sign in sign up
<<
Home , Neurological disorder

RESEARCH ARTICLE Specific Neurological Phenotypes in Spectrum Disorders Are Associated with Sex Representation Esther Ben-Itzchak, Shay Ben-Shachar, and Ditza A. Zachor

Autism spectrum disorder (ASD) is a heritable disorder occurring predominantly in males. The aim of this study was to compare sex differences in the prevalence of specific neurological phenotypes commonly described in ASD. The study included 663 participants, aged 18 months to 15 years, diagnosed with ASD. Neurological and behavioral assessments were performed using standardized tests, and obtaining medical, developmental, and familial histories from the parents. Phenotypes under investigation were macro- and microcephaly, developmental regression, minor neurological and musculoskeletal deficits (MNMD), and . Male : female ratio in the ASD group was 6.7:1. No sex differences in autism severity, cognitive ability, and adaptive functioning were noted. Mean head circumference percentile for males (50.1 ± 25.6) was significantly larger than females (43.4 ± 30.2). Micro- and were more frequent in ASD than expected (5.9%; 18.1%, respectively). Microcephaly in females (15.1%) was significantly more prevalent than in males (4.5%). The prevalence of macrocephaly in both sexes did not differ significantly. Regression was noted in 30.2% of the females with ASD, significantly higher than in males (18.9%). MNMD was documented in 73.8% of the females, significantly higher than in males (57.1%). M:F ratio decreased in a group with two or more phenotypes (3.6:1), while male predominance was more significant in the group without phenotypes (13.6:1). Neurological phenotypes associated with ASD are more prevalent in females than in males, resulting in more complex clinical and neurological manifesta- tions in females. Therefore, involvement of different etiologies is suggested in ASD in females. Autism Res 2013, ••: ••–••. © 2013 International Society for Autism Research, Wiley Periodicals, Inc.

Keywords: disorder; male:female ratio; microcephaly; macrocephaly; developmental regression, minor neurological and musculoskeletal deficits, seizures.

Introduction Minderaa, & Hadders-Algra, 2011; Ming, Brimacombe, & Wagner, 2007]. The most commonly reported neurologi- Autism spectrum disorders (ASD) are a group of neuro- cal deficits were hypotonia, abnormal reflexes, and behavioral disorders characterized by a classic triad of cranial dysfunction [De Jong et al., 2011]. Abnor- symptoms, including impaired social and communication mal neurological signs might indicate involvement of skills, and stereotyped behaviors. It is well accepted now multiple networks associated with ASD that may that ASD represents a heterogeneous group of conditions. affect neurological outcome. has been docu- It is not surprising, therefore, that some individuals show mented in about 5–46% of individuals with ASD in other phenotypes in addition to the core symptoms. ASD different studies, suggesting that autism and epilepsy is a highly heritable disorder, with heritability indices share a common neurobiological basis [Kagan-Kushnir, estimated at 60–92%, suggesting a major role for genetic Roberts, & Snead, 2005; Tuchman & Rapin, 2002]. factors in the etiology of ASD [Schaaf & Zoghbi, 2011]. Male predominance in ASD has been noted in all epi- Accelerated head growth and relative increased head demiological studies and is a well-established fact in the circumference have been documented in ASD in numer- field of ASD. The male : female (M:F) ratio has been ous studies [Courchesne & Pierce, 2005; Fombonne, Roge, reported in many studies in the range of 4–5:1 Claverie, Courty, & Fremolle, 1999; Minshew & Williams, [Fombonne, 2003; Holtmann, Bolte, & Poustka, 2007; 2007]. Developmental regression has been estimated at Newschaffer et al., 2007]. 15–30% in individuals with ASD [Baird et al., 2008; Parr In general, sex differences in autism severity and cog- et al., 2011]. Children with ASD are often described with nitive abilities have been examined in previous research. clumsiness and motor skills delay [Zachor, Ilanit, & Ben Studies looking at sex differences in autism severity Itzchak, 2010]. In addition, soft neurological signs were found conflicting results even when controlling for IQ. frequently documented in ASD [De Jong, Punt, De Groot, Several studies reported more severe symptoms in girls

From the Department of Communication Disorders, Ariel University Center of Samaria, Ariel, Israel (E.B-I.); The Autism Center, Department of Pediatrics, Assaf Harofeh Medical Center, Zerifin, Israel (E.B-I., D.A.Z.); The Genetic Institute, Tel Aviv Sourasky Medical Center, Tel-Aviv, Israel (S.B-S.);TelAviv University, Tel-Aviv, Israel (D.A.Z.) Received July 31, 2012; accepted for publication June 17, 2013 Address for correspondence and reprints: Esther Ben-Itzchak, The Autism Center, Assaf Harofeh Medical Center, Zerifin 70300, Israel. E-mail: [email protected] Published online in Wiley Online Library (wileyonlinelibrary.com) DOI: 10.1002/aur.1319 © 2013 International Society for Autism Research, Wiley Periodicals, Inc.

