databases to create a large symptoms indicating a stable higher female-to-male pre-
sample of girls and boys with an autism diagnosis to evalu- ponderance of depressive symptoms (Hankin et al., 2015;
ate sex differences in scores on standardized measures of Salk et al., 2016). Thus, sex differences in internalizing
autism symptoms. When matched for age, IQ, and lan- problems increase during adolescence and this seems to be
guage level, they found minimal sex differences (effect driven predominantly by depressive problems.
sizes < .20); boys received more severe scores on parent- ADHD symptoms on average decline from approxi-
reported and clinician-administered measures of restricted mately age 12, with inattention remaining relatively stable
repetitive behavior. A meta-analysis by Mahendiran et al. or declining at a modest rate, and hyperactivity/impulsiv-
(2019) of studies that included a normative comparison ity waning more strongly and remitting more abruptly
group on sex differences in social and communication (Hartman et al., 2016; Vos et al., 2021). Studies on sex-
symptoms in children and adolescents with autism again specific ADHD trajectories through adolescence are scarce
reported no significant differences between males and and showed contradictory results (e.g. Malone et al., 2010;
females but noted significant heterogeneity in the reviewed Murray et al., 2019). The developmental course of ADHD
studies. Wood-Downie and colleagues (2021) conducted a symptoms leads to a decline over time in the sex ratio of
systematic review and meta-analysis of gender differences ADHD (male-to-female ratio of 3:1 in childhood, closer to
on narrow construct domains like social attention and peer 1:1 in adulthood; Huang et al., 2016).
relationships and found that autistic females had better Most studies on the course of externalizing problems
social interaction and communication skills than autistic (aggression and oppositional behavior) in childhood and
males; for non-autistic individuals, similar gender differ- adolescence have also demonstrated a decline of symp-
ences were found. Groups were not matched for IQ. They toms with age (e.g. Bongers et al., 2004; Roskam, 2019).
concluded that standardized measures that are based on With regard to sex differences in externalizing problems in
broad constructs might not capture these differences which adolescents, Fernandez et al. (2014) found that boys
potentially contribute to under recognition of autism in showed higher levels of externalizing symptoms than girls
females. in a large sample from the general population. These
Overall, sex differences in social and communication symptoms decreased in both groups from age 11 to 15,
symptoms as measured by standardized diagnostic instru- indicating stable higher externalizing problems in males
ments in autistic persons seem to be absent. Sex differ- than females. Other studies converge on this pattern
ences found on some subdomains of social interaction, and (Bongers et al., 2004; Karriker-Jaffe et al., 2008) and show
communication seem to be alike in autistic and non-autis- that higher aggression in males than females is still present
tic persons. More repetitive and stereotyped behavior has in young adulthood (Boyd et al., 2015).
been reported in males than in females with autism; how- In children and adolescents with autism, higher levels
ever, with regard to this symptom domain, we did not find of depressive and anxiety symptoms than in peers in the
a study which included a normative comparison group. general population or peers with other developmental dis-
abilities have been reported (Gotham et al., 2015; Kim
The course of internalizing, ADHD, et al., 2000; Verheij et al., 2015; Woodman et al., 2016).
and externalizing symptoms during Sex differences in comorbid psychopathology in autistic
adolescents have not been investigated thoroughly, with
adolescence in males and females the findings so far summarized mostly based on cross-sec-
Multiple studies that examined the course of internalizing tional data, which were heterogeneous in terms of age,
problems throughout adolescence in the general popula- intelligence, and autism severity, and with inconsistent
tion have shown that these increase in early adolescence, results (Holtmann et al., 2007; Nasca et al., 2019; Oswald
peak in mid-adolescence and decrease into young adult- et al., 2016; Pisula et al., 2017). In a longitudinal study,
hood (e.g. Petersen et al., 2018). Females developed higher Gotham and colleagues (2015) compared trajectories of
levels of depressive and anxiety symptoms than males depressive and anxiety symptoms in participants with
around the onset of puberty (Hankin et al., 1998; Lewinsohn autism (n = 56) and with non-spectrum developmental
et al., 1998). Distinct gender-specific developmental tra- delay (n = 109) from school-age (age 6) through young
jectories of anxiety and depression during adolescence adulthood (age 24) and found that parent-rated symptom
have been identified. With regard to anxiety, females levels tended to start and remain higher in the autism group
Horwitz et al. 1719

compared to the comparison group. In both diagnostic females) from the clinical cohort of the Tracking
groups, anxiety and depressive symptoms were higher in Adolescents’ Individual Lives Survey (TRAILS; Huisman
males than in females at study outset and remained rela- et al., 2008; Oldehinkel et al., 2015), and a sex, age, and
tively stable over time, while symptom level in females SES-matched comparison group derived from the TRAILS
increased throughout adolescence causing the sexes to population cohort (n = 152). TRAILS is a prospective
converge on anxiety and depressive symptom in adult- study aiming to explain the development of mental health
hood. These results diverge from the aforedescribed stud- from early adolescence (approximately age 11) into adult-
ies in the general population, where boys and girls exhibit hood, with bi- or triennial assessments since. In this study,
around equal internalizing symptoms until adolescence, we used five assessment waves (T1–T5) between age 11
and subsequently, girls develop higher levels of internal- and 22. The population cohort started in 2001 and consists
izing symptoms than boys. We did not find studies on the of 2230 individuals who were selected from primary
developmental course of ADHD and externalizing symp- schools in five municipalities in the North of the Netherland
toms in autistic adolescents. including both urban and rural areas. The clinical cohort
started in 2004 and consists of 543 individuals who had
been referred to a child psychiatric outpatient clinic in the
Aims and hypotheses of this study
Northern Netherlands any time before the age of 11. The
This study aimed to determine if sex differences in devel- TRAILS cohort data were linked to the Psychiatric Case
opmental trajectories of psychopathology in persons with Register Northern Netherlands (PCRNN). The PCRNN
milder forms of autism differ from normative sex-specific registers specialist child, adolescent, and adult mental
developmental trajectories of psychopathology. To this health care consumption in the three Northern provinces of
end, we documented sex-specific developmental trajecto- the Netherlands, which overlaps with the area from which
ries of autistic and non-autistic symptoms from childhood TRAILS participants were recruited. Children and their
through young adulthood in persons with a Diagnostic and parents gave consent to link their TRAILS data to health
statistical manual of mental disorders: 4th Edition (DSM- care records in the PCRNN. The PCRNN contained data
IV-TR) ASD classification (American Psychiatric from 2000 up (4 years preceding the start of TRAILS) on
Association [APA], 2000). We compared these with nor- clinical diagnoses, and the potential presence of an autism
mative levels of these symptoms in typically developing diagnosis was identified in the PCRNN until age 19.
adolescents who were matched for IQ and socioeconomic Because only three of the 155 TRAILS CC participants
status (SES). Based on the literature, we expected higher with an ASD classification had an Autistic Disorder clas-
scores for autistic persons with all domains of psychopa- sification (DSM-IV classification 299.0), we confined our
thology, but specifically in autistic males, higher levels of study group to those participants with milder forms of
restricted and repetitive behavior than in females with ASD (DSM-IV classification 299.80: Asperger’s Disorder
ASD. This difference would be absent in the normative or Pervasive Developmental Disorder—Not Otherwise
comparison group due to low levels of these symptoms Specified).
(Kaat et al., 2021; Van Wijngaarden-Cremers et al., 2014).
Furthermore, based on the female phenotype hypothesis, Matching
more internalizing and less externalizing symptoms in
females compared to autistic males would be expected. We matched an equally sized subsample of the TRAILS
Furthermore, based on Gotham et al.’s (2015) longitudinal general population sample (normative sample) to the sam-
study, a more refined developmental expectation would be ple of autistic persons on characteristics potentially linked
that higher internalizing problems in early adolescence are to autism which may confound conclusions: that is, sex,
found in males, while females with and without autism IQ, and SES. To this end, we first ensured that we had an
increase in internalizing problems such that the female equal number of male and female participants in the clini-
phenotype hypothesis would hold particularly in late ado- cal and normative sample. IQ was a matching variable
lescence. Additional and more refined expectations on sex given its well-known association with severity of autism
differences throughout adolescence, in particular with and comorbid psychiatric problems. Furthermore, we used
regard to ADHD and externalizing symptoms, could not be SES as a matching variable, as family and social character-
formulated due to the overall absence of knowledge on istics have been implicated in influencing symptom mani-
developmental differences in autistic males and females. festation as well as clinical interpretation of such symptoms
in autistic girls (Kreiser et al., 2014), although few studies
have specifically investigated SES in this connection, and
Methods its association with symptom severity of autism is cur-
rently not clear (Mahendiran et al., 2019). Finally, since
Sample TRAILS is a homogeneous age cohort, and followed up at
The study is based on data from persons with a clinical same age-intervals over time, age matching was not a pri-
DSM-IV-TR ASD classification (n = 152; 111 males, 41 ori necessary, although we checked if the two samples
1720 Autism 27(6)

