Age- and sex-related emotional and behavioral problems in children with autism spectrum disorders: Comparison with control children

Authors

  • Fumie Horiuchi MD,

    Corresponding author
    1. Department of Neuropsychiatry and Neuroscience, Ehime University Graduate School of Medicine, Toon, Japan
    • Correspondence: Fumie Horiuchi, MD, Department of Neuropsychiatry and Neuroscience, Ehime University Graduate School of Medicine, Shitsukawa, Toon-city, Ehime 791-0295, Japan. Email: matsufu@m.ehime-u.ac.jp

    Search for more papers by this author
  • Yasunori Oka MD, PhD,

    1. Center for Sleep Medicine, Ehime University Graduate School of Medicine, Toon, Japan
    Search for more papers by this author
  • Hiroyuki Uno PhD,

    1. Department of Education for Disabled Children, Hyogo University of Teacher Education, Kato-gun, Japan
    Search for more papers by this author
  • Kentaro Kawabe MD,

    1. Department of Neuropsychiatry and Neuroscience, Ehime University Graduate School of Medicine, Toon, Japan
    Search for more papers by this author
  • Fumi Okada MA,

    1. Department of Neuropsychiatry and Neuroscience, Ehime University Graduate School of Medicine, Toon, Japan
    2. Center for Developmental Clinical Psychology and Psychiatry, Nagoya University, Nagoya, Japan
    Search for more papers by this author
  • Isao Saito MD, PhD,

    1. Department of Public Health, Ehime University Graduate School of Medicine, Toon, Japan
    2. Department of Basic Nursing and Health Science, Ehime University Graduate School of Medicine, Toon, Japan
    Search for more papers by this author
  • Takeshi Tanigawa MD, PhD,

    1. Department of Public Health, Ehime University Graduate School of Medicine, Toon, Japan
    Search for more papers by this author
  • Shu-ichi Ueno MD, PhD

    1. Department of Neuropsychiatry and Neuroscience, Ehime University Graduate School of Medicine, Toon, Japan
    Search for more papers by this author

Abstract

Aim

Children with autism spectrum disorders (ASD) often present with emotional and behavioral problems, which could change the clinical course, especially during childhood, and affect future quality of life. The aim of this study was to clarify the age- and sex-related differences of these problems in ASD.

Methods

The study subjects were 173 patients with ASD (age: 4–16 years) and 173 age- and sex-matched community children (control group). The parent version of the Strengths and Difficulties Questionnaire was used for comparison of the emotional and behavioral problems between the two groups.

Results

The Strengths and Difficulties Questionnaire scores were significantly higher in children with ASD than controls at all ages. The score of total difficulties was significantly higher in girls with ASD than in boys, while the score in male controls was significantly higher than in female controls. Age-related differences in emotional and behavioral problems were observed both in children with ASD and controls, but the characteristics were different: in children with ASD, emotional symptoms and peer problems in both sexes and conduct problems in girls increased significantly with age, while none of the problems in the controls changed with age except for a decrease in the score of hyperactivity/inattention developmentally in both sexes. Prosocial behaviors of children with ASD and controls showed small changes with age.

Conclusion

Emotional and behavioral problems are common in children with ASD and showed age- and sex-related differences. Our study emphasizes the importance of recognizing those differences among children with ASD for early intervention.

Autism spectrum disorders (ASD) are characterized by difficulties in reciprocal social interaction skills, deficits in communication skills, stereotypic, obsessive, or repetitive behaviors, and restricted patterns of interests and activities.[1] Prevalence of autism and related ASD is up to 1% among children and adolescents, which is greater than previously recognized.[2]

Emotional and behavioral problems are common in children with ASD.[3] These problems often develop into comorbid psychiatric disorders, such as anxiety disorder, attention-deficit/hyperactivity disorder (ADHD), and oppositional defiant disorder (ODD). However, identifying emotional and behavioral problems that require specific intervention is often difficult. One reason for the difficulty is that children with ASD cannot provide sufficient information for clinicians to identify the presence of co-existent emotional and behavioral problems because of their impairments in social communication. The second reason is that the developmental trajectories of ASD may affect the symptom topography of emotional and behavioral problems that have been observed in the general population. The third reason is that emotional and behavioral problems of children with ASD change with age and are different among male and female patients.[4] However, there is little information on the extent of these problems in girls and boys with ASD, or the effect of age on emotional and behavioral problems. To determine the age-related differences in emotional and behavioral problems in children with ASD, comparison with that of control children is necessary.