INSAR Autism Research ••: ••–••, 2013 1 comparing to boys especially in the communication The study included 690 participants who underwent a domain [Carter et al., 2007; Hartley & Sikora, 2009], full diagnosis for ASD during the years 2002–2011. Par- while others reported that boys had more severe symp- ticipants younger than 18 months (n = 11) or with diag- toms especially in repetitive stereotyped behaviors noses of specific syndromes (n = 16) were excluded from [Lord, Schopler, & Revicki, 1982; Mandy et al., 2012; the study. The final cohort included 663 participants (577 McLennan, Lord, & Schopler, 1993; Szatmari et al., 2012]. males, 86 females) within the age range of 18 months to Others found no sex differences in autism symptom 15 years. (M = 44.1 m. ± 28.7) diagnosed with ASD. Of severity [Holtmann et al., 2007; Pilowsky, Yirmiya, the cohort, 536 participants (80.8%) were 18 months to 5 Shulman, & Dover, 1998; Zwaigenbaum et al., 2012]. years old, 106 (15.9%) were between 5 and 10 years old, Earlier studies, especially those conducted prior to the use and 21 (3.3%) were older than 10 years. This research was of DSM-IV, found that females had higher rates of intel- approved by the institutional review board of the medical lectual compared with boys [Lord & Schopler, center as required. 1985; Volkmar, Szatmari, & Sparrow, 1993]. However several later studies did not find significant sex differ- Procedure ences [Mandy et al., 2012; Zwaigenbaum et al., 2012]. Dworzynski, Ronald, Bolton, and Happé [2012] reported Participants were referred to the autism center for a com- that girls, but not boys, meeting criteria for ASD showed prehensive assessment of a possible diagnosis of ASD. The more additional problems, such as low intellectual level evaluation included a neurological assessment, and and behavioral difficulties. Researchers suggested sex bias behavioral, cognitive, and functional evaluations. Assess- in the diagnosis of ASD. In the absence of additional ments were conducted by a skilled interdisciplinary team. intellectual or behavioral problems, girls are less likely Pediatric neurologists obtained medical, developmental, than boys to meet diagnostic criteria for ASD at equiva- and familial histories from the parents, and conducted a lently high levels of autistic-like traits. comprehensive of all the Only a few studies have examined the association participants. between M:F ratio and specific neurological phenotypes The diagnosis of autism/ASD was obtained by using in ASD. For example, Miles, Hadden, Takahashi, and two standardized tests: the Autism Diagnosis Interview- Hillman [2000] tested 137 individuals with autism, and Revised (ADI-R) [Rutter, Le Couteur, & Lord, 2003] and found no significant difference in sex ratio between the Autism Diagnosis Observation Schedule (ADOS) normocephalic (M:F 6.9:1) and macrocephalic (M:F 4.3:1) [Lord, Rutter, DiLavore, & Risi, 1999], and by meeting individuals. However, they did find that sex ratio was criteria for autism/ASD based on DSM-IV [American closer to normal (M:F 1.5:1) in individuals with micro- Psychiatric Association, 1994]. All the professionals cephaly. Epilepsy in ASD was associated with sex (female) involved in the diagnostic process established reliability [Amiet et al., 2008; Steffenburg & Gillberg, 1986; as required. Cognitive and developmental abilities (IQ/ Tuchman, Rapin, & Shinnar, 1991] and with intellectual DQ) were assessed using the Mullen Scales of Early Learn- disability [Amiet et al., 2008; Bolton et al., 2011]. ing (Mullen, 1995) , Bayley Scales of Infant Development However, the M:F ratio has not been specifically exam- (Bayley, 1993), Wechsler Preschool and Primary Scale of ined in the subpopulation of ASD and epilepsy. Further- Intelligence (Wechsler, 1989), Stanford–Binet Intelligence more, previous studies have not described M:F ratio Scales (Thorndike, Hagen, & Sattler, 1986), Kaufman differences in ASD with developmental regression [Parr Assessment Battery for Children-II (Kaufman, 1983), or et al., 2011; Shumway et al., 2011]. Wechsler Intelligence Scale for Children III and IV As ASD occurs more frequently in males, we hypoth- (Wechsler, 2003), according to the child’s age and lan- esized that the relatively small group of females with ASD guage level. Since 2006, developmental and cognitive may have a different clinical expression than males. In tests have been obtained, and therefore DQ/IQ scores this study, we aimed at examining the prevalence of were available on about half of the ASD cohort (n = 323). females compared with males in several known ASD- Adaptive skills were assessed using the Vineland Adap- related neurological phenotypes. The results of this study tive Behavior Scales (VABS) [Sparrow, Balla, & Cicchetti, may provide better characterization of ASD in females, 1984]. Since 2004, VABS assessments have been added to and insights regarding ASD etiology. each evaluation of ASD, and therefore VABS scores were available only for 528 participants. Methods Head circumference measurements were performed Participants using standard methods [Deutsch & Farkas, 1994; Deutsch & Joseph, 2003] by a senior child neurologist, The study was conducted at a tertiary autism center that and were plotted on normative head circumference provides diagnosis and treatment services, and is growth charts and converted to percentile values involved in research in the field of ASD. [Nellhaus, 1968]. Head circumference data were available