Table 1. Demographic characteristics of persons with and without ASD, comparing males and females at T1a (mean (SD)).

Males (n = 222) Females (n = 82) Total (n = 304) Difference f/m

ASD (n = 152) Age 11.00 (0.44) 11.02 (0.57) 11.01 (0.48) n.s.
IQ 101.67 (15.45) 96.16 (12.27) 100.15 (14.81) p < 0.05
SES 0.97 (0.68) −0.02 (0.62) 0.64 (0.67) n.s.
No ASD (n = 152) Age 11.23 (0.60) 11.35 (.54) 11.26 (0.48) n.s.
IQ 101.77 (14.94) 95.83 (13.71) 100.16 (14.81) p < 0.05
SES 0.91 (0.68) −0.56 (0.61) 0.52 (0.66) n.s.
Total (n = 304) Age 11.11 (0.45) 11.18 (0.58) 11.13 (0.55) n.s.
IQ 101.72 (15.16) 96.00 (12.92) 100.16 (14.79) p < 0.05
SES 0.09 (0.54) −0.38 (0.61) 0.06 (0.66) n.s.

ASD: autism spectrum disorder; IQ: intelligence quotient; SES: socioeconomic status; SD: standard deviation.
a
Sample size at different waves: T2: n = 140 (103 males, 37 females); T3: n = 110 (80 males, 30 females); T4: n = 117 (85 males, 32 females); T5: n = 115
(82 males, 33 females).

were comparable on age, after matching. We first matched autism. Furthermore, subscales (3) Fear of and resistance
the clinical and normative samples on IQ such that the IQ to change (i.e. Change) and (4) Sensory stimulation/motor
of the matching participant from the normative sample was stereotypies (i.e. Sensory) representing the restricted and
within 2 IQ points of the IQ of the autistic participant. repetitive behaviors and interest domains of autism.
Next, of those participants who met this criterion, the par- Finally, subscales (5) Orientation problems in time, place,
ticipant who had the SES closest to the autistic participant or activity (i.e. Orientation) and (6) Not optimally tuned to
was selected. After matching, mean IQ in the clinical sam- the social situation (i.e. Tuned) capture aspects of daily-
ple was 100.15 (SD 14.81), in the normative sample mean life executive functioning and self-regulation. These scales
IQ was 100.16 (SD 14.81) (Table 1), which were similar are less specific of autistic behavior, and results are put in
(p = 1.00); mean SES in the clinical sample was 0.64 (SD Supplementary Information. The CSBQ was designed to
0.67) and in the normative sample 0.52 (SD 0.66; Table 1), include both subtle forms of autistic behaviors and severe
which were similar (p = 0.87). Table 1 shows further that autistic behaviors and has a 3-point Likert-type scale to
within the clinical and normative samples, IQ was lower in also allow for quantitative change rather than presence/
females than males, but SES was comparable. absence (Hartman et al., 2006). Results are reported at
mean item level, ranging from 0 to 2. The validity of the
CSBQ has been shown in relation to autistic behavior and
Instruments
brain volumetric correlates of autism (Hartman et al.,
To assess cognitive ability, the Vocabulary and Block 2006; O’Dwyer et al., 2014). The CSBQ has been used in
Design subtests of the Revised Wechsler Intelligence various studies with a longitudinal design in which the
Scales for Children (Wechsler, 1974) were administered at course of autistic symptoms was examined (e.g. Oerlemans
baseline. IQ was estimated from the scores on these sub- et al., 2018), or as a measure to monitor the outcome of
tests. Family SES was assessed at baseline using five indi- interventions which aim to target autistic behavior (e.g.
cators: family income, educational level of the father and Pater et al., 2021).
the mother, and occupational level of both parents using Internalizing (depression and anxiety), ADHD and
the standard Classification of Occupations. We created externalizing (conduct disorder) symptoms were measured
an SES variable by averaging the indicators after with the DSM-based scales of the parent-reported Child
standardization. Behavior Checklist (CBCL; Achenbach, 2001) and the
To capture the heterogeneity of problems of children self-report Youth Self Report (YSR; Achenbach, 2001).
and adolescents with autism, we used the Children’s Social From age 18, there was a transition from the CBCL to the
Behavior Questionnaire (CSBQ; Hartman et al., 2006; Adult Behavior Checklist (ABCL; Rescorla & Achenbach,
Luteijn et al., 2000). The parent-reported, 49-item CSBQ 2004) and from the YSR to the Adult Self Report (ASR;
is a quantitative measure of autistic traits, with subscales Rescorla & Achenbach, 2004). For externalizing prob-
that allow a differentiated description of multitype autistic lems, we chose conduct disorder symptoms which have
problems. We report on the four subscales which are the been identified both for the CSBQ, YSR, ASBQ, and the
most characteristic of autism as well as two additional sub- ASR. We used self-report (i.e. YSR/ASR) as the main out-
scales: (1) Reduced contact and social interest (i.e. Social) come measure, in particular, given its indispensability for
and (2) Difficulties in understanding social information affective problems and anxiety, but put results from parent
(i.e. Understanding), representing the social domain of report in Supplementary Information. Results are reported
Horwitz et al. 1721

