Factors influencing the probability of a diagnosis of autism spectrum disorder in girls versus boys

Abstract

In order to shed more light on why referred girls are less likely to be diagnosed with autism spectrum disorder than boys, this study examined whether behavioral characteristics influence the probability of an autism spectrum disorder diagnosis differently in girls versus boys derived from a multicenter sample of consecutively referred children aged 2.5–10 years. Based on information from the short version of the Developmental, Dimensional and Diagnostic Interview and the Autism Diagnostic Observation Schedule, 130 children (106 boys and 24 girls) received a diagnosis of autism spectrum disorder according to Diagnostic and Statistical Manual of Mental Disorders (4th ed., text rev.) criteria and 101 children (61 boys and 40 girls) did not. Higher overall levels of parent-reported repetitive and restricted behavior symptoms were less predictive of an autism spectrum disorder diagnosis in girls than in boys (odds ratio interaction = 0.41, 95% confidence interval = 0.18–0.92, p = 0.03). In contrast, higher overall levels of parent-reported emotional and behavioral problems increased the probability of an autism spectrum disorder diagnosis more in girls than in boys (odds ratio interaction = 2.44, 95% confidence interval = 1.13–5.29, p = 0.02). No differences were found between girls and boys in the prediction of an autism spectrum disorder diagnosis by overall autistic impairment, sensory symptoms, and cognitive functioning. These findings provide insight into possible explanations for the assumed underidentification of autism spectrum disorder in girls in the clinic.

Keywords

autism spectrum disorder diagnosis girls sex/gender differences

Introduction

Boys are more likely to be diagnosed with autism spectrum disorder (ASD) than girls, but reasons for this discrepancy remain unclear. It has been estimated that boys are four times more likely to be diagnosed with ASD than girls, with estimates rising to 6–8:1 in samples with an average IQ or higher (Fombonne, 2003, 2005). In contrast, recent population studies suggest a lower male-to-female ratio in the range of 2–3:1 (Lai et al., 2015). This has raised the question whether females with ASD are underidentified in clinical samples, contributing to exaggerated male-to-female ratios (Kreiser and White, 2014; Lai et al., 2015), in addition to a real discrepancy in the occurrence of ASD between boys and girls possibly due to a female protective effect (e.g. Jacquemont et al., 2014; Robinson et al., 2013). Similar findings have also been reported for attention-deficit/hyperactivity disorder (ADHD; Taylor et al., 2016; Willcutt, 2012). The underidentification hypothesis for ASD is supported by findings that girls are less likely to be diagnosed with ASD than boys despite demonstrating similar levels of autistic symptoms (Dworzynski et al., 2012; Russell et al., 2011). Furthermore, there is evidence that girls with ASD are diagnosed later than boys (Begeer et al., 2013; Giarelli et al., 2010), suggesting that some girls with ASD are missed at an early age using current diagnostic practices. Therefore, a better understanding of how gender influences the expression and diagnosis of ASD is needed to improve the identification and treatment of girls with ASD.

Because ASD samples have been predominantly male, our current understanding of ASD and the behavioral criteria used to diagnose ASD may be biased toward males. Therefore, a different behavioral expression of the underlying biological liability for ASD in females compared to males, possibly due to different sociocultural influences, may contribute to the difficulty of identifying ASD in girls (Kreiser and White, 2014). Although results have been mixed, findings suggest that girls with ASD show similar or heightened levels of social-communication difficulties compared to boys, but less repetitive and restricted behavior (RRB; Frazier et al., 2014; Hartley and Sikora, 2009; Van Wijngaarden-Cremers et al., 2014). In addition, it is possible that girls show other types of RRB symptoms than boys that are less well identified using current assessment tools (Hiller et al., 2016; Mandy et al., 2012). To date, little is known about whether sensory symptoms, which are newly added in the Diagnostic and Statistical Manual of Mental Disorders (5th ed.; DSM-5) as part of the RRB domain, also differ between boys and girls with ASD.

The clinical presentation and identification of girls with ASD may also be affected by symptoms outside the ASD core domains. The few studies that have explored this indicate that girls are less likely to show externalizing problems (Hiller et al., 2014; Mandy et al., 2012; May et al., 2012) and more likely to show internalizing problems than boys (Hartley and Sikora, 2009; Mandy et al., 2012; Solomon et al., 2012), but results have been inconsistent and seem to vary according to the informant used. Moreover, few studies used comparison groups, so little is known about whether these differences are specific for individuals with ASD.

Research regarding gender differences in behavioral characteristics in ASD samples is complicated by a possible bias in the ascertainment of girls with ASD, risking a circularity of reasoning (Lai et al., 2015). A population study found that girls were more likely to be diagnosed with ASD than boys when they had higher levels of additional behavioral and cognitive difficulties despite showing similarly elevated levels of autistic symptoms (Dworzynski et al., 2012). This could reflect a diagnostic bias, resulting in girls with ASD being more likely to be overlooked in the absence of additional problems (Dworzynski et al., 2012). Therefore, the field could benefit from a different approach to examining the identification of ASD in females, not only focusing on phenotypic differences between boys and girls within ASD samples but also on how autistic symptoms and associated problems influence the probability of receiving an ASD diagnosis in girls versus boys.

This study aimed to contribute to a better understanding of gender differences in the expression of behavioral characteristics associated with a diagnosis of ASD in a sample of predominantly cognitively able children who had been consecutively referred to one of six participating mental health centers. First, we investigated differences in the proportions of boys and girls who received a positive screen for ASD or a best-estimate consensus diagnosis of ASD based on gold-standard diagnostic assessment. Second, we investigated whether overall autistic impairment, RRB symptoms, sensory symptoms, emotional and behavioral problems, and cognitive functioning differentially influenced the probability of the best-estimate ASD diagnosis in girls versus boys. Because previous studies found different results regarding the phenotypic characteristics of boys and girls with ASD according to the informant used (e.g., Mandy et al., 2012), we included both parent and teacher ratings of autistic symptoms and emotional/behavioral problems. Based on the study of Dworzynski et al. (2012), we hypothesized that higher levels of emotional and behavioral problems and lower levels of cognitive functioning would increase the probability that girls receive an ASD diagnosis, whereas this would be less so in boys. In addition, because many studies have found lower levels of RRB symptoms in girls with ASD than in boys with ASD (e.g. Van Wijngaarden-Cremers et al., 2014), we hypothesized that RRB symptoms would be less predictive of an ASD diagnosis in girls than in boys.