2 Ben-Itzchak et al./Specific phenotypes in females with ASD INSAR on 581 participants. Children with ASD were classified 16.4, SD = 10.8), communication (M = 10.7, SD = 3.9), further according to whether or not their parents had and restricted, repetitive, and stereotyped patterns of reported a history of loss of skills (developmental regres- behaviors (M = 4.8, SD = 2.4); ADOS severity scale scores sion) on the ADI-R. The definition of loss within the (M = 7.4, SD = 2.1); adaptive skill scores according to the ADI-R is twofold: it requires that any loss be coded only if VABS (M = 69.6, SD = 11.1); and developmental/cognitive the skill was established initially for at least 3 months, abilities measured (M = 78.4, SD = 21.7). The M:F ratio in and the loss of skill must have lasted at least 3 months. In the ASD group was 6.7:1, showing a significant male this study, the occurrence of regression was explored preponderance. The severity of autism symptoms, the using the coding of definite loss (score = 2) of specified level of adaptive functioning, and the developmental/ skills in the language, social engagement, constructive or cognitive ability were compared between males and imaginary play, and motor skills in the ADI-R. Minor females to control for the possible impact of differences neurological and musculoskeletal dysfunction (MNMD) in these domains on the analyses of the specific neuro- referred to findings of abnormal neurological and mus- logical phenotypes. No significant differences were noted culoskeletal findings that were detected during a compre- between the percentage of verbal males (39.6%) and hensive neurological examination, without the use of females (37.2%), and therefore ADI scores between the standardized procedures, by two pediatric neurologists. sexes could be compared. Table 1 summarizes the charac- These findings included muscle tone (hypotonia, hyper- teristics of males and females in our cohort. No signifi- tonia), deep tendon reflexes (weak, exaggerated), cere- cant differences were noted between males and females in bellar dysfunction (Romberg test, Finger-to-nose test, all the above examined measures. Heel-to-shin test, diadochokinesis), and hyperlaxity of Of the male participants, 47.1% received a diagnosis of joints, not related to a major neurological disorder. Data ASD and 52.9% received a diagnosis of autism. Of the were available on 616 participants with ASD. Seizures female participants, 52.3% received a diagnosis of ASD were defined as one or more episodes of idiopathic sei- and 47.7% received a diagnosis of autism. The distribu- zures not related to a febrile illness or a traumatic event. tion of autism diagnostic categories, autism, and ASD did Documentation of seizures was based on the neurological not differ between males and females [χ2(1) = 1.2, assessment by one of the certified pediatric neurologist at P > 0.05]. the autism center, previous reports from the neurologist The overall mean head circumference centile in the who made the diagnosis of seizures, and on the results ASD group was 49.2% ± 26.4. We divided the ASD group of the Electroencephalogram recording if available. An into two subgroups: one with HC >3rd percentile abnormal EEG recording was not required for a child to (n = 620) and a microcephaly group with HC ≤3rd per- be considered to have a disorder. Data on seizures centile (n = 39). There were no age differences between were available on 634 participants with ASD. the groups with and without microcephaly. The study revealed that the ASD group had higher representation Data Analysis (5.9%) of microcephaly than the 3% expected in the general population [χ2(1) = 19.3, P < 0.001]. We then To investigate the differences in M:F ratio between the divided the ASD group into two subgroups: one with HC ASD subgroups with and without the specific phenotypes <97th percentile (n = 540) and a macrocephaly group (microcephaly, macrocephaly, developmental regression, MNMD, seizures), chi-square tests were used. To compare the head circumference percentiles of males and females Table 1. Autism Symptoms Severity and Adaptive Functioning in the ASD group while controlling for age, one-way in Males and Females with ASD analysis of covariance was used. To compare the percent- ages of males and females with microcephaly or macro- Males Females cephaly with the expected percentage in the population nM(SD) nM(SD) F (3%), chi-square goodness-of-fit tests were used. This analysis compares between the observed distribution and Age (months) 577 43.8 (29.1) 86 44.9 (26.6) .11 ADOS severity scale 577 7.4 (2.1) 86 7.4 (2.1) .00 the expected distribution of head circumference in the ADI-R general population [Nellhaus, 1968] Social 577 16.3 (11.2) 86 16.9 (7.2) 1.9 Communication 10.8 (3.8) 10.1 (4.2) Repetitive behavior 4.8 (2.4) 4.3 (2.0) Results VABS composite scores 460 69.9 (10.9) 68 67.3 (12.3) 2.9 IQ/DQ 283 78.8 (21.8) 40 75.7 (21.7) .7

The characteristics of the ASD group were examined in ADOS, Autism Diagnosis Observation Schedule; ADI-R, Autism Diagnosis different domains: autism severity according to the ADI-R Interview-Revised; VABS, Vineland Adaptive Behavior Scales; IQ/DQ, cog- algorithm scores in reciprocal social interaction (M = nitive and developmental abilities.

