Hyperactivity and comorbidity in Japanese children with attention-deficit/hyperactivity disorder

Authors


Kazuhiro Takahashi, MD, Department of Neuropsychiatry, Osaka City University Medical School, 1-4-3 Asahimachi, Abenoku, Osaka 545-8585, Japan. Email: kazuhiro@med.osaka-cu.ac.jp

Abstract

Abstract  No previous study about comorbidity of attention-deficit/hyperactivity disorder (ADHD) in Japan have carried out both a comprehensive investigation using a structured interview and a comparison between ADHD subtypes. The aim of the present study was to clarify the relationship between hyperactivity and disruptive behavior disorder (DBD) in ADHD by comparing a hyperactivity group (HG) with a non-hyperactivity group (non-HG). After diagnosis was carried out by strict exclusion, the 41 ADHD subjects (6–14 years old; IQ, 70–121) diagnosed according to DSM-IV were divided into HG (n = 24) and non-HG (n = 17), and compared for comorbidities and psychopathologies. This was done via semistructured interview with children and parents and questionnaires to parents and teachers. The results demonstrate that (i) most ADHD children had comorbidity (e.g. DBD or anxiety disorder); (ii) the HG had a significantly higher rate of DBD than the non-HG, but the total number of anxiety disorders was not different between subgroups, and (iii) the HG generally had more serious psychopathologies both at home and at school than the non-HG. Both groups had more serious externalizing and internalizing problems at home than at school. The present study provides evidence of a strong relationship between hyperactivity and DBD.

INTRODUCTION

Attention-deficit/hyperactivity disorder (ADHD) is one of the most common psychiatric disorders of childhood. It is defined by the symptoms of inattention and/or hyperactivity/impulsivity. The prevalence of ADHD is reported to be 3–7% in children in USA,1 and approximately 7.7% in Japan (using parents' rating).2 The DSM-IV3 divides ADHD into three diagnostic subtypes based on the number of symptoms on two dimensions, inattention and hyperactivity/impulsivity. These are predominantly inattentive subtype (ADHD/I), predominantly hyperactive/impulsive subtype (ADHD/HI) and combined subtype (ADHD/C).

Several researchers reported the following findings. The rates of problems differ among ADHD subtypes with regard to behavior and academic performance.4–8 ADHD/C has a higher comorbidity rate with disruptive behavior disorder (DBD) (oppositional defiant disorder [ODD] or conduct disorder [CD]) than ADHD/I,4–7 but these two groups do not differ with regard to internalizing disorders.7 The parents of ADHD/C children tend to report more externalizing, delinquent, and aggressive behavior problems.5

However, because there have been few reports on ADHD subtypes in Japan, it is unknown whether or not these results in the West are valid in Japan.

Several previous studies have not strictly excluded pervasive developmental disorder (PDD). In addition,although it is well known that the intelligence and age of a child affects their comorbidity diagnosis, many previous ADHD comorbidity studies included children with mental retardation (MR) or children ranging widely in age (from preschool to adolescent) as subjects. Ishii et al.9 investigated the comorbidity of each ADHD subtype, and Saito investigated ADHD comorbidity using a semistructured interview.10 However, there has been no study investigating the comorbidity of each ADHD subtype comprehensively using a standardized structured interview, in Japan until now.

The aim of the present study was to clarify the relationship between hyperactivity and DBD in ADHD children. After having strictly excluded PDD, we diagnosed the subtypes and comorbidities of ADHD children according to semistructured interview. We also compared the hyperactivity group (HG) with the non-hyperactivity group (non-HG) for comorbidity and psychopathology.

METHODS

The subjects consisted of children who met ADHD criteria defined by DSM-IV, among 257 children aged 6–15 years (elementary school or junior high school students) under treatment at the child psychiatry outpatient clinic of the Osaka City University Hospital between January 2003 and January 2006.

Children having MR (IQ < 70 by Wechsler Intelligence Scale for Children−3rd edition; WISC-III), an acute psychotic state, cerebral palsy, or epilepsy were excluded. We carried out strict PDD exclusion (i) by interviewing parents carefully about their children's development history and any history of autistic behaviors their children exhibited since infancy via the questionnaires; (ii) by obtaining the information carefully from teachers via questionnaires about PDD; (iii) by having the interviews carried out by more than two child psychiatrists having experience of PDD medical exami-nation and treatment, and (iv) by observing clinical courses for more than 3 months. Ultimately, we excluded 54 PDD and 18 MR subjects from 257 children.

Two child psychiatrists having experience of ADHD medical examination diagnosed ADHD and the comorbidity of ADHD by performing clinical examination and follow up of each child patient, using the Japanese version of the Schedule for Affective Disorders and Schizophrenia for School-Age Children–present and lifetime version (K-SADS-PL-J)11 for each child patient and their parent, and using questionnaires (Japanese version of Attention-Deficit/Hyperactivity Disorder Rating Scale−4th edition, ADHD RS-IV-J;12 Child Behavior Checklist, CBCL;13 Teacher's Report Form, TRF13) rated by parents and teachers. As a result, 46 children were confidently diagnosed as ADHD.