Methods

Participants

Participants were derived from the Social Spectrum Study, a prospective multicenter cohort of clinically referred children with a focus on ASD. This study will be described in more detail elsewhere (Duvekot et al., 2016). Figure 1 presents the flow of the participants through the various stages of the study and the proportions of boys and girls at each stage. First, all children aged 2.5–10 years who had been referred to one of six child and adolescent mental health services (CAMHS) in the South-West of the Netherlands were routinely screened for ASD using the parent-reported Social Responsiveness Scale (2nd ed.; SRS; Constantino and Gruber, 2012). Participating CAMHS included both secondary and tertiary (specialized) mental health services. Screening took place during a 6-month window varying between April 2011 and July 2012. Children were referred by the general practitioner or other medical doctor for a variety of problems, including concentration/hyperactivity problems (24%), behavior problems (24%), social/contact problems (23%), anxiety/mood problems (12%), learning/cognitive problems (7%), or other developmental concerns (10%). The reason for referral did not differ significantly by gender (χ²(5, 3311) = 13.1, p = 0.13).

Figure 1.

Flowchart of the participants.

Second, 428 of the 1281 screened children (118 girls and 310 boys) were identified as at risk of ASD (total raw score ⩾ 75 on the parent-reported SRS) and approached to participate in a comprehensive diagnostic assessment for ASD, including the short version of the Developmental, Dimensional and Diagnostic Interview (3Di; Santosh et al., 2009), the second edition of the Autism Diagnostic Observation Schedule (ADOS-2; De Bildt et al., 2013; Lord et al., 2012) and several questionnaires (including the Repetitive Behavior Scale–Revised (RBS-R) and the Short Sensory Profile (SSP), see the “Measures” section for further information). In addition, we asked a random selection of children who screened negative (n = 240; 76 girls and 164 boys) to participate in the same assessments in order to enable generalization of the results to the total sample of screened children.

To be included in the current analyses, both a 3Di and ADOS assessment had to be available for the child in order to establish a best-estimate consensus diagnosis of ASD. Of the children who were selected and did not meet this inclusion criterion (n = 437), 348 did not participate in any diagnostic assessment and 89 were excluded because there was only one diagnostic assessment available. The final sample of children who completed full diagnostic assessment (n = 231) consisted of 64 girls and 167 boys aged 2–12 years at the time of diagnostic assessments. Screen-positive children were more likely to participate than screen-negative children (χ²(1) = 13.91, p < 0.001). There was no statistically significant difference in the participation rates of screen-negative boys and girls, χ²(1) = 1.89, p = 0.17, or screen-positive boys and girls, χ²(1) =0.22, p = 0.64 (see Figure 1). IQ scores of the final sample ranged from 50 to 145, but the majority of the sample had IQ scores in the normal range: 11% of the children showed an indication of an intellectual disability (Full-Scale IQ, Verbal IQ, or Performance IQ < 70), with no gender difference in this proportion, χ²(1230) = 0.59, p = 0.53.

Diagnostic procedure

The comprehensive diagnostic assessment included the short version of the Developmental, Dimensional and Diagnostic Interview (3Di; Santosh et al., 2009) and the ADOS-2 (De Bildt et al., 2013; Lord et al., 2012), and an IQ assessment was performed if IQ was unknown or based on an assessment that had been conducted more than 2 years ago. In addition, parents completed questionnaires regarding characteristics of the child (see section “Measures”), parent, and family. At the time of these assessments, written informed consent was obtained from the participating families. The study was approved by the Ethics Committee of the Erasmus Medical Center (MEC-2011-078).

Information from the 3Di and ADOS assessments was used to determine the presence or absence of an ASD diagnosis (i.e. autistic disorder, Asperger’s syndrome, or Pervasive Developmental Disorder–Not Otherwise Specified (PDD-NOS)) according to the Diagnostic and Statistical Manual of Mental Disorders (4th ed.; text rev.; DSM-IV-TR). All administrators of the 3Di and ADOS had achieved the standard of research reliability. If both the 3Di and ADOS had been performed as part of the research protocol (n = 176, 76%), the diagnosis was based on consensus by two of the psychologists of the research team who administered these instruments. First, they each independently rated a DSM-IV-TR symptom checklist based on the information from the specific instrument they administered. Then, they discussed their checklists and exchanged information until they reached consensus about the presence of each symptom and formed a best-estimate consensus ASD diagnosis based on information from both instruments. Inter-rater reliability between the indication of an ASD diagnosis based on the DSM-IV-TR symptom checklist that was based on information from each instrument and the consensus diagnosis was good: kappa = 0.78 for the checklist based on the 3Di and kappa = 0.70 for the checklist based on the ADOS (similar findings for boys and girls). In 55 of the 231 cases (24%), the 3Di and/or ADOS had been conducted during the clinical evaluation, mostly at a tertiary CAMHS specialized in ASD. In these cases, the diagnosis that was formed by a team of experienced clinicians at the CAMHS was used. There was no difference in the proportion of boys and girls who received a diagnosis determined by the research or clinical team (χ(1) = 0.07, p = 0.79).

Measures

Autistic symptoms

The SRS-2 (Constantino and Gruber, 2012) is a 65-item questionnaire designed to assess the severity of autistic symptoms. In this study, we used the version for school-aged children as well as the version for preschool children, which are largely similar with a few differences in the content of items to make them more age-appropriate. Parents and teachers/day care providers completed the questionnaire as part of the routine referral procedure. Items are scored on a 4-point Likert scale ranging from 0 (“not true”) to 3 (“almost always true”). A total score is created by summing all 65 items, with higher scores indicating more overall autistic impairment. In a general population sample (n = 1104), boys obtained higher scores on the SRS than girls with an effect size (Cohen’s d) of 0.19 for parent ratings and of 0.37 for teacher ratings (Constantino and Gruber, 2012). Internal consistency of the total score was high in this study (Cronbach’s α = 0.95 for the parent-report and teacher-report). The SRS has been found to discriminate well between children with ASD and children with other psychiatric problems (e.g. Bölte et al., 2011; Charman et al., 2007; Constantino and Gruber, 2012), supporting its validity as an indicator of ASD symptom severity. Because the SRS predominantly contains items related to social-communication impairment, we also used the RBS-R and the SSP to assess symptoms that fall under the domain of RRB.

The RBS-R

RBS-R (Bodfish et al., 2000) is a parent-reported questionnaire designed to assess a variety of restricted and repetitive behaviors (i.e. self-injurious behavior, stereotypic behavior, compulsive behavior, ritualistic behavior, insistence on sameness, and restricted interests) that are characteristic of individuals with ASD with both lower and higher levels of intellectual functioning (Bishop et al., 2013; Lam and Aman, 2007). It consists of 43 items that are rated on a 4-point Likert scale ranging from 0 (“behavior does not occur”) to 3 (“behavior occurs and is a serious problem”). We used the total score including 38 of the original 43 items that was based on a factor analysis (Lam and Aman, 2007) as an overall indicator of severity of RRB. Consistent with previous research (Esbensen et al., 2009; Lam and Aman, 2007), internal consistency of the total scale in our sample was good, with a Cronbach’s alpha of 0.93. Solomon et al. (2012) reported no gender differences between typically developing girls and boys (both n = 19) on this measure.