INSAR Ben-Itzchak et al./Specific phenotypes in females with ASD 3 with HC ≥97th percentile (n = 119). There were no age resentation was significantly higher than the expected differences between the groups with and without macro- 3% {males [χ2(1) = 594.6, P < 0.001]; females [χ2(1) = 28.3, cephaly. The study revealed that the ASD group had P < 0.001)]} (Fig. 1). higher representation than expected of macrocephaly Assessing the developmental trajectory (yes/no regres- (18.1%) [χ2(1) = 513.5, P < 0.001]. sion) in the ASD group revealed that 20.4% of the sample Comparing head circumference percentiles in males had a history of regression. There were no age differences and females while controlling for age revealed that between the groups with and without developmental the mean head circumference percentile for males regression. The rate of developmental regression was sig- (50.1 ± 25.6, n = 501) was significantly larger than for nificantly higher for females (30.2%) than for males females (43.4 ± 30.2; n = 80) [F(1, 579) = 6.2, P < 0.01, (18.9%). The M:F ratio in the developmental regression η2 = 0.021]. To examine whether the difference in mean group (4.2:1) was significantly lower than in the non- head size is not a function of having more females at the regressive group (7.8:1) (Table 2). Examining the specific extremes of small head size, we examined the distribution domains of regression revealed that the M:F ratio in the of head size in the female group using skew and kurtosis group with (n = 92) and without language regression did analyses. Values of skew (Z = 1.18) and kurtosis (Z = not differ significantly. However, the M:F ratio (2.5:1) in −1.06) were found in the proper range (−1.96 < Z<+1.96). the group with social regression (n = 67) was significantly Given these results, we further investigated the M:F ratio different from the M:F ratio (7.9:1) in the group without in different head circumference (HC) groups. M:F ratio regression, indicating more females in the social regres- was significantly lower (2.0:1) in the microcephaly group sion group [χ2(1) = 15.5, P < 0.001] (Fig. 2). compared with the >3rd percentile group (7.5:1). These Of the ASD group, 59.3% of the participants had minor results indicate an increased rate of microcephaly among neurological deficits (MNMD). Hyperlaxity of joints females with ASD compared with males (Table 2). We (52.5%) and hypotonia (26.3%) were the most frequent then compared the M:F ratio in the ≥97th percentile abnormalities, followed by abnormal deep tendon reflex (defined as macrocephaly) group (9.8:1) with the <97th (21.1.0%) and cerebellar dysfunction (18.2%), while percentile group (6.2:1) in the ASD sample. The differ- hypertonia (1.6%) was infrequently found. The group ences in M:F ratio in these two groups were not statisti- with MNMD was older (M = 46.5 ± 30.6 months) than the cally significant (Table 2). group without MNMD (M = 41.1 ± 25.9) [F(1, 615) = 5.4, To examine the prevalence of macro- and microcephaly P < 0.05, η2 = 0.009]. Looking at the sex distribution, sig- in males and females separately, the differences between nificantly higher rates of females (73.8%) in comparison the observed and expected HC in the general population to males (57.1%) had MNMD. Calculating the M:F ratio, were compared using χ2 goodness-of-fit test. Micro- a significantly lower ratio was found for the group with cephaly was noted in 4.5% of the males in the ASD group MNMD (5.2:1) compared with the group without MNMD and was significantly higher than the 3% expected by (11:1) (Table 2). definition [χ2(1) = 4.3, P < 0.05]. For females, the rate of Of the ASD group, 5.4% were reported to have seizures, microcephaly was five times more pronounced than significantly more than the 1% expected in the general expected, with 15.1% of the female group having HC population [χ2(1) = 57.0, P < 0.001]. In general, the group ≤3rd percentile [χ2(1) = 42.1, P < 0.001]. Macrocephaly with seizures was older (M = 64.3 ± 40.2 months) than the was more frequent in males (18.8%) than in females group without seizures (M = 43.4 ± 27.9) [F(1, 633) = 17.1, (12.8%), but this difference was not statistically signifi- P < 0.001, η2 = 0.026]. The rate of seizures among females cant. Among both males and females, macrocephaly rep- with ASD was 9.5%, which was twice as high as the rate of

Table 2. Distribution of Neurological Phenotypes among Males and Females in the ASD Cohort and M:F ratio for the Groups with and without the Neurological Phenotypes

Males Female Phenotype+ Phenotype+ M:F M:F Phenotype N % N % Phenotype+ Phenotype−χ2(1)

Microcephalic 26 4.5 13 15.1 2.0:1 7.5:1 15.0*** Macrocephalic 108 18.8 11 12.8 9.8:1 6.2:1 1.9 Seizures 26 4.7 8 9.5 3.3:1 6.9:1 3.3∧ Developmental regression 109 18.9 26 30.2 4.2:1 7.8:1 5.9* MNMD 306 57.1 59 73.8 5.2:1 11.0:1 8.0***

ASD, autism spectrum disorder; MNMD, minor neurological and musculoskeletal deficits. ***P < .001, *P < .05, ∧P < .1.

4 Ben-Itzchak et al./Specific phenotypes in females with ASD INSAR Figure 1. Percentage of head circumference percentiles (≤3%; ≥97%) in males and females with autism spectrum disorder. The horizontal line represents the expected three centiles.

Figure 2. Percentage of developmental regression in language, socialization, and motor skills in males and females with autism spectrum disorder. ***P < .001.