at mean item level, ranging from 0 to 2. The Achenbach without autism. A three-way interaction effect for Sensory
scales have been used extensively in research on internal- stimulation/motor stereotypies indicated that scores were
izing and externalizing symptoms in adolescents and have higher in males with autism than females with autism and
demonstrated good validity and reliability (Rescorla & decreased for males with age (Table 2, Figure 1(d), ST1,
Achenbach, 2004). The evidence for the validity of both and SF1d), while there were no sex differences in individ-
the syndrome scales and DSM-oriented scales of the CBCL uals without autism. The subscales Orientation problems
in individuals with ASD is strong (Magyar & Pandolfi, in time, place or activity, and Not optimally tuned to the
2017; Pandolfi et al., 2012). social situation showed no sex differences (Table 2, Figure
The CSBQ, CBCL, and YSR were administered at T1 1(e) and (f), ST1, and SF1e/f).
(mean age 11.13 years), T2 (mean age 13.12 years), and T3
(mean age 16.02 years). At T4 (mean age 19.05 years), the
Internalizing symptoms
CSBQ and ASR were administered, and at T5 (mean age
22.06 years) the ABCL. Table 3 provides model-estimated means of internaliz-
ing, ADHD, and externalizing symptoms, and Figure 2
plots the sex differences in persons with and without
Data analysis ASD. Supplementary Table ST2 provides the regression
CSBQ, CBCL, ABCL, YSR, and ASR scores were ana- model estimates, and Supplementary Figure SF2 pro-
lyzed using linear mixed modeling (Gardiner et al., 2009). vides the estimated means separately for males and
Models contained a random intercept and random slope for females with and without ASD. During adolescence, as
age and fixed effects for age at assessment, autism status, indicated by two-way sex-by-age interaction effects,
sex, and all possible interactions between these variables. self-reported affective and anxiety symptoms increased
The covariance between the random intercept and slope in females, but decreased in males (Table 3, Figure 2(a)
was freely estimated. Age at assessment was continuous and (b), ST2, and SF2a/b). There were no three-way
and the normative group and males served as reference interaction effects between diagnostic group, sex, and
categories in all models. age on either of these two internalizing subscales indi-
In a post hoc analysis, we added IQ as a covariate to cating that these sex differences were not specific for
adjust for the lower IQ difference that was found to be pre- ASD. The parent ratings of affective and anxiety symp-
sent between in autistic females compared to autistic males toms yielded similar findings (Table 3, Figure 2(e) and
(and through matching also in males and females without (f), ST2, and SF2e/f).
ASD). All models were estimated in SPSS, version 26,
using maximum likelihood estimation (West, 2009).
To enhance interpretation, we standardized all CSBQ,
ADHD symptoms
CBCL, ABCL, YSR, and ASR subdomain scores and There were no sex differences in self- or parent-reported
mean-centered age. scores of ADHD symptoms (Table 3, Figure 2(c) and (g),
There was no community involvement in the reported ST2, and SF2c/g).
study.
Externalizing symptoms
Results No sex differences were found in self-reported conduct
Table 2 provides model estimated means of autistic symp- problems as shown in Table 3, Figure 2(d), ST2, and SF2d.
toms, and Figure 1 plots the sex differences in persons The parent ratings yielded the same conclusion (Table 3,
with and without ASD. Supplementary Table ST1 provides Figure 2(h), ST2, and SF2h).
the regression model estimates and Supplementary Figure
SF1 the estimated means separately for males and females
Post hoc analyses
with and without ASD. No sex differences for the sub-
scales Reduced contact and social interest and Difficulties Adding IQ as a covariate in post hoc analyses did not alter
in understanding social information were found (Table 2, any of the conclusions drawn from the findings of our
Figure 1(a) and (b), Supplementary ST1, and SF1a/b). In main analyses (see the regression model estimates in ST3
contrast, a three-way interaction effect for the Fear of and and ST4 as compared to the estimates in ST1 and ST2,
resistance to change subscale (Table 2, Figure 1(c), ST1, respectively).
and SF1c) indicated that scores were similar for males and
females without autism but higher in autistic males com-
Discussion
pared to autistic females. These scores decreased in autis-
tic males during adolescence and increased in autistic The aim of this study was to identify possible sex-specific
females relative to the low stable scores in individuals manifestations of autism during the course of adolescence
1722 Autism 27(6)

Table 2. Model-estimated means of CSBQ domains in standardized scores of males and females with and without ASD between
childhood and young adulthood.

Wave 1 (age Wave 2 (age Wave 3 (age Wave 4 (age

~11 years) ~13 years) ~16 years) ~19 years)
CSBQ Reduced contact and social interest Males without ASD −0.57 −0.53 −0.47 −0.41
Females without ASD −0.68 −0.63 −0.56 −0.49
Males with ASD 0.45 0.46 0.48 0.51
Females with ASD 0.32 0.35 0.40 0.45
Difficulties in understanding social Males without ASD −0.50 −0.53 −0.56 −0.59
situations
Females without ASD −0.57 −0.57 −0.57 −0.57
Males with ASD 0.67 0.55 0.38 0.20
Females with ASD 0.57 0.52 0.46 0.40
Fear and resistance to change Males without ASD −0.39 −0.40 −0.40 −0.41
Females without ASD −0.55 −0.54 −0.52 −0.50
Males with ASD 0.46 0.39 0.28 0.17
Females with ASD 0.34 0.42 0.54 0.65
Sensory stimulation/motor Males without ASD −0.22 −0.28 −0.36 −0.45
stereotypes
Females without ASD −0.37 −0.40 −0.44 −0.49
Males with ASD 0.71 0.52 0.24 −0.04
Females with ASD 0.07 0.06 0.04 0.02
Not optimally tuned into the social Males without ASD −0.41 −0.46 −0.54 −0.62
situation
Females without ASD −0.41 −0.43 −0.47 −0.50
Males with ASD 0.70 0.54 0.31 0.08
Females with ASD 0.63 0.52 0.34 0.17
Orientation problems in time, place, Males without ASD −0.42 −0.44 −0.47 −0.50
or activity
Females without ASD −0.59 −0.62 −0.65 −0.69
Males with ASD 0.68 0.57 0.40 0.23
Females with ASD 0.51 0.40 0.23 0.06

CSBQ: Children’s Social Behavior Questionnaire; ASD: autism spectrum disorder.