The Short Sensory Profile

SSP (McIntosh et al., 1999) is a parent-reported questionnaire consisting of 38 items assessing the frequency of the child’s reactions to different sensory experiences. It is a shortened version of the Sensory Profile (Dunn, 1999). The items are scored on a 5-point Likert scale (1 = “always,” 2 = “frequently,” 3 = “occasionally,” 4 = “seldom,” 5 = “never”). We used the total score as an indicator of overall sensory processing ability. Note that lower scores reflect more sensory processing difficulties. In this study, the internal consistency of the total scale was 0.92. Criterion-validity of the SSP has been supported by findings that score on the SSP discriminated between children with ASD and typically developing children or children with developmental delays (Tomchek and Dunn, 2007). In a general population sample (n = 1115), gender differences on the Sensory Profile were found to be negligible (effect sizes < 0.10; Dunn and Westman, 1997).

Emotional and behavioral problems

The Child Behavior Checklist (CBCL), a parent-reported questionnaire, and a parallel form for teachers, the Teacher Report Form (TRF), were collected as part of the routine referral procedure to assess a broad variety of emotional and behavioral problems. Items are scored on a 3-point scale ranging from 0 (“not true”) to 3 (“very true”). The CBCL and TRF have one version for children aged 1.5–5 years old (Achenbach and Rescorla, 2000) and one for children aged 6–18 years (Achenbach and Rescorla, 2001; Verhulst and Van der Ende, 2013). Both versions were used in this study. In addition to the Total Problems score, we used the two empirically derived broadband scales: Internalizing and Externalizing problems. In the general population, boys score higher than girls on the Total and Externalizing Problems Scale and lower on the Internalizing Problems Scale (Crijnen et al., 1997). Because of the two age versions, we used T scores that are based on normative data to make the scores more comparable. The CBCL and TRF have good psychometric properties (Achenbach and Rescorla, 2000, 2001; Verhulst and Van der Ende, 2013) which has also been confirmed in ASD samples (Pandolfi et al., 2009, 2012).

Cognitive functioning

Because we obtained the majority of intelligence quotient (IQ) scores from the patient file (74%), a variety of tests were used to estimate the level of cognitive functioning: the Wechsler Intelligence Scale for Children (third Dutch ed.; WISC-III-NL; Kort et al., 2005), the Wechsler Preschool and Primary Scale of Intelligence (third Dutch ed.; WPPSI-III-NL; Hendriksen and Hurks, 2009), the Snijders-Oomen Nonverbal Intelligence Test–Revised (SON-R; Tellegen et al., 1998), the Bayley Scales of Infant Development (Dutch ed.; 2nd ed.; BSID-II-NL; Van der Meulen et al., 2004) or, as part of the research protocol (26%), the Wechsler Abbreviated Scale of Intelligence (WASI; Axelrod, 2002). All of these tests are standardized with a mean score of 100 and a standard deviation of 15. The WASI Full-Scale IQ has shown good concurrent validity with the WISC-III Full-Scale IQ (r = 0.87; Wechsler, 1999), the SON-R with the Wechsler Intelligence Scale for Children–Revised (WISC-R) Full Scale (r = 0.74; Tellegen et al., 1998), and the mental development index of the BSID-II with the WPPSI-R Full-Scale IQ (r = 0.73; Bayley, 1993).

Statistical analyses

Chi-square analyses were used to investigate differences in the probability of attaining a positive screen for ASD on the parent-reported SRS and a best-estimate consensus diagnosis of ASD (Aim 1). Differences in mean levels of characteristics between boys and girls in the total sample were tested using the complex samples general linear modeling procedure in SPSS 20 (SPSS Inc., Chicago, IL).

To investigate gender differences in the factors influencing the diagnosis of ASD (Aim 2), a series of logistic regression analyses were performed using the presence of an ASD diagnosis as the outcome variable and gender and age as covariates. In separate analyses, predictors included were (a) autistic symptoms: parent-reported SRS total score, teacher-reported SRS total score, RBS-R total score, and SSP total score; (b) emotional/behavioral problems: internalizing, externalizing, and total problem T scores of the CBCL and TRF; and (c) cognitive functioning: Verbal, Performance, and Full-Scale IQ scores. First, the main effects of these predictors were investigated. Then, to test whether the associations between these predictors and the probability of an ASD diagnosis differ for boys and girls, we included interaction terms between the predictors and gender. An interaction effect would indicate that there is a difference between boys and girls in how this characteristic is associated with an ASD diagnosis. To facilitate interpretation, the continuous predictors were standardized (transformed to z-scores).

The logistic regression analyses were performed using Mplus 7.3 (Muthén & Muthén, 1998–2012) using maximum likelihood estimation with robust standard errors (MLR) as estimator. Mplus uses the full information maximum likelihood (FIML) estimation to produce robust parameter estimates for the missing data based on available information in the data. To maximize this information, we added auxiliary variables that were related to the predictors in the model as additional dependent variables (Graham, 2003). The parent-reported SRS (0% missing), the teacher-reported SRS (12% missing), CBCL (7% missing), and TRF (20% missing) were used as auxiliary variables in each other’s models. The same was done for the Verbal IQ (16% missing), Performance IQ (10% missing), and Full-Scale IQ (6% missing). For the RBS-R (20% missing) and SSP (20% missing), we used the RRB scales of the parent-reported SRS, 3Di, and ADOS (calibrated score) and the total score of the CBCL as auxiliary variables. Because we oversampled children with a positive screen for ASD compared to children with a negative screen, we used sampling weights in our analyses to increase the generalizability of the results to the referred population from which we sampled. Statistical significance was determined at p < 0.05.

Results

Gender differences in ASD ascertainment

A similar proportion of clinically referred boys (310 of 885, 35%) and girls (118 of 396, 30%) screened positive on the parent-reported SRS (χ² = 3.36, p = 0.07). Of the 231 children who underwent full diagnostic assessments (consisting of the ADOS and 3Di), 106 boys and 24 girls (unweighted counts) received a best-estimate diagnosis of ASD. Correcting for sampling weights, 55% of the boys (weighted n = 154) and 25% of the girls (weighted n = 27) were estimated to have a best-estimate consensus diagnosis of ASD, indicating that boys were 2.18 times more likely to receive an ASD diagnosis than girls (χ² = 15.978, p < 0.001). Children who did not receive an ASD diagnosis (weighted n = 209) had a range of psychiatric diagnoses reported in the patient file, with ADHD as the most common diagnosis (39%), followed by anxiety/mood disorders (11%). The rates of these non-ASD diagnoses did not differ by gender (χ²(3, 297) = 4.4, p = 0.39).