INSAR Ben-Itzchak et al./Specific phenotypes in females with ASD 5 4.7% in the male population. The M:F ratio in the group relatively more representation of females. We did not with a history of seizures was lower (3.3:1) than the sex observe the same pattern in the group with macro- ratio of the ASD group without seizures (6.9:1), showing a cephaly, where the M:F ratio was not significantly differ- trend for significant differences (Table 2). Since the peaks ent from the rest of the ASD group. Similar findings of seizures in ASD may occur after the age of 4, the regarding the role of sex in ASD with microcephaly were percentage of seizures was investigated in this age group previously reported in a relatively small cohort [Miles separately and was found to be higher (10.2%) than for the et al., 2000]. Previous studies reported increased rate of entire group. However, the M:F ratio for this age range for microcephaly in ASD [Miles et al., 2005, −5%; Fombonne groups with and without seizures was not different from et al., 1999, −15.1%], but only Miles et al. [2005] reported the findings in the entire ASD group. more female representation in this group. Microcephaly Overall, 24.3% of ASD group (n = 161) did not have any is, in many instances, a sign of abnormal brain develop- of the examined phenotypes, 50.7% (n = 336) had a ment, seen in multiple genetic and nongenetic condi- single neurological phenotype, and 25.1% (n = 166) had tions. Our finding of a higher proportion of microcephaly two or more of the examined phenotypes. M:F ratio sig- in females compared with males is highly interesting as nificantly decreased as the number of the examined phe- ASD is usually associated with enlarged head circumfer- notypes increased. M:F ratio for the ASD group without ence. These results suggest that a subgroup of ASD with phenotypes was 13.6:1, for the ASD group with one phe- microcephaly is more prevalent in females and might notype it was 7.6:1, and for the ASD group with two or point to a unique etiology for ASD in this group. more phenotypes it was 3.6:1. No age differences were We further examined another common clinical condi- noted between the groups with and without the neuro- tion unique for ASD, developmental regression, noted in logical phenotypes. 15–30% of cases with autism in different studies [Parr et al., 2011; Shumway et al., 2011; Stefanatos, 2008]. In our ASD group, about 20% had a history of regression, Discussion which is in accordance with previous reports. The phe- nomenon of regression was significantly more common ASD is a heterogeneous group, associated with multiple in females than in males. This pattern was especially clinical presentations and etiologies. ASD is considered noted in social development regression, but no differ- one of the most common neurodevelopmental disorders, ences in sex ratio were found for language development affecting about 1% of the population. Although ASD has regression. Currently, there is no evidence for clear patho- been proven to have a major genetic component, only genetic mechanisms that can explain developmental about 20–30% of cases have a known etiology [Schaaf & regression in ASD. Genetic and environmental factors Zoghbi, 2011]. During the last few years, a vast effort has have been suggested as a possible etiology for regression been made to assess biological and medical conditions in ASD. Interestingly, one of the known that co-occur frequently in ASD, and can shed light on that is associated with developmental regression and the pathogenesis of ASD. The present study, conducted symptoms of autism among females is Rett syndrome, on a large series of children with ASD, found a decreased which is caused by in the MeCP2 gene. The M:F ratio, with relatively more females in the majority of high prevalence of developmental regression in females clinical phenotypes that are associated with autism. in our study strengthens again the possible contribution Accelerated head growth in early childhood resulting of unique genetic or neurological processes in females in relatively enlarged head circumference has been with ASD, which are less prominent in affected males. reported in numerous studies. It is hypothesized that the We further looked at brain abnormalities in ASD as brain in children with autism undergoes an abnormal manifested in MNMD found during the medical exami- growth trajectory that includes a period of early over- nation. High rates of joint hyperlaxity, hypotonia, abnor- growth. This theory is supported by and mal DTR, and cerebellar dysfunction were documented, neuropathological findings [Courchesne, Carper, & suggesting involvement of various parts of the brain in Akshoomoff, 2003; Courchesne et al., 2007; Mraz, Green, ASD. These findings are in accordance with recent studies Dumont-Mathieu, Makin, & Fein, 2007; Webb et al., that documented frequent specific dysfunctions in ASD 2007]. In general, our study found that the representation (95%) in comparison to the general population (20%) [De of both sexes in the macro- and microcephalic groups was Jong et al., 2011]. The group with MNMD was older than higher than the 3% expected in the general population the group without MNMD, which may be explained by (18.1% and 5.9% respectively). Regarding sex differences the increased minor neurological signs, i.e. cerebellar dys- in head circumference, we show for the first time that the function, with age. About three quarters of the females average head circumference in females with ASD is and half of the males had neurological dysfunctions, smaller than in males. Moreover, among the group with adding to our previous findings that point to more exten- microcephaly, the M:F ratio was significantly lower, with sive involvement of brain dysfunction in females.