and determine if these differed from normative sex-spe- results confirm that found developmental sex differences
cific developmental trajectories of psychopathology. for anxiety and affective problems, even though they tend
Sex differences specific for autism were only found in to be higher in (both men and) women with than without
the autism domain of restrictive repetitive behavior (Fear autism, are not autism specific.
of and resistance to change and Sensory stimulation / In recent years, the notion of a specific female autistic
motor stereotypies), such that unlike persons without symptom profile has been developed in the literature (e.g.
autism, autistic males had higher scores than autistic Hull et al., 2020; Mandy et al., 2012) which has received
females. This difference was stable over time for sensory much attention both in the scientific and in the lay press.
stimulation/motor stereotypies, but disappeared during the Although the first delineations of this female phenotype
course of adolescence for resistance to change, due to an assumed subtle sex differences in behavior (more restricted
increase of these problems in autistic females. No sex dif- repetitive behavior in males, more internalizing problems
ferences were found on the CSBQ social-communicative in females, more externalizing problems in males, superior
scale scores. With regard to broader-than-autism psycho- motor skills in females (Mandy et al., 2012), more recent
pathology, sex differences were alike for the autism and publications presume a broader set of typical female autis-
the normative groups: a decrease over time in parent- and tic characteristics, also including social-communicative
self-reported affective and anxiety symptoms in males, behavior (Hull et al., 2020). Our results clearly support the
while in females, these symptoms increased over time. former, more limited female autistic symptom profile: sex
Such similarities in course of symptoms in autism and differences in severity and course of social-communica-
non-autism participants have also been reported in previ- tive behavior were not confirmed in our present analyses.
ous studies that included individuals with non-spectrum Our results are in line with previous findings that autistic
developmental delays (Gotham et al., 2015). Thus, our males demonstrate more restricted repetitive behavior than
Horwitz et al. 1723

(a) CSBQ: reduced contact and social interest (b) CSBQ: difficules in understanding social situaons
0.00 0.30

-0.02 0.25
Standard deviation difference

Standard deviaon difference

0.20
-0.04
0.15
-0.06 0.10
-0.08 0.05

-0.10 0.00
-0.05
-0.12
-0.10
-0.14 Persons without ASD Persons without ASD
Persons with ASD -0.15 Persons with ASD
-0.16 -0.20
10 11 12 13 14 15 16 17 18 19 20 10 11 12 13 14 15 16 17 18 19 20
Age Age
(c) CSBQ: fear and resistance to change (d) CSBQ: sensory smulaon/motor stereotypies
0.60 0.20
0.50 0.10
Standard deviation difference

Standard deviaon difference

0.40 0.00
-0.10
0.30
-0.20
0.20
Persons without ASD -0.30
0.10 Persons with ASD -0.40
0.00
-0.50
-0.10 -0.60
Persons without ASD
-0.20 -0.70 Persons with ASD
-0.30 -0.80
10 11 12 13 14 15 16 17 18 19 20 10 11 12 13 14 15 16 17 18 19 20
Age Age
(e) CSBQ: not opmally tuned into the social situaon (f) CSBQ: orientaon problems in me, place, or acvity
0.20 -0.13
Standard deviation difference

Standard deviaon difference

0.15 -0.15

0.10 -0.17

0.05 -0.19

0.00 -0.21

-0.05 -0.23 Persons without ASD

Persons without ASD
Persons with ASD
Persons with ASD
-0.10 -0.25
10 11 12 13 14 15 16 17 18 19 20 10 11 12 13 14 15 16 17 18 19 20
Age Age

Figure 1. Model-estimated CSBQ female versus male differences in standard deviations using linear mixed modeling between
childhood and young adulthood in persons with and without ASD.
CSBQ: Children’s Social Behavior Questionnaire; ASD: autism spectrum disorder.

autistic females in adolescence. In addition, we found a both in the autism and the normative samples, while no sex
sex-specific course through adolescence of the Fear of and differences in conduct problems and ADHD symptoms
resistance to change symptoms, which increased in were identified. In sum, the sex-specific autistic symptom
females through adolescence such that the initially higher profile that emerges from our study is confined to higher
scores in males disappeared. So far, no research has been restricted repetitive behavior in men and higher resistance
conducted on the course of restricted behavior in this age to change in early adolescent boys—the latter restricted to
group in autism in a longitudinal design. Thus, we extended early adolescence due to subsequent increase of these
previous findings. With respect to the sex differences in symptoms in girls. Broader anxiety and affective symp-
internalizing and externalizing symptoms reported by toms showed the well-known developmental increase in
Mandy et al. (2012) and others, we found that a sex-spe- girls during adolescence, but this was not confined to indi-
cific course of anxiety and affective symptoms was present viduals with autism (Fernandez Castelao & Kröner-Herwig
1724 Autism 27(6)

Table 3. Model-estimated means of YSR and CBCL domains in standardized scores of males and females with and without ASD
between childhood and young adulthood.

Wave 1 (age Wave 2 (age Wave 3 (age Wave 4/5 (age

~11 years) ~13 years) ~16 years) ~19/22 years)
YSR/ASR Affective problems Males without ASD −0.155 −0.241 −0.37 −0.499
Females without ASD −0.273 −0.193 −0.073 0.047
Males with ASD 0.184 0.104 −0.016 −0.136
Females with ASD 0.452 0.556 0.712 0.868
Anxiety problems Males without ASD −0.222 −0.264 −0.327 −0.39
Females without ASD −0.175 −0.073 0.08 0.233
Males with ASD 0.064 0.006 −0.081 −0.168
Females with ASD 0.578 0.656 0.773 0.89
ADHD problems Males without ASD −0.171 −0.203 −0.251 −0.299
Females without ASD −0.055 −0.121 −0.22 −0.319
Males with ASD 0.349 0.259 0.124 −0.011
Females with ASD 0.376 0.26 0.086 −0.088
Conduct problems Males without ASD 0.187 0.083 −0.073 −0.229
Females without ASD −0.094 −0.198 −0.354 −0.51
Males with ASD 0.32 0.218 0.065 −0.088
Females with ASD 0.033 −0.037 −0.142 −0.247
CBCL/ACBL Affective problems Males without ASD −0.466 −0.468 −0.471 −0.477
Females without ASD −0.636 −0.544 −0.406 −0.13
Males with ASD 0.371 0.357 0.336 0.294
Females with ASD 0.231 0.355 0.541 0.913
Anxiety problems Males without ASD −0.453 −0.443 −0.428 −0.398
Females without ASD −0.544 −0.446 −0.299 −0.005
Males with ASD 0.386 0.318 0.216 0.012
Females with ASD 0.481 0.521 0.581 0.701
ADHD problems Males without ASD −0.214 −0.31 −0.454 −0.742
Females without ASD −0.471 −0.525 −0.606 −0.768
Males with ASD 0.794 0.624 0.369 −0.141
Females with ASD 0.582 0.434 0.212 −0.232
Conduct problems Males without ASD −0.293 −0.285 −0.273 −0.249
Females without ASD −0.466 −0.426 −0.366 −0.246
Males with ASD 0.382 0.372 0.357 0.327
Females with ASD 0.067 0.069 0.072 0.078

The YSR/ASR were assessed at waves 1–4 and the CBCL/ACBL at waves 1, 2, 3, and 5. YSR: youth self report; ASR: adult self report; CBCL: Child
Behavior Checklist; ABCL: Adult Behavior Checklist; ADHD: attention-deficit/hyperactivity disorder.