Gender differences in the total sample

Table 1 shows the descriptive characteristics of the total included sample and non-ASD and ASD groups by gender. There were several general gender differences in the total sample, irrespective of ASD diagnosis. Compared to boys, girls were on average older (Wald F = 10.40 (230), p = 0.001), had higher levels of average IQ scores (Wald F ⩾ 3.96 (192), p < 0.05), more internalizing problems as reported by parents on the CBCL (Wald F = 5.54 (215), p = 0.02), and lower levels of autistic symptoms as reported by teachers on the SRS (Wald F = 7.63 (203), p = 0.006). These findings resembled gender differences in the initial screened sample, but for IQ this could not be examined because these data were not present for all screened children.

Table 1.

Descriptive statistics of boys and girls in the total sample and grouped by ASD diagnosis.

Characteristics	Total sample						Non-ASD						ASD
	Boys			Girls			Boys			Girls			Boys			Girls
	N	M	SD	N	M	SD	N	M	SD	N	M	SD	N	M	SD	N	M	SD
Age screening (years)	167	6.7	2.3	64	7.9	2.0	61	6.8	2.3	40	8.2	2.1	106	6.6	2.3	24	7.3	1.8
Age diagnostic assessment (years)	167	7.5	2.4	64	8.9	2.1	61	7.7	2.4	40	9.2	2.2	106	7.4	2.4	24	8.2	1.9
SRS parent total	167	70.5	27.7	64	66.4	30.7	61	60.5	25.8	40	57.6	32.1	106	78.9	26.6	24	92.8	26.6
SRS teacher total	146	69.5	30.0	58	54.2	25.1	54	61.8	27.8	36	52.0	23.1	92	76.2	30.3	22	60.4	38.6
CBCL internalizing	156	63.0	9.9	60	65.7	8.6	59	62.7	9.9	40	64.5	8.6	97	63.3	9.9	20	70.0	8.1
CBCL externalizing	156	64.7	10.8	60	65.2	9.4	59	65.2	10.9	40	64.3	9.4	97	64.2	10.7	20	68.5	9.6
CBCL total	156	67.1	8.5	60	66.8	7.2	59	67.4	8.4	40	65.6	7.6	97	66.8	8.7	20	71.1	6.0
TRF internalizing	133	60.8	8.8	53	60.7	9.2	50	61.0	10.0	34	60.8	9.8	83	60.6	8.1	19	60.3	8.1
TRF externalizing	133	60.0	10.3	53	62.2	9.4	50	60.5	10.1	34	62.9	8.3	83	59.5	10.3	19	60.2	11.1
TRF total	133	61.9	9.1	53	62.2	6.8	50	62.6	9.8	34	62.7	5.6	83	61.3	8.6	19	60.7	8.7
RBS-R total	133	12.7	14.0	53	11.0	14.6	50	8.9	12.0	32	9.8	16.4	83	16.6	14.4	21	14.3	11.2
SSP total^a	133	149.0	20.5	53	156.7	25.8	50	152.6	20.7	32	163.3	22.5	83	145.6	19.8	21	137.6	26.3
VIQ	136	96.0	15.1	57	103.8	17.6	53	94.2	15.4	36	105.3	17.0	83	97.8	14.9	21	98.9	19.1
PIQ	151	96.6	16.9	58	101.5	17.5	56	95.3	14.9	37	101.1	19.2	95	97.7	17.9	21	102.9	13.8
FSIQ	158	94.3	16.8	59	102.9	18.6	57	93.0	15.9	37	103.9	18.8	101	95.3	17.4	22	99.9	18.9

SRS: Social Responsiveness Scale; RBS-R: Repetitive Behavior Scale–Revised; SSP: Short Sensory Profile; CBCL: Child Behavior Checklist; TRF: Teacher Report Form; VIQ: Verbal IQ; PIQ: Performance IQ; FSIQ: Full-Scale IQ; SD: standard deviation; ASD: autism spectrum disorder.

Means (M) are weighted and counts (N) are unweighted.

Higher scores indicate less sensory processing difficulties.

Gender differences in factors related to an ASD diagnosis

Table 2 shows the results of the prediction of an ASD diagnosis by standardized measures of various behavioral characteristics and their interactions with gender (see Supplement 1 for the results using unstandardized predictors, available online). Parent-reported and teacher-reported autistic symptoms on the SRS and parent-reported sensory symptoms on the SSP significantly predicted an ASD diagnosis irrespective of gender. A significant interaction effect with gender was found for the RBS-R total scale, indicating that higher scores of restricted and repetitive behavior tended to be less predictive of an ASD diagnosis in girls than in boys. The odds ratio (OR) of 0.41 indicates that an increase of one standard deviation on the RBS-R total scale increased the odds of an ASD diagnosis in girls (OR = 1.10, 95% confidence interval (CI) = 0.48–2.45) less than half of what it increased the odds in boys (OR = 2.67, 95% CI = 1.50–4.75). In addition, there was a significant interaction between gender and the total score on the CBCL, indicating that girls were more likely to be diagnosed with ASD when they had higher total levels of behavioral problems (OR = 2.40, 95% CI = 1.13–5.29), whereas this effect was not present in boys (OR = 0.98, 95% CI = 0.70–1.38). To illustrate these interactions, Figures 2 and 3 show the mean levels on these scales in boys and girls with and without ASD. No main effect nor interaction effect with gender was found for IQ and TRF scores.

Table 2.

Logistic regression analyses predicting the probability of an ASD diagnosis.

Characteristics	Main effects			Interaction with gender
	OR	95% CI	p	OR	95% CI	p
SRS parent total	2.13	1.54–2.95	<0.001	1.38	0.34–2.05	0.40
SRS teacher total	1.72	1.19–2.50	0.004	0.88	0.97–1.02	0.77
RBS-R total	1.76	0.94–3.30	0.08	0.41	0.18–0.92	0.03
SSP total^a	0.58	0.40–0.85	0.005	0.63	0.30–1.31	0.21
CBCL internalizing	1.20	0.86–1.65	0.28	2.00	0.88–4.56	0.10
CBCL externalizing	1.02	0.75–1.39	0.88	1.73	0.90–3.32	0.10
CBCL total	1.13	0.84–1.53	0.43	2.44	1.13–5.29	0.02
TRF internalizing	1.01	0.69–1.48	0.96	0.81	0.37–1.76	0.59
TRF externalizing	0.91	0.61–1.34	0.62	0.71	0.26–1.94	0.50
TRF total	0.89	0.61–1.31	0.56	0.67	0.22–2.03	0.48
Verbal IQ	1.02	0.75–1.39	0.90	0.56	0.28–1.11	0.09
Performance IQ	1.07	0.78–1.47	0.67	0.93	0.48–1.81	0.83
Full-Scale IQ	1.08	0.78–1.50	0.65	0.73	0.37–1.44	0.37

SRS: Social Responsiveness Scale; RBS-R: Repetitive Behavior Scale–Revised; SSP: Short Sensory Profile; CBCL: Child Behavior Checklist; TRF: Teacher Report Form; VIQ: Verbal IQ; PIQ: Performance IQ; FSIQ: Full-Scale IQ; OR: odds ratio; CI: confidence interval; ASD: autism spectrum disorder.