Failure to detect co-existent emotional and behavioral problems, rather than ASD-related symptoms, can forestall specific intervention and increase the risk of further functional impairment. In fact, the presence of emotional and behavioral problems is associated with poor clinical outcome.[5] Thus, early detection of such problems is important to avoid suffering of the patients and their families.[6] It is vital to have a screening instrument to detect emotional and behavioral problems in children with ASD for early intervention.

Several rating scales are available to detect ASD, such as the Modified Checklist for Autism in Toddlers (M-CHAT),[7] the Social Communication Questionnaire (SCQ),[8] and the Social Responsiveness Scale (SRS).[9] However, there are only a few screening measures of emotional and behavioral problems designed specifically for children with ASD. The Strengths and Difficulties Questionnaire (SDQ) is a screening instrument of emotional and behavioral problems in children aged 4–16 years and their worldwide application for screening, clinical, and research purposes has been facilitated by the availability of officially authorized translated versions in over 60 languages.[10] The SDQ might be useful for screening emotional and behavioral problems in children with developmental disorders, including ASD.

The aims of this study were: (i) to identify any modifiable factors associated with emotional and behavioral problems in 4–16-year-old children; (ii) to delineate sex differences in emotional and behavioral problems by comparing SDQ scores in children with ASD and control children; and (iii) to characterize the age-related differences of emotional and behavioral problems in children with ASD in comparison with control children.

Methods

Participants and instruments

ASD group

The ASD group consisted of 173 children with ASD, including 84 children aged 4–6 years, 32 children aged 7–9 years, 38 children aged 10–12 years, and 19 children aged 13–16 years (Table 1). These children represented all those referred to Ehime University Hospital for assessment of developmental disorders, including ASD, between 2003 and 2012 and were later diagnosed clinically with ASD based on the DSM-IV-TR criteria by board-certified child psychiatrists. None of the children had any perinatal events or uncontrolled seizure disorders.

Table 1. Characteristics of participating subjects
Age (years): Mean (SD)Control groupASD group
7.92 (3.28)7.92 (3.28)
  1. ASD, autism spectrum disorders.
Number of participants (boys/girls)
All173 (129/44)173 (129/44)
4–6 years84 (63/21)84 (63/21)
7–9 years32 (28/4)32 (28/4)
10–12 years38 (24/14)38 (24/14)
13–16 years19 (14/5)19 (14/5)
IQ (n = 148)  
Full IQ: Mean (SD)88.3 (20.1)
IQ range40–132

The parent version of the SDQ was completed by the parents or caregivers. The SDQ describes the child's functioning during the preceding 6 months and is used for the detection of emotional and behavioral problems.[11, 12] It comprises 25 items, organized into the following five subscales: emotional symptoms, conduct problems, hyperactivity/inattention, peer relationship problems, and prosocial behavior. Each item is accompanied by a response option ‘0 = not true’, ‘1 = somewhat true’, and ‘2 = certainly true’. The scores of the 20 items on emotional symptoms, conduct problems, hyperactivity/inattention and peer relationship problems were summed to generate the ‘total difficulties score’, ranging from 0 to 40. The scores of prosocial behavior were summed to generate the ‘strength score’. A high total difficulties score reflected difficulties, whereas a high score for the prosocial behavior subscale reflected strengths.[10] The cut-off values for total and subscale scores of the SDQ were set so that roughly 80% of children were normal, 10% were on the borderline range, and the remaining 10% had clinical range of SDQ. The cut-off values for the English version, including borderline range and clinical range, were used in this study.[10] To assess intellectual functioning, the Wechsler Intelligence Scale for Children-Third Edition (WISC-III) or the Wechsler Intelligence Scale for Children-Fourth Edition (WISC-IV) was used to measure IQ for children aged > 5 years. There were no missing data on the SDQ or WISC among the ASD children.

Control group

The control subjects were recruited from a pool of children from the local community who participated in the Toon Child Study, which was designed to review the lifestyle and behavior of children.[13] The Toon Child Study was conducted in 2009 and all children of Toon city aged 3–18 years (n = 4212) were involved. The total population of Toon city during that year was approximately 35 000, with about 280 births/year. Questionnaires including the SDQ were provided to the parents or primary caregivers. The number of answered questionnaires was 3687, and subjects with any missing data (n = 247) were excluded (the proper response rate: 81.6%). There were no significant differences between subjects with missing data and those without missing data (n = 3440) in terms of sex (χ2 = 0.261, d.f. = 1, P = 0.610) and age (P = 0.814). For this analysis, the subjects were enrolled by stratified two-stage sampling, then 173 age- and sex-matched children were selected.