2014; Hankin et al., 2015; Ohannessian et al., 2017). Sex- females (Supplementary Table S3–S4)). Note that in our
specific patterns of externalizing problems were not sample with an average IQ of around 100, the diagnostic
identified. barriers for autistic girls such as being more successful at
Males had a somewhat higher IQ than females in our camouflaging their problems as suggested in the literature
autism group (and therefore also in our matched normative (Hull et al., 2020) may have led to underdiagnosis of
group) which was not accounted for in our main analysis. female participants. As a consequence, we cannot fully
The difference in IQ identified in our samples is in line rule out that when more women with a higher IQ had been
with the literature showing that autistic females are over- present in our autism group, the results regarding sex dif-
represented at the lower end of IQ and underrepresented at ferences would have been more in line with findings in
the higher end (Kaat et al., 2021). This may have likewise other studies that suggested a more pronounced female
played a role in previous studies on sex differences in autism phenotype. That said, the composition of our autism
autism that have for the most part not studied potential group and our findings give a good representation of the
confounding by IQ (Mahendiran et al., 2019). To further sex differences we encounter in clinical practice, including
explore this, we performed post hoc analysis, and showed females with lower IQ.
that the results regarding sex differences did not differ The parallel increase in Fear and resistance to change
when adjusted for the IQ differences between males and scores and anxiety scores found in autistic girls leads to the
Horwitz et al. 1725

(a) YRS: affecve problems (b) YSR: anxiety problems

1.20 1.20

1.00 1.00
Standard deviation difference

Standard deviaon difference

0.80
0.80
0.60
0.60
0.40
0.40
0.20
0.20
0.00

-0.20 Persons without ASD 0.00 Persons without ASD

Persons with ASD Persons with ASD
-0.40 -0.20
10 11 12 13 14 15 16 17 18 19 20 10 11 12 13 14 15 16 17 18 19 20
Age Age
(c) YSR: ADHD problems (d) YSR: conduct problems
0.15 0.00

-0.05
Standard deviation difference

Standard deviaon difference

0.10
-0.10
0.05
-0.15
0.00
-0.20
-0.05
-0.25
-0.10 Persons without ASD -0.30 Persons without ASD
Persons with ASD Persons with ASD
-0.15 -0.35
10 11 12 13 14 15 16 17 18 19 20 10 11 12 13 14 15 16 17 18 19 20
Age Age
(e) CBCL: affecve problems (f) CBCL: anxiety problems
1.00 1.00

0.80
Standard deviation difference

Standard deviaon difference

0.80
0.60
0.60
0.40
0.40
0.20
0.20
0.00

-0.20 Persons without ASD 0.00 Persons without ASD

Persons with ASD Persons with ASD
-0.40 -0.20
10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
Age Age
(g) CBCL: ADHD problems (h) CBCL: conduct problems
0.05 0.10
0.05
0.00
Standard deviation difference

Standard deviaon difference

0.00
-0.05
-0.05
-0.10 -0.10

-0.15 -0.15
-0.20
-0.20
-0.25
-0.25 Persons without ASD
-0.30 Persons without ASD
Persons with ASD
Persons with ASD
-0.30 -0.35
10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
Age Age

Figure 2. Model-estimated YSR and CBCL female versus male differences in standard deviations using linear mixed modeling
between childhood and young adulthood in persons with and without ASD.
The YSR/ASR were assessed at waves 1–4 and the CBCL/ACBL at waves 1, 2, 3, and 5. YSR: youth self report; ASR: adult self report; CBCL: Child
Behavior Checklist; ABCL: Adult Behavior Checklist; ADHD: attention-deficit/hyperactivity disorder.
1726 Autism 27(6)

question if, in autism (i.e. unlike anxiety, fear, and resist- that a female autism group with a more “masculine” autism
ance to change did not increase in females without autism), presentation was included. However, as mentioned, par-
these symptom domains are related. Insistence on same- ticipants with autism were on the milder part of the autism
ness has been suggested to function as a coping mecha- spectrum in our sample, which diminishes chances of hav-
nism for anxiety or distress. Lidstone et al. (2014) showed ing missed substantial numbers of autistic females. In gen-
that, in 2- to 17-year olds with autism, insistence on same- eral, it holds that only if diagnostic criteria—that are
ness was associated with anxiety, this relation being medi- characteristic for both sexes—are agreed upon and assess-
ated by sensory sensitivity and sensory avoiding. Baribeau ment instruments are changed accordingly this situation of
and colleagues (2020) studied longitudinal trajectories of potential underdiagnosis in autistic females can be
insistence on sameness and anxiety in 421 children with improved.
autism at ages 3, 6, and 11 years and found that these tra- We conclude that in adolescents with milder forms of
jectories were similar in severity and direction for most of autism and an average IQ subtle sex differences are found
the participants. The group with trajectories with the high- that are not present in the normative sample: higher scores
est insistence on sameness and anxiety scores had propor- on the Sensory stimulation/motor stereotypes scale in
tionally more girls than the other groups. In line with these males than in females throughout adolescence, and an
findings, our results also suggest a sex-specific association increase over time in scores on the Fear of and resistance
between Fear and resistance to change symptoms and to change scales in females such that the higher scores of
anxiety in autism, thus extending the findings of Baribeau males in early adolescence were similarly high in males
and colleagues in childhood to adolescence. Further and females with autism at the start young adulthood.
research on this topic should clarify the strength of the Furthermore, with regard to broader-than-autism psycho-
relationship between the restrictive repetitive domain of pathology, a decrease over time in affective and anxiety
autism and specific subtypes of anxiety (for instance sepa- symptoms in males was found, while in females, these
ration anxiety; see Black et al., 2017). Girls with autism in symptoms increased over time. While this sex-specific
late adolescence may particularly benefit from interven- developmental pattern held for both individuals with and
tions focused on anxiety, intolerance of uncertainty and without autism, increasing anxiety and affective problems
sensory hypersensitivity (Jenkinson et al., 2020; Wigham during adolescence are nonetheless an area of concern in
et al., 2015). the clinical care of autistic women. Note, however, that sex
An important limitation of this study is that the initial differences between men and women with autism are more
autism diagnosis was made clinically and assessment was subtle than the differences between individuals with and
not systematically confirmed by us at inclusion in the without autism, which were broadly present for both the
study using standardized diagnostic measures. The quality internalizing and externalizing problem domains.
of diagnostic information on autism in TRAILS depends
on clinical diagnostic practice, which in the Netherlands Funding
accords with the national clinical guideline stating that a The author(s) received no financial support for the research,
best-estimate clinical diagnosis should be based on inte- authorship, and/or publication of this article.
grated information from different sources such as anamne-
sis, hetero-anamnesis, and observation (Kan et al., 2013). ORCID iD
The around-to-above threshold-levels of SCQ scores for Ernst Horwitz https://orcid.org/0000-0002-9204-1162
autism at the time of inclusion in the study at age 11
reported in our previous study appears to confirm the pres- Supplemental material
ence of significant autistic behavior in the autism group,
although almost all received a clinical diagnosis of PDD- Supplemental material for this article is available online.
NOS (Horwitz et al., 2020). Nonetheless, in particular in
light of the current higher threshold for diagnosing autism References
in DSM-5, it is likely that autism may have been diagnosed Achenbach, L. A. (2001). Manual for the ASEBA school-age
more often in our sample than if DSM-5 criteria would forms & profiles. Research Center for Children, Youth and
have been the standard. Families, University of Vermont.
Another limitation relates to the possibility of a diag- American Psychiatric Association. (2000). Diagnostic and statis-
tical manual of mental disorders (4th ed.).
nostic bias against girls, particularly among those without
Bargiela, S., Steward, R., & Mandy, W. (2016). The experiences
intellectual disability who score high on autistic symptoms of late-diagnosed women with autism spectrum conditions:
but fail to meet the diagnostic criteria for autism (Loomes An investigation of the female autism phenotype. Journal of
et al., 2017; Ratto et al., 2018). According to the female Autism and Developmental Disorders, 46(10), 3281–3294.
autistic phenotype/camouflaging hypothesis, girls with a Baribeau, D. A., Vigod, S., Pullenayegum, E., Kerns, C. M.,
female-specific autism presentation are likely to be missed Mirenda, P., Smith, I. M., Valliancourt, T., Volden, J.,
out on a timely diagnosis. In our study, we cannot rule out Waddell, C., Zwaigenbaum, L., Bennett, T., Duku, E.,
Horwitz et al. 1727