6 Ben-Itzchak et al./Specific phenotypes in females with ASD INSAR A history of seizures was documented in about 5.4% of reported that girls with ASD tend to have a lower mean IQ the entire ASD population and in about 10% for children and higher rates of severe compared above the age of 4 in our study. These rates are nearly with boys [Lord & Schopler, 1985; Volkmar et al., 1993]. 5–10 times more than the reported rate (1%) in the However, several recent studies that focused on very general population. Higher rates of epilepsy have long young children diagnosed with ASD reported less pro- been reported in ASD, with prevalence estimates that nounced or no sex differences in cognitive abilities vary from 5% to 46% [Kagan-Kushnir et al., 2005; [Carter et al., 2007; Hartley & Sikora, 2009; Mandy et al., Tuchman & Rapin, 2002]. Our finding of a seizure rate 2012; Zwaigenbaum et al., 2012]. The differences lying in the lower range in ASD may stem from the age of between the results of earlier vs. more recent studies the studied population and the type of referrals. In our might be related to the use of different diagnostic mea- study, the ASD population included mostly young chil- sures, before and after the implementation of Interna- dren (mean age = 44.1 months); therefore, the described tional Classification of (ICD)-10 and DSM-IV second peak of epilepsy among individuals with ASD that diagnostic criteria. Several studies assessing autism appears during mid-adolescence was not reflected in our severity using standardized tests reported no sex population. The finding that the ASD group with a differences [Holtmann et al., 2007; Pilowsky et al., 1998; history of seizures was older than the group without Zwaigenbaum et al., 2012]. Other studies did report sex seizures supports the above explanation. In addition, the differences in autism severity, but with conflicting result; tertiary autism center receives referrals for diagnosis of some reported more severe symptoms in males [Lord possible ASD in a heterogenic population with general et al., 1982; McLennan et al., 1993], while others developmental problems, while most of the studies on reported more severe presentation in females [Carter autism and epilepsy came from specialty and et al., 2007; Hartley & Sikora, 2009]. In several studies, epilepsy clinics that assess populations with seizures as females had less repetitive stereotyped behavior com- the main problem [Hughes & Melyn, 2005; Rossi, pared with males with similar levels of social and com- Parmeggiani, Bach, Santucci, & Visconti, 1995; Tuchman munication disorders [Mandy et al., 2012; Szatmari et al., & Rapin, 2002], and thus are at a higher risk for ascer- 2012]. tainment bias regarding the prevalence of seizures. In the Social-communication impairments and stereotyped current study, the history of seizures was about twice behaviors characterize, by definition, all the individuals more frequent in females than in males with ASD. with ASD, both males and females. However, ASD is com- However, these differences were almost statistically sig- monly associated with other phenotypes. Miles et al. nificant, which may be explained by the small number of [2005] divided autism to essential and complex individuals with seizures. A trend of a greater frequency groups; the latter included children with significant of seizures in females with ASD compared with males dysmorphology and microcephaly. The “complex autism was described in previous studies [Bolton et al., 2011; group” had poorer outcome, with lower IQ, more seizures Steffenburg & Gillberg, 1986; Tuchman et al., 1991]. and abnormal EEG, more brain magnetic resonance Amiet et al. [2008], in their meta-analysis on epilepsy in imaging (MRI) abnormalities, and relatively more females females with autism, reported that the pooled prevalence than the “essential autism” group. In the current study, of epilepsy in all the studies they reviewed was of 34.5% we examined specific neurological phenotypes, including in females vs. 18.5% in males. The M:F ratio of autism microcephaly, in ASD in a sex-specific manner, in a large, comorbid with epilepsy was close to 2:1, whereas M:F well-characterized population with ASD. We showed that ratio of autism without epilepsy was 3.5:1. the abnormal neurological phenotypes are consistently Overall, a minority of the ASD population had none of associated with female sex. the ASD-related phenotypes. However, sex ratio in this This study has some limitations, including the lack of group revealed extreme male predominance (13.6:1) that complete DQ/IQ data on all the participants, and the use might represent idiopathic ASD. In contrast, the trend of comprehensive neurological examination but not a toward normalization of M:F ratio when one or more of standardized protocol for assessment of MNMD. the examined phenotypes exist may suggest different eti- The convergence of our findings indicates that, in ologies for this subgroup. These findings complement the addition to the core symptoms, females with ASD study conclusions that secondary complex etiologies are present with a higher prevalence of clinical and neuro- more associated with female ASD. logical manifestations compared with males. The Our study found no sex differences in autism severity, investigated neurological phenotypes, microcephaly, as well as cognitive and adaptive skills, in a large ASD developmental regression, and other neurological signs population. Therefore, sex differences found in the rep- are likely associated with more extensive neurological resentation of neurological phenotypes in this study deficits. It is possible, therefore, that whereas “pro- cannot be explained by differences in these behavioral, totypic” ASD is typically seen in males, the ASD mani- cognitive, and functional domains. Earlier studies festation in many females may result from a different