The study was approved by the Institutional Review Board of Ehime University Graduate School of Medicine. Questionnaires were provided to the parents or primary caregivers, together with written information about the project. A signed informed consent was obtained from the parents/caregivers of each participant.

Statistical analysis

To elucidate the main factor(s) that affect the total difficulties, variables such as age, sex, and IQ were entered into multiple regression analysis. Backwards stepwise regression was performed, and a P-value greater than 0.05 was used for variable removal. The exact test was used to analyze the proportion of individuals who scored ‘clinical range’, ‘borderline range’, and ‘normal range’. The mean scores of girls and boys with ASD, and male and female control children, were compared by the Mann–Whitney U-test. Both the ASD subjects (n = 173) and the control subjects (n = 173) were divided into four age groups. The data for each of the four age groups of children with ASD and control children were compared by the Student's t-test. To clarify the age-related differences in emotional and behavioral problems in the two sex groups, differences among the four groups were assessed using Pearson correlation. All statistical analyses were conducted using spss 11.5 for Windows (ibm, Chicago, IL, USA). Significance level was set at P < 0.05.

Results

Demographic data

The mean age of the children with ASD at the time of evaluation was 7.92 ± 3.3 years and 129 of the 173 children (74.6%) were male (Table 1). The mean IQ score of those who were tested by the WISC-III or WISC-IV (n = 148) was 88.3 ± 20.1. Twenty-eight children with ASD (male/female = 18/10) out of 148 children with ASD (male/female = 114/34) had intellectual disability. The remaining 25 children could not be tested due to cognitive limitation (n = 23) or children's refusal (n = 2). The control children were not given the cognitive ability test.

There were no comorbid psychiatric disorders, except for one girl with well-controlled epilepsy.

Effects of IQ on the total difficulties score of SDQ in children with ASD

Multiple regression analysis of the total score of SDQ and various demographic factors showed that the total score of SDQ correlated significantly with age (B = 0.330, β = 0.204, P = 0.011) and sex (B = –2.802, β = –0.224, P = 0.006), but not with IQ (B = –0.006, β = –0.022, P = 0.781).

Patterns of emotional and behavioral problems in children with ASD

Table 2 shows the range of the total and subscale scores of SDQ using the cut-off values. The percentages of children with ASD in total difficulties and all subscales were significantly higher than those of control children (P ≤ 0.005), indicating more emotional and behavioral problems in patients than the controls. Moreover, the percentages of children in the clinical range were more than 40% in the hyperactivity/inattention, peer problems, and prosocial behavior scales among boys with ASD and in all subscales among girls with ASD.

Table 2. Proportion of ASD and control children in the normal, borderline, and clinical range of the SDQ
SDQ scalesNormal rangeBorderline rangeClinical range
RawControlASDRawControlASDRawControlASD
  1. Data are numbers of patients (percentages). *P < 0.05, compared with the control of the same group.
  2. ASD, autism spectrum disorders; SDQ, Strengths and Difficulties Questionnaire.
Boys         
Total difficulties0–13100 (77.5)*41 (31.8)14–1612 (9.3)31 (24.0)*17–4017 (13.2)57 (44.2)*
Emotional symptoms0–3105 (81.4)*85 (65.9)410 (7.8)20 (15.5)*5–1014 (10.9)24 (18.6)*
Conduct problems0–282 (63.6)*60 (46.5)320 (15.5)30 (23.3)*4–1027 (20.9)39 (30.2)*
Hyperactivity/inattention0–5102 (79.1)*54 (41.9)613 (10.1)21 (16.3)*7–1014 (10.9)54 (41.9)*
Peer problems0–290 (69.8)*31 (24.0)320 (15.5)*15 (11.6)4–1019 (14.7)83 (64.3)*
Prosocial behavior6–1068 (52.7)*36 (27.9)529 (22.5)26 (20.2)0–432 (24.8)67 (51.9)*
Girls         
Total difficulties0–1342 (95.5)*9 (20.5)14–162 (4.5)11 (25.0)*17–400 (0)24 (54.5)*
Emotional symptoms0–338 (86.4)*20 (45.5)44 (9.1)6 (13.6)*5–102 (4.5)18 (40.9)*
Conduct problems0–235 (79.5)*15 (34.1)32 (4.5)9 (20.5)*4–107 (15.9)20 (45.5)*
Hyperactivity/inattention0–541 (93.2)*10 (22.7)60 (0)5 (11.4)*7–103 (6.8)29 (65.9)*
Peer problems0–237 (84.1)*12 (27.3)35 (11.4)4 (9.1)4–102 (4.5)28 (63.6)*
Prosocial behavior6–1026 (59.1)*14 (31.8)58 (18.2)*4 (9.1)0–410 (22.7)26 (59.1)*