Elsabbagh, M., Georgiades, S., Ungar, W. J., Zait, A. Z., crucial time window: Illustrating the need to go beyond
& Szatmari, P. (2020). Co-occurring trajectories of anxiety childhood. Neuroscience and Biobehavioralreviews, 71,
and insistence on sameness behaviour in autism spectrum 529–541.
disorder. The British Journal of Psychiatry: The Journal of Hartman, C. A., Luteijn, E., Serra, M., & Minderaa, R. (2006).
Mental Science, 218, 20–27. Refinement of the Children’s Social Behavior Questionnaire
Black, K. R., Stevenson, R. A., Segers, M., Ncube, B. L., Sun, (CSBQ): An instrument that describes the diverse prob-
S. Z., Philipp-Muller, A., Bebko, J. M., Barense, M. D., & lems seen in milder forms of PDD. Journal of Autism and
Ferber, S. (2017). Linking anxiety and insistence on same- Developmental Disorders, 36(3), 325–342.
ness in autistic children: The role of sensory hypersensitiv- Holtmann, M., Bölte, S., & Poustka, F. (2007). Autism spectrum
ity. Journal of Autism and Developmental Disorders, 47(8), disorders: Sex differences in autistic behaviour domains and
2459–2470. coexisting psychopathology. Developmental Medicine and
Bongers, I. L., Koot, H. M., van der Ende, J., & Verhulst, F. C. Child Neurology, 49(5), 361–366.
(2004). Developmental trajectories of externalizing behav- Horwitz, E. H., Schoevers, R. A., Greaves-Lord, K., de Bildt,
iors in childhood and adolescence. Child Development, A., & Hartman, C. A. (2020). Adult manifestation of milder
75(5), 1523–1537. forms of autism spectrum disorder; autistic and non-autistic
Boyd, A., Van de Velde, S., Vilagut, G., de Graaf, O’Neill, S., psychopathology. Journal of Autism and Developmental
Florescu, S., Alonso, J., & Kovess-Masfety, V. EU-WMH Disorders, 50(8), 2973–2986.
Investigators. (2015). Gender differences in mental disor- Huang, C. L., Weng, S., & Ho, C. (2016). Gender ratios of
ders and suicidality in Europe: Results from a large cross- administrative prevalence and incidence of attention-def-
sectional population-based study. Journal of Affective icit/hyperactivity disorder (ADHD) across the lifespan: A
Disorders, 173, 245–254. nationwide population-based study in taiwan. Psychiatry
Demetriou, E. A., Pepper, K. L., Park, S. H., Pellicano, L., Research, 244, 382–387.
Song, Y. J. C., Naismith, S. L., Hickie, I. B., Thomas, E. Huisman, M., Oldehinkel, A. J., de Winter, A., Minderaa, R. B.,
E., & Guastella, A. J. (2021). Autism spectrum disorder: de Bildt, A., Huizink, A. C., Verhust, F. C., & Ormel, J.
An examination of sex differences in neuropsychological (2008). Cohort profile: The Dutch “TRacking adolescents”
and self-report measures of executive and non-executive individual lives’ survey’; TRAILS. International Journal of
cognitive function. Autism: The International Journal of Epidemiology, 37(6), 1227–1235.
Research and Practice, 25(8), 2223–2237. Hull, L., Petrides, K. V., & Mandy, W. (2020). The female autism
Fernandez Castelao, C., & Kröner-Herwig, B. (2014). phenotype and camouflaging: A narrative review. Journal of
Developmental trajectories and predictors of externalizing Autism and Developmental Disorders, 7(4), 306–317.
behavior: A comparison of girls and boys. Journal of Youth Jenkinson, R., Milne, E., & Thompson, A. (2020). The rela-
and Adolescence, 43(5), 775–789. tionship between intolerance of uncertainty and anxiety
Fombonne, E. (2020). Camouflage and autism. Journal of Child in autism: A systematic literature review and meta-anal-
Psychology & Psychiatry, 61(7), 735–738. ysis. Autism: The International Journal of Research and
Fountain, C., Winter, A. S., & Bearman, P. S. (2012). Six devel- Practice, 24(8), 1933–1944.
opmental trajectories characterize children with autism. Kaat, A. J., Shui, A. M., Ghods, S. S., Farmer, C. A., Esler, A.
American Academy of Pediatrics. N., Thurm, A., Georgiades, S., Kanne, S. M., Lord, C., Kim,
Gardiner, J. C., Luo, Z., & Roman, L. A. (2009). Fixed effects, Y. S., & Bishop, S. L. (2021). Sex differences in scores on
random effects and GEE: What are the differences? Statistics standardized measures of autism symptoms: A multisite
in Medicine, 28(2), 221–239. integrative data analysis. Journal of Child Psychology &
Gotham, K., Brunwasser, S. M., & Lord, C. (2015). Depressive Psychiatry, 62(1), 97–106.
and anxiety symptom trajectories from school age through Kan, C., Geurts, H., Bosch, K., van den Forceville, E., van
young adulthood in samples with autism spectrum disorder Manen, J., van den Bosch, A., Hermens, M., Peters, A.,
and developmental delay. Journal of the American Academy Majo, M. C., Shields, L., Zon, N., & van Duin, D. (Eds.).
of Child and Adolescent Psychiatry, 54(5), 369.e–376.e3. (2013). Dutch guideline for the assessment and treatment of
Hankin, B. L., Abramson, L. Y., Moffitt, T. E., Silva, P. A., autism spectrum disorder in adults. De Tijdstroom.
McGee, R., & Angell, K. E. (1998). Development of depres- Karriker-Jaffe, K., Foshee, V. A., Ennett, S. T., & Suchindran, C.
sion from preadolescence to young adulthood: Emerging (2008). The development of aggression during adolescence:
gender differences in a 10-year longitudinal study. Journal Sex differences in trajectories of physical and social aggres-
of Abnormal Psychology, 107(1), 128–140. sion among youth in rural areas. Journal of Abnormal Child
Hankin, B. L., Young, J. F., Abela, J. R. Z., Smolen, A., Jenness, Psychology, 36(8), 1227–1236.
J. L., Gulley, L. D., Technow, J. R., Gottieb, A. B., Cohen, Kim, J. A., Szatmari, P., Bryson, S. E., Streiner, D. L., & Wilson,
J. R., & Oppenheimer, C. W. (2015). Depression from F. J. (2000). The prevalence of anxiety and mood problems
childhood into late adolescence: Influence of gender, devel- among children with autism and asperger syndrome. Autism:
opment, genetic susceptibility, and peer stress. Journal of The International Journal of Research and Practice, 4(2),
Abnormal Psychology, 124(4), 803–816. 117–132.
Hartman, C. A., Geurts, H. M., Franke, B., Buitelaar, J. K., & Kreiser, N., & White, S. (2014). ASD in females: Are we over-
Rommelse, N. N. J. (2016). Changing ASD-ADHD symp- stating the gender difference in diagnosis? Clinical Child &
tom co-occurrence across the lifespan with adolescence as Family Psychology Review, 17(1), 67–84.
1728 Autism 27(6)