Boldface type indicates that interactions with gender reached significance (p < 0.05). Predictor variables are standardized (z-scores).

Higher scores indicate less sensory processing difficulties.

Figure 2.

Mean levels of total RRB symptoms on the Repetitive Behavior Checklist–Revised (RBS-R) in boys and girls with and without ASD.

Figure 3.

Mean levels of total emotional and behavioral problems on the Child Behavior Checklist (CBCL) in boys and girls with and without ASD.

Discussion

This study examined whether there are differences in how individual characteristics influence an ASD diagnosis in clinically referred girls versus boys. We found that higher overall levels of parent-reported RRB symptoms were less predictive of an ASD diagnosis in girls than in boys. In contrast, higher overall levels of parent-reported emotional and behavioral problems increased the probability of an ASD diagnosis in girls, but not in boys. No gender differences were found in the prediction of an ASD diagnosis by overall levels of autistic symptoms, sensory symptoms, and cognitive functioning. These findings may contribute to our understanding of why girls are less likely to be diagnosed with ASD than boys.

In our sample of clinically referred children, similar proportions of boys and girls were identified as having elevated ASD symptoms as indicated on the SRS. Since we do not know of any other study that screened for ASD in a clinically referred sample irrespective of referral reason, we cannot say whether our screening rates are consistent with other studies. However, despite similar screening rates in boys and girls, girls were less likely to receive an ASD diagnosis based on the standardized diagnostic instruments. This could mean that girls with ASD are at risk of being underidentified using current diagnostic instruments. Consistently, using an adult sample, Lai et al. (2011) reported that only 21% of women with ASD met criteria for an ASD classification on the ADOS, compared to 58% of the men with ASD. A possible reason is that females with ASD may be better at masking their problems during a short observation (e.g. Hiller et al., 2016; Kreiser and White, 2014; Rynkiewicz et al., 2016). So, diagnostic instruments and/or their manuals may need to be adapted to improve the identification of ASD in girls. For instance, some scoring items may need to be adapted to provide examples that are more characteristic of girls. In addition, administrators may need to gain more training/experience in scoring these instrument in girls. A more profound adaptation would be the development of gender-specific cut-offs. This could lead to more girls with ASD being identified, but possibly also to an overinclusion of girls who deviate too much from the conceptualization of ASD. Clearly, before more specific recommendations can be made about possible adaptations, more information is needed about why females are less likely to reach the diagnostic threshold and whether adaptations truly result in an improved discriminative ability of these instruments in girls.

As hypothesized, we found that overall RRB symptoms were less strongly associated with ASD in girls than in boys. Thus, whereas there was a notable contrast in the level of RRB symptoms between boys with ASD and non-ASD, with boys with ASD showing higher levels of RRB symptoms, this was not the case for girls. One possible explanation is that girls with ASD are characterized by lower levels of RRB symptoms than boys, suggesting a quantitative difference, in line with findings of previous studies that investigated mean differences in RRB symptoms between boys and girls with ASD (Frazier et al., 2014; Mandy et al., 2012; Szatmari et al., 2012). However, we did not find an overall quantitative difference in the level of RRB symptoms between referred boys and girls, which may imply that it is not just a matter that girls have a lower likelihood of ASD because of lower levels of RRB symptoms in general. It also possible that RRB symptoms in girls with ASD are qualitatively different from those in boys and are therefore not adequately captured by current instruments or less likely to be recognized by clinicians as being characteristic of ASD (Hiller et al., 2014, 2016; Mandy et al., 2012). For example, girls with ASD may be less likely to show stereotyped use of objects (e.g. lining up toys) and their restricted interests may concern topics that are socially accepted for girls (e.g. horses or Barbie dolls; Attwood, 2007) or that seem random (e.g. rocks, stickers, and pens; Hiller et al., 2014). It should be noted that the measure for RRB symptoms used in this study, the RBS-R, only contains few items on restricted interests and may therefore lack sensitivity to a differential expression of this domain in cognitively able girls with ASD. A failure of instruments to capture the expression of these symptoms in girls could also contribute to apparent quantitative differences between boys and girls with ASD (Van Wijngaarden-Cremers et al., 2014). Therefore, further research at a more detailed level is needed to advance our understanding of the expression of RRB symptoms in girls with ASD.

In contrast, sensory symptoms, which are added to the RRB domain in the DSM-5, were positively associated with an ASD diagnosis in both boys and girls. In the light of the finding that overall RRB symptoms did not contribute to an ASD diagnosis in girls, this might suggest that the evaluation of sensory symptoms is particularly important for the evaluation of ASD in girls. Few studies have yet compared the expression of sensory symptoms associated with ASD between girls and boys. In a study among adults, there was preliminary evidence that women with ASD showed more sensory symptoms on the Autism Diagnostic Interview–Revised (ADI-R) than men with ASD (Lai et al., 2011). This needs to be examined further, with attention to different types of sensory symptoms.

The finding that higher overall levels of emotional and behavioral problems increased the probability that girls, but not boys, received an ASD diagnosis is in line with a previous general population study (Dworzynski et al., 2012). There are several possible explanations for this finding. First, this could indicate that girls with ASD are vulnerable for experiencing high levels of co-occurring emotional and behavioral problems. Case reports described that cognitively able girls with ASD can be sensitive to social expectations and sometimes copy social behaviors from peers or characters in books or television shows (Bargiela et al., 2016). In that way, they camouflage their limitations, but this takes a lot of their energy, which may lead to drained or edgy feelings and behaviors that are noticeable in the home setting. Second, this finding may reflect a diagnostic bias, indicating that girls with ASD without these problems are overlooked. Qualitative studies have also reported how some girls have struggled a long time before they received an ASD diagnosis (Bargiela et al., 2016; Cridland et al., 2014). The timely identification of girls with ASD may not only be important to provide them with services to improve their outcomes, which is an important subject for further investigation (Wong et al., 2015), but also to provide these girls with a sense of belonging and understanding of their difficulties (Bargiela et al., 2016). Third, and not necessarily in contradiction with the former explanation, girls without high levels of emotional and behavioral problems may be better able to compensate for possibly elevated ASD problems (Dworzynski et al., 2012; Mandy et al., 2012). However, it is debatable whether this latter group—if they would not meet diagnostic criteria for ASD when subtle variations in the expression of symptoms are taken into account—should be regarded in the light of the autism phenotype. In addition, further research is needed to investigate whether these girls are at risk of developing more problems or significant distress over time or whether they really function better (Kreiser and White, 2014).