Sex differences in SDQ scores in children with ASD and in control children

The mean scores of total difficulties, hyperactivity/inattention, and conduct problems were significantly higher in girls with ASD than in boys (Table 3). On the other hand, the scores of total difficulties and hyperactivity/inattention in control boys were significantly higher than those of girls.

Table 3. Differences in mean SDQ scores between boys and girls with ASD and control subjects
SDQ scalesControlASD
Boys (n = 129)Girls (n = 44)PBoys (n = 129)Girls (n = 44)P
  1. *P-values for differences between boys and girls of the same group (by Mann–Whitney U-test).
  2. ASD, autism spectrum disorders; SDQ, Strengths and Difficulties Questionnaire.
Total difficulties9.75 (5.8)7.11 (3.5)0.013*16.08 (5.1)18.48 (5.4)0.020*
Emotional symptoms1.87 (2.0)1.32 (1.6)0.1752.93 (2.1)3.80 (2.7)0.077
Conduct problems2.33 (1.8)1.80 (1.3)0.1422.78 (1.9)3.34 (1.7)0.041*
Hyperactivity/inattention3.70 (2.3)2.73 (2.0)0.006*5.91 (2.3)6.70 (2.1)0.029*
Peer problems1.85 (1.6)1.27 (1.1)0.0684.45 (2.3)4.64 (2.5)0.612
Prosocial behavior5.71 (2.0)6.02 (2.0)0.3464.28 (2.5)4.30 (2.8)0.862

Age-related differences in emotional and behavioral problems in children with ASD

The total and subscale scores of the SDQ stratified by age are shown in Table 4. In each age group, the total SDQ difficulty score was significantly higher in children with ASD than control children and this was true in both boys and girls. The total SDQ scores of girls and boys with ASD increased with age (boys: r = 0182, P = 0.039; girls: r = 0.483, P = 0.001). On the other hand, the same scores for control children remained unchanged during development (boys: r = –0.168, P = 0.057; girls: r = –0.025, P = 0.871).