Lai, M., Hull, L., Mandy, W., Chakrabarti, B., Nordahl, C. W., disorder: Evidence from a large sample of children and ado-
Lombardo, M. V., Amies, S. H., Szatmari, P., Baron-Cohen, lescents. Journal of Autism and Developmental Disorders,
S., & Happe, F. (2020). Commentary: “camouflaging” in 42(7), 1304–1313.
autistic people—reflection on fombonne (2020). Journal of Mandy, W., Pellicano, L., St Pourcain, B., Skuse, D., & Heron,
Child Psychology and Psychiatry, 62, Article 13926. J. (2018). The development of autistic social traits across
Lai, M., Lombardo, M. V., Ruigrok, A. N., Chakrabarti, B., childhood and adolescence in males and females. Journal
Auyeung, B., Szatmari, P., Happe, F., Baron-Cohen, S., & of Child Psychology and Psychiatry, and Allied Disciplines,
MRC AIMS Consortium. (2017). Quantifying and explor- 59(11), 1143–1151.
ing camouflaging in men and women with autism. Autism: May, T., Adesina, I., McGillivray, J., & Rinehart, N. J. (2019).
The International Journal of Research and Practice, 21(6), Sex differences in neurodevelopmental disorders. Current
690–702. Opinion in Neurology, 32(4), 622–626.
Legerstee, J. S., Verhulst, F. C., Robbers, S. C. C., Ormel, J., Murray, A. L., Booth, T., Eisner, M., Auyeung, B., Murray, G.,
Oldehinkel, A. J., & van Oort, F. V. A. (2013). Gender— & Ribeaud, D. (2019). Sex differences in ADHD trajec-
specific developmental trajectories of anxiety during ado- tories across childhood and adolescence. Developmental
lescence: Determinants and outcomes—The TRAILS study. Science, 22(1), Article e12721.
Canadian Academy of Child and Adolescent Psychiatry Nasca, B. C., Lopata, C., Donnelly, J. P., Rodgers, J. D., &
/ Académie canadienne de psychiatrie de l’enfant et de Thomeer, M. L. (2019). Sex differences in externalizing and
l’adolescent. internalizing symptoms of children with ASD. Journal of
Lewinsohn, P. M., Gotlib, I. H., Lewinsohn, M., Seeley, J. R., & Autism and Developmental Disorders, 50, 3245–3252.
Allen, N. B. (1998). Gender differences in anxiety disorders O’Dwyer, L., Tanner, C., van Dongen, E. V., Greven, C. U.,
and anxiety symptoms in adolescents. Journal of Abnormal Bralten, J., Zwiers, M. P., Franke, B., Oosterlaan, J.,
Psychology, 107(1), 109–117. Hesenfeld, D., Hoekstra, P., Hartman, C. A., Rommelse,
Lidstone, J., Uljarevic, M., Sulliva, J., Rodgers, J., McConachie, N., & Buitelaar, J. K. (2014). Brain volumetric correlates
H., Freeston, M., Le Couteur, A., Prior, M., & Leekam, S. of autism spectrum disorder symptoms in attention deficit/
(2014). Relations among restricted and repetitive behaviors, hyperactivity disorder. PLOS ONE, 9(6), 1–13.
anxiety and sensory features in children with autism spec- Oerlemans, A. M., Rommelse, N. N. J., Buitelaar, J. K., &
trum disorders. Research in Autism Spectrum Disorders, Hartman, C. A. (2018). Examining the intertwined devel-
8(2), 82–92. opment of prosocial skills and ASD symptoms in adoles-
Loomes, R., Hull, L., & Mandy, W. P. L. (2017). What is the male- cence. European Child & Adolescent Psychiatry, 27(8),
to-female ratio in autism spectrum disorder? A systematic 1033–1046.
review and meta-analysis. Journal of the American Academy Ohannessian, C. M., Milan, S., & Vannucci, A. (2017). Gender
of Child and Adolescent Psychiatry, 56(6), 466–474. differences in anxiety trajectories from middle to late ado-
Louwerse, A., Eussen, M. L. J. M., Van, d. E., de Nijs, P. F. lescence. Journal of Youth and Adolescence, 46(4), 826–
A., Van Gool, A. R., Dekker, L. P., Verheij,C., Verheij, F., 839.
Verhulst, F. C., & Greaves-Lord, K. (2015). ASD symptom Oldehinkel, A. J., Rosmalen, J. G., Buitelaar, J. K., Hoek, H.
severity in adolescence of individuals diagnosed with PDD- W., Ormel, J., Raven, D., Reijneveld, V. A., Veenstra,
NOS in childhood: Stability and the relation with psychiat- R., Verhulst, F. C., Vollebergh, W. M., & Hartman, C. A.
ric comorbidity and societal participation. Journal of Autism (2015). Cohort profile update: The TRacking adolescents’
and Developmental Disorders, 45(12), 3908–3918. individual lives survey (TRAILS). International Journal of
Luteijn, E., Luteijn, F., Jackson, S., Volkmar, F., & Minderaa, Epidemiology, 44(1), 76–76.
R. (2000). The children’s social behavior questionnaire for Oswald, T. M., Winter-Messiers, M., Gibson, B., Schmidt, A.
milder variants of PDD problems: Evaluation of the psycho- M., Herr, C. M., & Solomon, M. (2016). Sex differences
metric characteristics. Journal of Autism and Developmental in internalizing problems during adolescence in autism
Disorders, 30(4), 317–330. spectrum disorder. Journal of Autism and Developmental
Magyar, C. I., & Pandolfi, V. (2017). Utility of the CBCL DSM- Disorders, 46(2), 624–636.
oriented scales in assessing emotional disorders in youth Pandolfi, V., Magyar, C. I., & Dill, C. A. (2012). An initial
with autism. Research in Autism Spectrum Disorders, 37, psychometric evaluation of the CBCL 6-18 in a sample of
11–20. youth with autism spectrum disorders. Research in Autism
Mahendiran, T., Brian, J., Dupuis, A., Muhe, N., Wong, P., Spectrum Disorders, 6(1), 96–108.
Iaboni, A., & Anagnostou, E. (2019). Meta-analysis of sex Pater, M., Spreen, M., & van Yperen, T. (2021). The develop-
differences in social and communication function in chil- mental progress in social behavior of children with autism
dren with autism spectrum disorder and attention-deficit/ spectrum disorder getting music therapy. A multiple case
hyperactivity disorder. Frontiers in Psychiatry, 10, Article study. Children & Youth Services Review, 120, 1057–1067.
804. Petersen, I. T., Lindhiem, O., LeBeau, B., Bates, J. E., Pettit, G.
Malone, P. S., Van Eck, K., Flory, K., & Lamis, D. A. (2010). S., Lansford, J. E., & Dodge, K. A. (2018). Development of
A mixture-model approach to linking ADHD to adolescent internalizing problems from adolescence to emerging adult-
onset of illicit drug use. Developmental Psychology, 46(6), hood: Accounting for heterotypic continuity with vertical
1543–1555. scaling. Developmental Psychology, 54(3), 586–599.
Mandy, W., Chilvers, R., Chowdhury, U., Salter, G., Seigal, A., Pisula, E., Pudło, M., Słowińska, M., Kawa, R., Strząska, M.,
& Skuse, D. (2012). Sex differences in autism spectrum Banasiak, A., & Wolańczyk, T. (2017). Behavioral and
Horwitz et al. 1729