We found some important informant differences. Only emotional and behavioral problems reported by parents, but not by teachers, increased the probability that girls received an ASD diagnosis. Although Dworzynski et al. (2012) did find that girls for whom the teacher reported high levels of total behavioral and hyperactivity problems were more likely to be diagnosed with ASD, these findings are difficult to compare to our own because they used a general population sample and did not include parent ratings. For ADHD, some differences were found between boys and girls with ADHD in the general population that could not be detected in clinical samples, possibly because referred girls are more severely affected (Gaub and Carlson, 1997). The informant discrepancy in our study could indicate that teachers are less likely to recognize difficulties in girls with ASD. Teachers reported lower levels of autistic symptoms in referred girls than in referred boys, which is consistent with previous studies using ASD samples (Hiller et al., 2014; Mandy et al., 2012) and general population samples (Constantino and Gruber, 2012; Posserud et al., 2006). Previous studies also found that teachers reported lower levels of externalizing behavior in girls than in boys with ASD (Hiller et al., 2014; Mandy et al., 2012). The “camouflaging” abilities of some cognitively able girls with ASD may also contribute to the discrepancy between teacher and parent ratings. So, caution is needed in relying on teacher ratings to screen for ASD in girls, though they still can provide valuable information in addition to parent ratings (Duvekot et al., 2015). More research is needed to better understand these informant effects in relation to gender differences and how these can be dealt with in order to improve the identification of ASD in girls.

In contrast to the study by Dworzynski et al. (2012), we did not find that lower levels of IQ were more strongly associated with an ASD diagnosis in girls than in boys. This is surprising, given that is often assumed that girls with ASD have a greater risk of cognitive impairments than boys with ASD (Fombonne, 2003, 2005; Frazier et al., 2014; Volkmar et al., 1993). However, several recent studies also failed to find gender differences for IQ (Hartley and Sikora, 2009; Mandy et al., 2012). Similar to these studies, our sample showed a wide range of IQ, but with the majority of children showing a normal/high IQ level because that was the target population of the majority of the participating CAMHS. It is possible that there is only an increase in the probability of an ASD diagnosis in girls with an intellectual impairment (IQ < 70). This should be examined further in samples with more individuals in the lower IQ range.

Several studies reported that girls with ASD have higher levels of internalizing problems than boys with ASD (Hartley and Sikora, 2009; Mandy et al., 2012; Oswald et al., 2016; Solomon et al., 2012). Although we found that girls had higher levels of internalizing problems than boys in the total sample, the absence of a significant interaction effect with gender in the prediction of an ASD diagnosis suggests that this may reflect a general gender difference in internalizing problems, consistent with findings of gender differences in internalizing symptoms in other referred populations (Compas et al., 1997) and the general population (Crijnen et al., 1997), rather than a specific vulnerability of girls with ASD to develop internalizing problems. Future research should also examine this in older samples, as it could be that ASD-specific gender differences in internalizing symptoms emerge later during adolescence (Oswald et al., 2016). However, even if internalizing problems are not specific for girls with ASD, the heightened vulnerability of girls to develop internalizing problems has implications for the treatment needs of girls with ASD.

Strengths and limitations

A strength of this study is that we screened a large sample of children who had been referred to multiple mental health services and used well-established standardized instruments for the diagnosis of ASD in a selection of this sample. Importantly, we did not only perform diagnostic assessment for ASD in children with a positive screen for ASD but also in a random selection of children with a negative screen. This design may have reduced the diagnostic bias that may be pronounced in clinical samples that recruited children with a previously established diagnosis of ASD. Moreover, because we did not use the presence of an ASD diagnosis as an inclusion criterion, our design was suitable to investigate factors that determine the probability of receiving an ASD diagnosis. Other strengths are that we assessed a wide variety of characteristics and used multiple informants for some assessments.

Our study also has several limitations. Since we used a clinically referred sample, our findings cannot be generalized to the general population. Certain gender biases may already have been present at referral. Consistent with the literature, the initial referred sample already consisted of fewer girls than boys, and referred girls were also significantly older than referred boys. This could reflect that girls are more likely to be referred for internalizing problems, which tend to increase with age, whereas boys are more likely to be referred for externalizing problems (Zwaanswijk et al., 2003). Furthermore, girls with ASD might only be referred if they have severe symptoms or symptoms that more closely resemble those in males. Therefore, it also important to investigate factors related to identification of ASD in general population samples.

In addition, we used two different versions of the CBCL. Because girls were older than boys, a greater proportion of girls (82%) than boys (64%) received the CBCL/6-18 instead of the CBCL/1.5-5. Although we have taken measures to account for the age difference between boys and girls and the use of different age versions of the CBCL by correcting for age in our analyses and using standardizing scores, we cannot rule out the possibility that this may have influenced our findings. Another limitation is that the number of girls with ASD included in the sample was small, which limited the power to detect small interaction effects. To limit the number of tests conducted in our small sample, we examined total scores rather than subscale scores for most measures. Moreover, we used conventional, standardized measures that may be biased toward symptoms that are characteristic of males with ASD and may therefore be less sensitive to detect subtle differences between boys and girls with ASD (Lai et al., 2015). So, further research using other methods is needed to investigate fine-grained differences between boys and girls at a more detailed level in larger samples.

Conclusion

Our results suggest that some individual behavioral characteristics (i.e. RRB symptoms and emotional and behavioral problems) affect the diagnosis of ASD differently in girls than in boys, possibly contributing to an underidentification of ASD in girls. One of the factors that may contribute to a lower probability of girls to be diagnosed with ASD is that RRB symptoms are not as predictive of an ASD diagnosis in girls as in boys. The finding that sensory symptoms were equally predictive of an ASD diagnosis in girls as in boys needs to be investigated further and suggest the importance of assessing sensory symptoms in the diagnostic evaluation of ASD in girls. We also found support that girls, but not boys, were more likely to be diagnosed with ASD if they had higher levels of emotional and behavioral problems. This highlights that it is important to be aware of high levels of co-occurring emotional and behavioral problems in girls with ASD and the possibility that girls with ASD who do not display high levels of co-occurring emotional and behavioral problems may be at risk of being overlooked. Further research is needed to also investigate the possibility that girls who display subclinical levels of autistic symptoms in the absence of these problems have compensatory abilities that prevent them from reaching the clinical threshold and whether these girls are at risk of developing more autistic or other difficulties over time.

Footnotes

Acknowledgements

The Social Spectrum Study was conducted by the Erasmus Medical Center in collaboration with Emergis, GGZ WNB, Lucertis, Riagg Rijnmond, and Yulius. In addition, the authors gratefully acknowledge the families that participated in the study and the graduate students, PhD-students, and research assistants who helped conducting the study.