Table 4. Chronological changes in the mean score of the SDQ
Age: n (M/F)4–6 years: 84 (63/21)7–9 years: 32 (28/4)10–12 years: 38 (24/14)13–16 years: 19 (14/5)rP**
Mean (SD)P*Mean (SD)P*Mean (SD)P*Mean (SD)P*
  1. *P-values for differences between children with ASD and controls of the same age group (by Student's t-test). **P values for differences among the four age groups by Pearson correlation.
  2. Total difficulties scores are sum of emotional symptoms, conduct problems, hyperactivity/inattention, and peer problems.
  3. ASD, autism spectrum disorders; SDQ, Strengths and Difficulties Questionnaire.
Total difficulties          
BoysASD14.84 (4.8)<0.00117.79 (4.4)<0.00116.54 (5.5)0.00117.36 (6.2)<0.0010.1820.039*
 Control10.27 (5.7) 10.11 (5.7) 10.38 (6.2) 5.64 (3.9) –0.1680.057
GirlsASD15.62 (4.2)<0.00119.25 (6.5)0.02421.43 (4.9)<0.00121.60 (5.1)0.0040.4830.001*
 Control7.38 (3.5) 8.50 (3.0) 5.43 (2.9) 9.60 (4.4) –0.0250.871
Emotional symptoms          
BoysASD2.49 (1.8)0.1793.14 (2.4)0.0263.33 (2.2)0.0622.79 (2.1)<0.0010.2210.012*
 Control2.02 (2.1) 1.86 (1.7) 2.17 (2.1) 0.71 (1.7) –0.1200.176
GirlsASD2.71 (2.1)0.0094.25 (2.9)0.0435.21 (3.1)<0.0014.00 (3.0)0.6480.2940.052
 Control1.19 (1.5) 0.50 (0.6) 1.07 (1.3) 3.20 (2.3) 0.2160.159
Conduct problems          
BoysASD2.62 (1.8)0.6143.39 (2.0)0.1232.88 (2.0)0.3732.14 (2.1)0.196–0.0110.900
 Control2.46 (1.7) 2.54 (2.1) 2.38 (1.9) 1.29 (1.1) –0.1450.102
GirlsASD2.67 (1.6)0.0253.75 (1.5)0.3823.93 (1.6)<0.0014.20 (1.3)0.1740.4150.005*
 Control1.67 (1.1) 2.75 (1.5) 1.36 (1.2) 2.80 (1.6) 0.1220.429
Hyperactivity/inattention          
BoysASD5.90 (2.1)<0.0016.68 (2.5)<0.0015.00 (2.4)0.0815.93 (2.3)<0.001–0.0660.461
 Control3.90 (2.3) 3.96 (2.2) 3.75 (2.4) 2.14 (1.4) –0.1880.033*
GirlsASD6.62 (2.4)<0.0017.50 (2.5)0.1346.64 (1.7)<0.0016.60 (2.1)0.0010.0140.929
 Control3.19 (2.3) 4.00 (3.2) 2.00 (0.9) 1.80 (0.8) –0.2880.058
Peer problems          
BoysASD3.83 (2.2)<0.0014.57 (1.9)<0.0015.33 (2.4)<0.0015.50 (2.3)<0.0010.2860.001*
 Control1.89 (1.6) 1.75 (1.8) 2.08 (1.7) 1.50 (1.5) –0.0230.792
GirlsASD3.62 (2.2)<0.0013.75 (2.6)0.1115.64 (2.3)<0.0016.80 (1.9)0.0020.4320.003*
 Control1.33 (1.1) 1.25 (0.5) 1.00 (1.0) 1.80 (1.6) –0.0070.966
Prosocial behavior          
BoysASD3.98 (2.6)<0.0014.61 (2.7)0.0094.33 (2.2)0.0564.86 (2.9)0.7170.1040.241
 Control5.79 (2.0) 6.32 (2.0) 5.46 (1.7) 4.50 (2.3) –0.1640.064
GirlsASD3.81 (2.2)<0.0015.50 (3.5)0.7354.57 (3.2)0.1864.60 (3.6)0.6650.1460.346
 Control6.43 (1.7) 4.75 (2.4) 6.00 (2.4) 5.40 (1.1) –0.1500.330

The scores of emotional symptoms of boys with ASD increased significantly with age (r = 0.221, P = 0.012) and the scores of children with ASD were higher than those of control children in both sexes. In contrast, the emotional score of male controls over 13 years of age was significantly lower than that of male controls aged 10–12 years. Accordingly, the gap between boys over 13 years with ASD and control boys of the same age was larger. The scores of female controls increased significantly and matched the level of children with ASD. Consequently, there seemed no emotional discrepancy between girls with ASD and control girls in this age group.

The score of conduct problems of boys with ASD did not increase with age (r = –0.011, P = 0.900) and there were no significant differences between children with ASD and control children. In contrast, the score of conduct problems in girls with ASD increased significantly with age (r = 0.415, P = 0.005). Particularly, the scores of girls with ASD aged 4–6 years and 10–12 years were significantly higher than those for control girls.

The scores of hyperactivity/inattention were higher in both girls and boys with ASD, compared with the controls. Age had no effect on these scores in children with ASD (boys: r = –0.066, P = 0.461; girls: r = 0.014, P = 0.929). However, the score for male and female control children significantly decreased with age (boys: r = –0.188, P = 0.033; girls: r = –0.288, P = 0.058). Thus, the discrepancy of scores between children with ASD and control children was significant in both sexes.

The scores for peer relationship problems were similar in both girls and boys with ASD. The scores of children with ASD were significantly higher than those of the control children for all age groups, and significantly increased with age in both sexes (boys: r = 0.286, P = 0.001; girls: r = 0.432, P = 0.003) and the scores were higher than those of the controls at all ages.

The scores of prosocial behavior in children with ASD were significantly lower than those of control children in 4–9-year-old boys and in 4–6-year-old girls. However, the scores showed no significant age-related differences in children with ASD or control children (ASD boys: r = 0.104, P = 0.241; ASD girls: r = 146, P = 0.346; control boys: r = –0.164, P = 0.241; control girls: r = –0.150, P = 0.330).