INSAR Ben-Itzchak et al./Specific phenotypes in females with ASD 7 disorder associated with more extensive brain dysfunc- frontal pyramidal and interneuron development and tion. It might be that females are more protected than connectivity. International Journal of Developmental Neuro- males from “prototypical” ASD [Levy et al., 2011], and science, 23, 153–170. therefore a higher prevalence of ASD that is a secondary Courchesne, E., Pierce, K., Schumann, C.M., Redcay, E., Buckwalter, J.A., et al. (2007). Mapping early brain develop- manifestation of other disorders is found in affected ment in autism. Neuron, 56, 399–413. females. De Jong, M.D., Punt, M., De Groot, E., Minderaa, R.B., & Recent advances in genetic testing, such as the current Hadders-Algra, M. (2011). Minor neurological dysfunction in extensive use of microarray techniques, may reveal new children with autism spectrum disorder. Developmental microdeletions/duplications in DNA that are responsible Medicine and Child Neurology, 53, 641–646. for ASD with complex neurological phenotypes. Such Deutsch, C.K., & Farkas, L.G. (1994). Quantitative methods of genetic aberrations affect both sexes equally resulting in dysmorphology diagnosis. In L.G. Farsak (Ed.), Anthropom- more female representation in this subgroup. etry of the head and face (pp. 151–158). New York: Raven. These compelling results should alert clinicians when Deutsch, C.K., & Joseph, R.M. (2003). Brief report: Cognitive assessing a female with ASD to conduct a more compre- correlates of enlarged head circumference in children with hensive neurological assessment, and to consider brain autism. Journal of Autism and Developmental Disorders, 33, imaging and EEG as part of the evaluation. 209–215. Dworzynski, K., Ronald, A., Bolton, P., & Happé, F. (2012). How The proposed hypothesis of a “complex ASD pheno- different are girls and boys above and below the diagnostic type in females” that involves extensive brain dysfunc- threshold for autism spectrum disorders? Journal of the tion should be further analyzed by other methods and American Academy of Child and Adolescents , 51, approaches, such as a functional MRI and a comprehen- 788–797. sive genetic evaluation using advanced methods. Further Fombonne, E. (2003). The prevalence of autism. Journal of the molecular investigation of the nature of ASD with American Medical Association, 289, 87–89. complex neurological phenotypes is required to improve Fombonne, E., Roge, B., Claverie, J., Courty, S., & Fremolle, J. our understanding of the mechanisms associated with (1999). Microcephaly and macrocephaly in autism. Journal of ASD, and might show better success than that seen in the Autism and Developmental Disorders, 29, 113–119. extensive studies performed on the general ASD popula- Hartley, S.L., & Sikora, D.M. (2009). Sex differences in autism tion, which predominantly consists of males. spectrum disorder: An examination of developmental func- tioning, autistic symptoms, and coexisting behavior prob- lems in toddlers. Journal of Autism and Developmental References Disorders, 39, 1715–1722. Holtmann, M., Bolte, S., & Poustka, F. (2007). Autism spectrum American Psychiatric Association. (1994). Diagnostic and statis- disorders: Sex differences in autistic behavior in autistic tical manual of mental disorders, fourth edition (DSM-IV). behavior domains and coexisting psychopathology. Develop- Washington, DC: American Psychiatric Association. mental Medicine and Child Neurology, 49, 361–366. Amiet, C., Gourfinkel-An, I., Bouzamondo, A., Tordjman, S., Hughes, J.R., & Melyn, M. (2005). EEG and seizures in autistic Baulac, M., et al. (2008). Epilepsy in autism is associated with children and adolescents: Further findings with therapeutic intellectual disability and gender: Evidence from a meta- implications. Clinical EEG and Neuroscience, 36, 15–20. analysis. , 64, 577–582. Kagan-Kushnir, T., Roberts, S.W., & Snead, O.C. (2005). Screen- Baird, G., Charman, T., Pickles, A., Chandler, S., Loucas, T., et al. ing electroencephalograms in autism spectrum disorders: (2008). Regression, developmental trajectory and associated Evidence-based guideline. Journal of Child Neurology, 20, problems in disorders in the autism spectrum: The SNAP 197–206. study. Journal of Autism and Developmental Disorders, 38, Kaufman, A.S. (1983). Kaufman Assessment Battery for Children 1827–1836. (K-ABC). Circle Pines, MN: American Guidance Services. Bayley, N. (1993). Bayley scales for infant development – Second Levy, D., Ronemus, M., Yamrom, B., Lee, Y.H., Leotta, A., et al. edition manual. San Antonio, TX: Psychological Corporation. (2011). Rare de novo and transmitted copy-number-variation Bolton, P.F., Carcani-Rathwell, I., Hutton, J., Goode, S., Howlin, in autistic spectrum disorders. Neuron, 70, 886–897. P., & Rutter, M. (2011). Epilepsy in autism: Features and Lord, C., Rutter, M., DiLavore, P.C., & Risi, S. (1999). Autism correlates. British Journal of Psychiatry, 198, 289–294. Diagnostic Observation Schedule – Generic. Los Angeles, CA: Carter, A.S., Black, D.O., Tewani, S., Connolly, C.E., Kadlec, M.B., Western Psychological Services. & Tager-Flusberg, H. (2007). Sex differences in toddlers with Lord, C., & Schopler, E. (1985). Differences in sex ratios in autism autism spectrum disorders. Journal of Autism and Develop- as a function of measured intelligence. Journal of Autism and mental Disorders, 37, 86–97. Developmental Disorders, 15, 185–193. Courchesne, E., Carper, R., & Akshoomoff, N. (2003). Evidence of Lord, C., Schopler, E., & Revicki, D. (1982). Sex differences in brain overgrowth in the first year of life in autism. Journal of autism. Journal of Autism and Developmental Disorders, 12, the American Medical Association, 290, 337–344. 317–330. Courchesne, E., & Pierce, K. (2005). Brain overgrowth in autism Mandy, H., Chilvers, R., Chowdhury, U., Salter, G., Seigal, A., & during a critical time in development: Implications for Skuse, D. (2012). Sex differences in autism spectrum disorder:

8 Ben-Itzchak et al./Specific phenotypes in females with ASD INSAR Evidence from a large sample of children and adolescents. Schaaf, C.P., & Zoghbi, H.Y. (2011). Solving the autism puzzle a Journal of Autism and Developmental Disorders, 42, 1304– few pieces at a time. Neuron, 70, 806–808. 1313. Shumway, S., Thurm, A., Swedo, S.E., Deprey, L., Barnett, L.A., McLennan, J.D., Lord, C., & Schopler, E. (1993). Sex differences et al. (2011). Brief report: Symptom onset patterns and func- in higher functioning people with autism. Journal of Autism tional outcomes in young children with autism spectrum and Developmental Disorders, 23, 217–227. disorders. Journal of Autism and Developmental Disorders, Miles, J.H., Hadden, T.N., Takahashi, T.N., & Hillman, R.E. 41, 1727–1732. (2000). Head circumference is an independent clinical Sparrow, S.S., Balla, D.A., & Cicchetti, D.V. (1984). Vineland finding associated with autism. American Journal of Medical Adaptive Behavior Scales: Interview Edition, Survey Form Genetics, 95, 339–350. Manual. Circle Pines, MN: American Guidance Service. Miles, J.H., Takahashi, T.N., Bagby, S., Sahota, P.K., Vaslow, D.F., Stefanatos, G.A. (2008). Regression in autistic spectrum disor- et al. (2005). Essential versus complex autism: Definition of ders. Review, 18, 305–319. fundamental prognostic subtypes. American Journal of Steffenburg, S., & Gillberg, C. (1986). Autism and autistic-like Medical Genetics, 135A, 171–180. conditions in Swedish rural and urban areas: A population Ming, X., Brimacombe, M., & Wagner, G.C. (2007). Prevalence of study. British Journal of Psychiatry, 149, 81–87. motor impairment in autism spectrum disorders. Brain Devel- Szatmari, P., Liu, X.Q., Goldberg, J., Zwaigenbaum, L., Paterson, opment, 29, 565–570. A.D., et al. (2012). Sex differences in repetitive stereotyped Minshew, N.J., & Williams, D.L. (2007). The new neurobiology behaviors in autism: Implications for genetic liability. Ameri- of autism: Cortex, connectivity, and neuronal organization. can Journal of Medical Genetics, 159B, 5–12. Archives of Neurology, 64, 945–950. Thorndike, R.M., Hagen, E.P., & Sattler, J.M. (1986). Stanford- Mraz, K.D., Green, J., Dumont-Mathieu, T., Makin, S., & Fein, D. Binet Intelligence Scale (4th ed.). Itasca, IL: Riverside (2007). Correlates of head circumference growth in infants Publishing. later diagnosed with autism spectrum disorders. Journal of Tuchman, R.F., & Rapin, I. (2002). Epilepsy in autism. Lancet Child Neurology, 22, 700–713. Neurology, 1, 352–358. Mullen, E. (1995). The Mullen Scales of Early Learning. Circle Tuchman, R.F., Rapin, I., & Shinnar, S. (1991). Autistic and dys- Pines, MN: American Guidance Service. phasic children. II: Epilepsy. Pediatrics, 88, 1219–1225. Nellhaus, G. (1968). Head circumference from birth to eighteen Volkmar, F.R., Szatmari, P., & Sparrow, S.S. (1993). Sex differences years. Practical composite international and interracial in pervasive developmental disorders. Journal of Autism and graphs. Pediatrics, 41, 106–114. Developmental Disorders, 23, 579–591. Newschaffer, C.J., Croen, L.A., Daniels, J., Giarelli, E., Grether, Webb, S.J., Nalty, T., Munson, J., Brock, C., Abbott, R., & J.K., et al. (2007). The epidemiology of autism spectrum dis- Dawson, G. (2007). Rate of head circumference growth as a orders. Annual Review of Public Health, 28, 235–258. function of autism diagnosis and history of autistic regres- Parr, J.R., Le Couteur, A., Baird, G., Rutter, M., Pickles, A., et al. sion. Journal of Child Neurology, 22, 182–1190. (2011). Early developmental regression in autism spectrum Wechsler, D. (1989). Wechsler Preschool and Primary Scale disorder: Evidence from an international multiplex sample. of Intelligence-Revised. New York: The Psychological Journal of Autism and Developmental Disorders, 41, 332– Corporation. 340. Wechsler, D. (2003). Wechsler Intelligence Scale for Children, Pilowsky, T., Yirmiya, N., Shulman, C., & Dover, R. (1998). The fourth edition. San Antonio, TX: The Psychological Autism Diagnostic Interview-Revised and the Childhood Corporation. Autism Rating Scale: Differences between diagnostic systems Zachor, D.A., Ilanit, Z., & Ben Itzchak, E. (2010). Autism severity and comparison between genders. Journal of Autism and and motor abilities correlates of imitation situations in chil- Developmental Disorders, 28, 143–151. dren with autism spectrum disorders. Research in Autism Rossi, P.G., Parmeggiani, A., Bach, V., Santucci, M., & Visconti, P. Spectrum Disorder, 4, 438–443. (1995). EEG features and epilepsy in patients with autism. Zwaigenbaum, L., Bryson, S.E., Szatmari, P., Brian, J., Smith, I.M., Brain Development, 17, 169–174. et al. (2012). Sex differences in children with autism spec- Rutter, M., Le Couteur, A., & Lord, C. (2003). The Autism Diag- trum disorder identified within a high-risk infant cohort. nostic Interview-Revised (ADI-R). Los Angeles, CA: Western Journal of Autism and Developmental Disorders, 42, 2585– Psychological Services. 2596.

INSAR Ben-Itzchak et al./Specific phenotypes in females with ASD 9