emotional problems in high-functioning girls and boys with symptoms during adolescence: The TRAILS study. Journal
autism spectrum disorders: Parents’ reports and adoles- of Child Psychology & Psychiatry, 50(10), 1209–1217.
cents’ self-reports. Autism : The International Journal of Van Wijngaarden-Cremers, P., Eeten, E., Groen, W., Deurzen,
Research and Practice, 21(6), 738–748. P., Oosterling, I., & Gaag, R. (2014). Gender and age dif-
Ratto, A. B., Kenworthy, L., Yerys, B. E., Bascom, J., ferences in the core triad of impairments in autism spectrum
Wieckowski, A. T., White, S. W., Wallace, G. L., Pugliese, disorders: A systematic review and meta-analysis. Springer.
C., Schultz, R. T., Ollendick, T. H., Scarpa, A., Seese, S., Verheij, C., Louwerse, A., Ende, J., Eussen, M., Gool, A.,
Register-Brown, K., Martin, A., & Anthony, L. G. (2018). Verheij, F., Verhulst, F. C., & Greaves-Lord, K. (2015). The
What about the girls? sex-based differences in autistic traits stability of comorbid psychiatric disorders: A 7 year follow
and adaptive skills. Journal of Autism and Developmental up of children with pervasive developmental disorder-not
Disorders, 48(5), 1698–1711. otherwise specified. Journal of Autism & Developmental
Rescorla, L. A., & Achenbach, T. M. (2004). The achenbach sys- Disorders, 45(12), 3939–3948.
tem of empirically based assessment (ASEBA) for ages 18 Volkmar, F. R., & Szatmari, P. (1993). Sex differences in per-
to 90 years. In M. E. Maruish (Ed.), The use of psychologi- vasive developmental disorders. Journal of Autism &
cal testing for treatment planning and outcomes assessment Developmental Disorders, 23(4), 579–591.
(pp. 115–152). Lawrence Erlbaum. Vos, M., Rommelse, N. N. J., Franke, B., Oosterlaan, J.,
Rivet, T., & Matson, J. (2011). Gender differences in core symp- Heslenfeld, D. J., Hoekstra, P. J., Klein, H., Faraone, S. V.,
tomatology in autism spectrum disorders across the lifes- Buitlaar, J. K., & Hartman, C. A. (2021). Characterizing
pan. Journal of Developmental & Physical Disabilities, the heterogeneous course of inattention and hyperactivity-
23(5), 399–420. impulsivity from childhood to young adulthood. European
Robinson, E. B., Munir, K., Munafò, M. R., Hughes, M., Child & Adolescent Psychiatry, 31, 1–11.
McCormick, M. C., & Koenen, K. C. (2011). Stability of Wechsler, D. (1974). Wechsler’s Intelligence Scale for Children
autistic traits in the general population: Further evidence for a Revised. Psychological Cooperation.
continuum of impairment. Journal of the American Academy West, B. T. (2009). Analyzing longitudinal data with the linear
of Child and Adolescent Psychiatry, 50(4), 376–384. mixed models procedure in SPSS. Evaluation & the Health
Roskam, I. (2019). Externalizing behavior from early childhood Professions, 32(3), 207–228.
to adolescence: Prediction from inhibition, language, par- Wigham, S., Rodgers, J., South, M., McConachie, H., &
enting, and attachment. Development and Psychopathology, Freeston, M. (2015). The interplay between sensory pro-
31(2), 587–599. cessing abnormalities, intolerance of uncertainty, anxiety
Salk, R. H., Petersen, J. L., Abramson, L. Y., & Hyde, J. S. (2016). and restricted and repetitive behaviours in autism spectrum
The contemporary face of gender differences and similari- disorder. Journal of Autism and Developmental Disorders,
ties in depression throughout adolescence: Development 45(4), 943–952.
and chronicity. Journal of Affective Disorders, 205, 28–35. Wood-Downie, H., Wong, B., Kovshoff, H., Cortese, S., &
Simonoff, E., Kent, R., Stringer, D., Lord, C., Briskman, J., Hadwin, J. A. (2021). Research review: A systematic
Lukito, S., Pickles, A., Charman, T., & Baird, G. (2019). review and meta-analysis of sex/gender differences in social
Trajectories in symptoms of autism and cognitive ability in interaction and communication in autistic and nonautistic
autism from childhood to adult life: Findings from a lon- children and adolescents. Journal of Child Psychology and
gitudinal epidemiological cohort. Journal of the American Psychiatry, 62(8), 922–936.
Academy of Child and Adolescent Psychiatry, 59, 1342– Woodman, A. C., Mailick, M. R., & Greenberg, J. S. (2016).
1352. Trajectories of internalizing and externalizing symptoms
Van Oort, F. V. A., Greaves-Lord, K., Verhulst, F. C., Ormel, J., & among adults with autism spectrum disorders. Development
Huizink, A. C. (2009). The developmental course of anxiety and Psychopathology, 28(2), 565–581.