Discussion

Using the SDQ, the present study identified the high prevalence of emotional and behavioral problems in children with ASD compared to age- and sex-matched control children. The study also demonstrated the age-related differences of emotional and behavioral problems in children with ASD and control children.

Previous reports showed significant correlation between emotional/behavioral difficulties and ASD symptoms. Matilla et al. reported that 74% of children aged 9–16 years with Asperger's syndrome (AS)/high-functioning autism (HFA) have co-existing psychiatric disorders.[14] Sturm et al. mentioned that 95% of children with ASD aged 5–12 years had attentional problems, 75% had motor difficulties, and 50% had impulsiveness based on examination of medical and psychiatric records.[15] Hartley et al. reported that the total problems' score of the Child Behavior Checklist (CBCL) was in the clinically significant range in one-third of young children with autistic disorders aged 1.5–5.8 years and that the highest percentages of clinically significant scores were for the withdrawal, attention, and aggression CBCL syndrome scales.[16] Mazzone et al. reviewed psychiatric comorbidities in AS/HFA reported during the period of 2000–2011. Their analysis identified several studies that reported the association of AS/HFA with psychiatric comorbidities, including depression in six articles, bipolar mood disorders in two, anxiety disorders in nine, obsessive–compulsive disorders in six, and ADHD in five.[17] However, most of these studies did not include control groups and covered limited age groups. Thus, little is known about the course of emotional and behavioral problems of children with ASD. To our knowledge, this is the first study that evaluated the age-related differences of emotional and behavioral problems in children with ASD aged 4–16 years relative to children from the local community.

The present study demonstrated that age and sex correlated significantly with the total difficulties score of SDQ. However, intelligence did not correlate with the total score of SDQ in children with ASD. Estes et al. mentioned that poor level of intellectual functioning may be a risk factor for different patterns of associated symptoms of ASD in later childhood; children with higher functioning at age 6 displayed increased internalizing symptoms by age 9, whereas children with lower functioning displayed higher hyperactivity, attention problems, and irritability at 9 years of age.[18] In another study, the prevalence of psychiatric disorders was 36% among children with intellectual disability and 8% among children without, and the prevalence was particularly high in patients with autism-spectrum disorders, hyperkinesis, and conduct disorders.[19] Our study showed no correlation between the level of intellectual functioning and emotional and behavioral problems in children with ASD. Emotional and behavioral problems might be strongly associated with core features of ASD rather than intellectual functioning.

The present results showed a trend for sex-related differences in SDQ scores in children with ASD and control children. Girls with ASD had significantly more difficulties, especially hyperactivity/inattention and conduct problems, than boys with ASD. Our findings are consistent with the results of previous studies; the CBCL scores of total difficulties and hyperactivity/inattention in girls with ASD were significantly higher than the respective scores of boys with ASD, and girls with ASD aged 1.5–3.9 years exhibited more sleep problems and anxiety or depression than boys with ASD.[20] Steinhausen and Metzke reported more pronounced behavioral abnormalities in their study of autistic girls than in autistic boys, such as self-absorbed behavior and anxiety.[21] Interestingly, reports of parents of children with high-functioning autism indicated significantly more symptoms in girls than in boys, particularly social problems, attention problems, and thought problems. Severe problems in girls with ASD might lead to adjustment difficulties, or might be a potential interpreting bias by parents who may expect more socially desired behaviors from daughters than from sons.[22] Considered collectively, thorough assessment and intervention in these domains are necessary. Comorbid behavior problems are often associated with greater distress to families and teachers than core autistic symptoms.[23] Girls with ASD and their families may be at a greater risk for distress rather than boys with ASD.

Early intervention to deal with emotional symptoms is crucial as they could lead to future development of depression if left untreated. This argument is based on previous studies, which showed that children or adults with ASD hardly express depressive mood clinically until the depression is quite severe and/or prolonged.[24] Many individuals with autism neither have sufficient language skills to express feelings, nor describe changes in mood or comment on the presence of biological symptoms of depression. For this reason, parents might not be able to notice their mood change.