Declaration of conflicting interests

F.C.V. is a contributing author of the Achenbach System of Empirically Based Assessment (ASEBA), from which he has received remuneration. K.G.-L. is a contributing author of the Dutch ADOS-2 manual, for which the mental health organization Yulius receives remuneration. The remaining authors have declared that they have no potential or competing conflicts of interest in relation to this work.

Funding

This research was funded by the Sophia Foundation for Scientific Research (SSWO project no. 958).

References

Achenbach

Rescorla

(2000) Manual for the ASEBA Preschool Forms & Profiles. Burlington, VT: University of Vermont, Research Center for Children, Youth, and Families.

Achenbach

Rescorla

(2001) Manual for the ASEBA School-age Forms & Profiles. Burlington, VT: University of Vermont, Research Center for Children, Youth, and Families.

Attwood

(2007) The Complete Guide to Asperger’s Syndrome. London: Jessica Kingsley Publishers.

Axelrod

(2002) Validity of the Wechsler abbreviated scale of intelligence and other very short forms of estimating intellectual functioning. Assessment 9: 17–23.

Bargiela

Steward

Mandy

(2016) The experiences of late-diagnosed women with autism spectrum conditions: an investigation of the female autism phenotype. Journal of Autism and Developmental Disorders. Epub ahead of print 25 July 2016. DOI: 10.1007/s10803-016-2872-8.

Bayley

(1993) Bayley Scales of Infant Development. 2nd ed. San Antonio, TX: The Psychological Corporation.

Begeer

Mandell

Wijnker-Holmes

. (2013) Sex differences in the timing of identification among children and adults with autism spectrum disorders. Journal of Autism and Developmental Disorders 43: 1151–1156.

Bishop

Hus

Duncan

. (2013) Subcategories of restricted and repetitive behaviors in children with autism spectrum disorders. Journal of Autism and Developmental Disorders 43: 1287–1297.

Bodfish

Symons

Parker

. (2000) Varieties of repetitive behavior in autism: comparisons to mental retardation. Journal of Autism and Developmental Disorders 30: 237–243.

10.

Bölte

Westerwald

Holtmann

. (2011) Autistic traits and autism spectrum disorders: the clinical validity of two measures presuming a continuum of social communication skills. Journal of Autism and Developmental Disorders 41: 66–72.

11.

Charman

Baird

Simonoff

. (2007) Efficacy of three screening instruments in the identification of autistic-spectrum disorders. British Journal of Psychiatry 191: 554–559.

12.

Compas

Oppedisano

Connor

. (1997) Gender differences in depressive symptoms in adolescence: comparison of national samples of clinically referred and non-referred youths. Journal of Consulting and Clinical Psychology 65: 617–626.

13.

Constantino

Gruber

(2012) Social Responsiveness Scale (SRS-2). 2nd ed. Los Angeles, CA: Western Psychological Services.

14.

Cridland

Jones

Caputi

. (2014) Being a girl in a boys’ world: investigating the experiences of girls with autism spectrum disorders during adolescence. Journal of Autism and Developmental Disorders 44: 1261–1274.

15.

Crijnen

Achenbach

Verhulst

(1997) Comparisons of problems reported by parents of children in 12 cultures: total problems, externalizing, and internalizing. Journal of the American Academy of Child and Adolescent Psychiatry 36: 1269–1277.

16.

De Bildt

Greaves-Lord

De Jonge

(2013) ADOS-2: Autisme diagnostisch observatieschema. Handleiding. [ADOS 2nd Edition: Autism Diagnostic Observation Schedule, Dutch Manual]. Amsterdam: Hogrefe.

17.

Duvekot

Hoopen

Slappendel

. (2016) Design and cohort characteristics of the social spectrum study: a multicenter study of the autism spectrum among clinically referred children. Journal of Autism and Developmental Disorders. Epub ahead of print 3 October 2016. DOI: 10.1007/s10803-016-2919-x.

18.

Duvekot

van der Ende

Verhulst

. (2015) The screening accuracy of the parent and teacher-reported social responsiveness scale (SRS): comparison with the 3Di and ADOS. Journal of Autism and Developmental Disorders 45: 1658–1672.

19.

Dunn

(1999) Sensory Profile-NL 3 t/m 10 jaar: Nederlandse vertaling A. Rietman. [Sensory Profile 3 to 10 years old: trans. A Rietman (Dutch)]. Amsterdam: Harcourt Test Publishers.

20.

Dunn

Westman

(1997) The sensory profile: the performance of a national sample of children without disabilities. American Journal of Occupational Therapy 51: 25–34.

21.

Dworzynski

Ronald

Bolton

. (2012) How different are girls and boys above and below the diagnostic threshold for autism spectrum disorders? Journal of the American Academy of Child and Adolescent Psychiatry 51: 788–797.

22.

Esbensen

Seltzer

Lam

. (2009) Age-related differences in restricted repetitive behaviors in autism spectrum disorders. Journal of Autism and Developmental Disorders 39: 57–66.

23.

Fombonne

(2003) Epidemiological surveys of autism and other pervasive developmental disorders: an update. Journal of Autism and Developmental Disorders 33: 365–382.

24.

Fombonne

(2005) The changing epidemiology of autism. Journal of Applied Research in Intellectual Disabilities 18: 281–294.

25.

Frazier

Georgiades

Bishop

. (2014) Behavioral and cognitive characteristics of females and males with autism in the Simons simplex collection. Journal of the American Academy of Child and Adolescent Psychiatry 53: 329–340.

26.

Gaub

Carlson

(1997) Gender Differences in ADHD: a Meta-Analysis and Critical Review. Journal of the American Academy of Child and Adolescent Psychiatry 36: 1036–1045.

27.

Giarelli

Wiggins

Rice

. (2010) Sex differences in the evaluation and diagnosis of autism spectrum disorders among children. Disability and Health Journal 3: 107–116.

28.

Graham

(2003) Adding missing-data-relevant variables to FIML-based structural equation models. Structural Equation Modeling 10: 80–100.

29.

Hartley

Sikora

(2009) Sex differences in autism spectrum disorder: an examination of developmental functioning, autistic symptoms, and coexisting behavior problems in toddlers. Journal of Autism and Developmental Disorders 39: 1715–1722.

30.

Hendriksen

JGM

Hurks

PPM

(2009) WPPSI-III-NL: Wechsler Preschool and Primary Scale of Intelligence. 3rd ed. Amsterdam: Pearson Assessment and Information B.V.

31.

Hiller

Young

Weber

(2014) Sex differences in autism spectrum disorder based on DSM-5 criteria: evidence from clinician and teacher reporting. Journal of Abnormal Child Psychology 42: 1381–1393.