Peer problems from the questionnaire, including ‘tends to play alone’, ‘has at least one good friend’, ‘generally liked by other children’, ‘picked on or bullied by other children’, and ‘get on better with adults than with other children’, increased markedly with advancing age both in girls and boys with ASD. Peer problems could be strongly related to the main features of ASD, such as difficulties in reciprocal social interaction skills or deficits in communication skills.

The subscale of prosocial behavior, which is an important component of the SDQ, includes ‘considerate of other people's feelings’, ‘shares readily with other children’, ‘helpful if someone is hurt, upset of feeling ill’, ‘kind to younger children’ and ‘often volunteers to help others’. The results of the present study showed significant differences between children with ASD and control children by 9 years of age in boys, and by 6 years of age in girls. The discrepant results may be due to differences in the speed in acquiring prosocial behaviors. Children with ASD could develop prosocial behaviors gradually, and girls with ASD could learn faster than their male counterparts.

We used SDQ to evaluate emotional and behavioral problems in ASD children because it is easy to complete by caregivers due to the size of the questionnaire, and it is a suitable screening tool for emotional and behavioral problems. Goodman and Goodman demonstrated that the total difficulties score provided an accurate and unbiased method for the assessment of mental health and that it closely predicted the prevalence of clinically rated child mental disorders in Britain.[25] Cut-off values of clinical, borderline and normal ranges were identified for children aged 4–16 years for the English version and aged 4–12 years for the Japanese version.[12] As the age range of our subjects was 4–16 years, we used the cut-off values for the English version in this study. In order to see if the use of the cut-off values for the English version is appropriate, we also analyzed our data based on the cut-off values for the Japanese version. In terms of the total difficulty score of SDQ, the clinical range of children increased in girls and boys with ASD and there were no statistically significant differences between the use of the English and Japanese cut-off values. We also found that significant differences were not observed with the use of the English or the Japanese cut-off values, except for the scores of conduct problems in boys and girls and the score of peer problems in girls; there were no significant differences observed in all three ranges of conduct problems in boys and the borderline range of conduct problems in girls between ASD and control children, and also in the borderline range of peer problems in boys between ASD and control children.

Our data showed that the clinical range of the total score of SDQ in children with ASD was 44.2% in boys, and was relatively higher in girls (54.5%). Moreover, the proportions of children with ASD not only in the clinical range but also in the borderline range were significantly higher than those of controls, especially for emotional problems, hyperactivity/inattention, and conduct problems. Considered together, the division of the SDQ score into three ranges, clinical, borderline and normal, is meaningful, highlighting the need to pay attention to children with ASD in both clinical and borderline ranges of the SDQ. The use of the SDQ is potentially useful to prevent such co-existing problems in advance.

The present study has certain limitations. First, the sample size of the boys and girls was different. ASD is approximately 3–4 times more prevalent in boys than girls and the majority of ASD research has focused on boys with ASD.[26] We focused on sex differences in this study, but the sample size was still small even though it was appropriate for statistical comparison. Second, we examined children with ASD referred to a University Hospital, which meant that the results of this study cannot be generalized to children with ASD who are not diagnosed as having ASD until they have problems later and are diagnosed in psychiatric hospitals or children's hospitals. Third, the emotional and behavioral problems were assessed only through parental reports. Comparison among self-perception of their behaviors, teacher's and parent's aspects would be of interest from a research point of view. Fourth, the control group was recruited from community children; in other words, a community sample could include subjects with psychiatric problems like ASD. Lastly, we used SDQ to describe emotional and behavioral problems, but it is difficult to distinguish such problems from the main features of ASD and clarify differences between developmental disorders, such as ASD or intellectual disabilities. Most of the clinical studies conducted so far have been cross-sectional, comparing one particular clinical measure at a single time-point across samples. Longitudinal studies are needed that closely follow the age-related differences of ASD and to detect subtle changes in behavior at different stages of development. Further studies of larger sample size and alternative measures of autistic symptoms and developmental functioning are needed to confirm the results of the present study. In addition, assessment of emotional and behavioral problems through self-report of children with ASD should be of interest.

In conclusion, the present study examined the age-related differences of emotional and behavioral problems of ASD compared with community children, and investigated sex differences. Age-related characteristics of emotional and behavioral problems in ASD were different from that of the control, and sex differences were evident. For early intervention, there is a need to clarify the age-related differences in the main features of ASD as well as the emotional and behavioral problems.

Acknowledgments

The authors appreciate the help from the Toon City Educational Committee. The authors declare no conflict of interest.

Ancillary