32.

Hiller

Young

Weber

(2016) Sex differences in pre-diagnosis concerns for children later diagnosed with autism spectrum disorder. Autism 20: 75–84.

33.

Jacquemont

Coe

Hersch

. (2014) A higher mutational burden in females supports a “female protective model” in neurodevelopmental disorders. American Journal of Human Genetics 94: 415–425.

34.

Kort

Schittekatte

Dekker

. (2005) WISC-III-NL: Wechsler Intelligence Scale for Children. 3rd ed. Amsterdam: Pearson Assessment and Information B.V.

35.

Kreiser

White

(2014) ASD in females: are we overstating the gender difference in diagnosis? Clinical Child and Family Psychology Review 17: 67–84.

36.

Lai

Lombardo

Auyeung

. (2015) Sex/gender differences and autism: setting the scene for future research. Journal of the American Academy of Child and Adolescent Psychiatry 54: 11–24.

37.

Lai

Lombardo

Pasco

. (2011) A behavioral comparison of male and female adults with high functioning autism spectrum conditions. PLoS ONE 6: e20835.

38.

Lam

Aman

(2007) The repetitive behavior scale-revised: independent validation in individuals with autism spectrum disorders. Journal of Autism and Developmental Disorders 37: 855–866.

39.

Lord

Rutter

DiLavore

. (2012) Autism Diagnostic Observation Schedule (ADOS-2) Manual (Part II). 2nd ed. Los Angeles, CA: Western Psychological Services.

40.

McIntosh

Miller

Shyu

(1999) Development and validation of the short sensory profile. In: Dunn

(ed.) Sensory Profile 3 t/m 10 jaar. Nederlandse vertaling A. Rietman [Sensory Profile 3 to 10 years: trans. A Rietman (Dutch)] Amsterdam: Harcourt Test Publishers, pp.59–73.

41.

Mandy

Chilvers

Chowdhury

. (2012) Sex differences in autism spectrum disorder: evidence from a large sample of children and adolescents. Journal of Autism and Developmental Disorders 42: 1304–1313.

42.

May

Cornish

Rinehart

(2012) Gender profiles of behavioral attention in children with autism spectrum disorder. Journal of Attention Disorders 20: 627–635.

43.

Oswald

Winter-Messiers

Gibson

. (2016) Sex differences in internalizing problems during adolescence in autism spectrum disorder. Journal of Autism and Developmental Disorders 46: 624–636.

44.

Pandolfi

Magyar

Dill

(2009) Confirmatory factor analysis of the child behavior checklist 1.5-5 in a sample of children with autism spectrum disorders. Journal of Autism and Developmental Disorders 39: 986–995.

45.

Pandolfi

Magyar

Dill

(2012) An initial psychometric evaluation of the CBCL 6-18 in a sample of youth with autism spectrum disorders. Research in Autism Spectrum Disorders 6: 96–108.

46.

Posserud

Lundervold

Gillberg

(2006) Autistic features in a total population of 7-9 year-old children assessed by the ASSQ (Autism Spectrum Screening Questionnaire). Journal of Child Psychology and Psychiatry, and Allied Disciplines 47: 167–175.

47.

Robinson

Lichtenstein

Anckarsater

. (2013) Examining and interpreting the female protective effect against autistic behavior. Proceedings of the National Academy of Sciences of the United States of America 110: 5258–5262.

48.

Russell

Steer

Golding

(2011) Social and demographic factors that influence the diagnosis of autistic spectrum disorders. Social Psychiatry and Psychiatric Epidemiology 46: 1283–1293.

49.

Rynkiewicz

Schuller

Marchi

. (2016) An investigation of the “female camouflage effect” in autism using a computerized ADOS-2 and a test of sex/gender differences. Molecular Autism 7: 10.

50.

Santosh

Mandy

WPL

Puura

. (2009) The construction and validation of a short form of the developmental, diagnostic and dimensional interview. European Child & Adolescent Psychiatry 18: 521–524.

51.

Solomon

Miller

Taylor

. (2012) Autism symptoms and internalizing psychopathology in girls and boys with autism spectrum disorders. Journal of Autism and Developmental Disorders 42: 48–59.

52.

Szatmari

Liu

Goldberg

. (2012) Sex differences in repetitive stereotyped behaviors in autism: implications for genetic liability. American Journal of Medical Genetics Part B: Neuropsychiatric Genetics 159B: 5–12.

53.

Taylor

Lichtenstein

Larsson

. (2016) Is there a female protective effect against attention-deficit/hyperactivity disorder? Evidence from two representative twin samples. Journal of the American Academy of Child and Adolescent Psychiatry 55: 504–51200.

54.

Tellegen

Winkel

Wijnberg-Williams

. (1998) Snijders-Oomen Niet-verbale Intelligentietest: SON-R 2½-7 Handleiding and verantwoordig [Snijders-Oomen Nonverbal Intelligence Test: SON-R 2½-7 Manual and Research Report]. Lisse: Swets & Zeitlinger B.V.

55.

Tomchek

Dunn

(2007) Sensory processing in children with and without autism: a comparative study using the short sensory profile. American Journal of Occupational Therapy 61: 190–200.

56.

Van der Meulen

Ruiter

SAJ

Lutje Spelberg

. (2004) Bayley Scales of Infant Development (BSID-II-NL). 2nd ed. Amsterdam: Harcourt Test Publishers.

57.

Van Wijngaarden-Cremers

Van Eeten

Groen

. (2014) Gender and age differences in the core triad of impairments in autism spectrum disorders: a systematic review and meta-analysis. Journal of Autism and Developmental Disorders 44: 627–635.

58.

Verhulst

Van der Ende

(2013) Handleiding ASEBA: Vragenlijsten voor leeftijden 6 tot en met 18 jaar. [ASEBA Manual: Questionnaires for ages 6 to 18 years] Rotterdam: ASEBA Nederland.

59.

Volkmar

Szatmari

Sparrow

(1993) Sex differences in pervasive developmental disorders. Journal of Autism and Developmental Disorders 23: 579–591.

60.

Wechsler

(1999) Wechsler Abbreviated Scale of Intelligence (WASI). San Antonio, TX: The Psychological Corporation.

61.

Willcutt

(2012) The prevalence of DSM-IV attention-deficit/hyperactivity disorder: a meta-analytic review. Neurotherapeutics 9: 490–499.

62.

Wong

Odom

Hume

. (2015) Evidence-based practices for children, youth, and young adults with autism spectrum disorder: a comprehensive review. Journal of Autism and Developmental Disorders 45: 1951–1966.

63.

Zwaanswijk

Verhaak

Bensing

. (2003) Help seeking for emotional and behavioural problems in children and adolescents: a review of recent literature. European Child & Adolescent Psychiatry 12: 153–161.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.03 MB