Because the methylphenidate used for ADHD treatment affects clinical symptoms, all evaluations in the present study were performed after stopping use of methylphenidate for at least 24 h to remove any influence of the drug.

The purpose of the study was explained, and written consent was obtained from each parent. We excluded five children from the 46 because three did not consent and two did not have sufficient information. Finally, 41 ADHD children participated in the present study, consisting of 34 boys and seven girls. The mean age of the subjects was 8.7 years (range, 6–14 years). The mean full-scale IQ score was 95 (range, 70–121; Table 1). These subjects were classified into ADHD/C (n = 21; 51.2%), ADHD/I (n = 17; 41.5%) and ADHD/HI (n = 3; 7.3%). The 3 subjects classified in ADHD/HI very closely resembled ADHD/C in symptomatology because they satisfied 5 inattention items of ADHD criteria in K-SADS-PL-J (if they met one more inattention items, ADHD/HI changed into ADHD/C). To investigate the relationship between hyperactivity and comorbidity or other psychopathologies, we combined data from subjects in the ADHD/HI and ADHD/C groups that satisfied the hyperactivity/impulsivity criteria of ADHD in DSM-IV, and compared 24 HG subjects (ADHD/HI and ADHD/C) with 17 non-HG subjects (ADHD/I) for comorbidities and psychopathologies. There was no significant difference in sex distribution, age or full-scale IQ score between the HG and the non-HG (Table 1).

Table 1.  Demographics of ADHD subjects
 TotalADHD subgroupsAnalysis
ADHD
(n = 41)
HG
(n = 24)
non-HG
(n = 17)
tP
  • † 

    Fisher's exact test.

  • ADHD, attention-deficit/hyperactivity disorder; Full IQ, full-scale intelligence quotient by Wechsler intelligence scale for children−3rd edition; HG, hyperactivity group; non-HG, non-hyperactivity group.

Age (years; mean ± SD)8.7 ± 2.38.4 ± 2.09.1 ± 2.6−0.9680.339
Male/female (n)34/722/212/5 0.105
Full IQ (mean ± SD)95.0 ± 12.196.5 ± 13.493.0 ± 10.30.8940.377

MEASURES

K-SADS-PL

The K-SADS-PL is a semistructured interview developed for the purpose of making psychiatric diagnoses of patients aged 6–18 years.14 For each child patient and their parent, an interviewer performs a detailed interview taking 2–3 h. The diagnosis process is known for its scrupulousness and high reliability. It was confirmed that K-SADS-PL-J11 scored well on Cohen's κ (ADHD = 1.00, DBD = 0.91, anxiety disorders = 0.76, tic disorders = 0.75) for interrater reliability by the simultaneous interview method and had concurrent validity for ADHD, DBD, and anxiety disorders using some evaluation scales.

ADHD RS-IV-J

Yamazaki produced this Japanese edition, ADHD RS-IV-J,12 from the ADHD rating scale-IV developed by DuPaul et al. in 1998.15 Inspection of the reliability and validity showed normal values. ADHD RS-IV-J is a questionnaire consisting of 18 items (nine hyperactivity/impulsivity items and nine inattention items). Parents, teachers, and the child (if >12 years old) evaluate behavioral symptoms using a four-point Likert-type scale (ranging from 0, rarely or never, to 3, very often).

CBCL, TRF

Itani et al. standardized the CBCL and TRF of Achenbach16 in the Japanese edition.13 These are 113-item questionnaires rated by each of the parents and teachers using the following scale: 0, not true (as far as you know); 1, somewhat or sometimes true; 2, very true or often true; and describe behavioral and emotional problems: withdrawn, somatic complaints, anxious/depressed, social problems, thought problems, attention problems, delinquent behavior and aggressive behavior. Raw scores are converted into T-scores (mean 50; SD 10) which were adjusted for age and gender.

Statistical analysis

All statistical analyses were performed using SPSS 11.5J for Windows (SPSS Japan, Tokyo, Japan). We used the χ2 test with continuity correction and Fisher's exact test for categorical comparisons of the data. To compare continuous variables, t-tests were used to examine group differences after the normality of the distribution of scores on all measures was examined. The Levene F-test was used to determine homogeneity of variance, and the t-test was modified for unequal variances when appropriate. Differences of evaluated scores between a parent and a teacher were tested using paired t-tests. Statistical significance was set at P < 0.05 (two-tailed).

RESULTS

Table 2 shows the present comorbidity diagnoses in each ADHD subgroup. Thirty-six subjects (87.8%) had a comorbid diagnosis. The mean number of comorbid disorders was 1.95 ± 1.18.

Table 2.  Comorbidity of subjects with ADHD
 TotalADHD subgroupsAnalysis
ADHD
(n = 41)
n (%)
HG
(n = 24)
n (%)
non-HG
(n = 17)
n (%)
χ2P
  • † 

    Fisher's exact test.

  • ‡ 

    Pearson's chi-squared test with continuity correction.

  • ADHD, attention-deficit/hyperactivity disorder; Full IQ, full-scale intelligence quotient by Wechsler intelligence scale for children−3rd edition; HG, hyperactivity group; non-HG, non-hyperactivity group.

Comorbid diagnosis
 Any comorbid ADHD36 (88)23 (96)13 (76) 0.141
 Any disruptive behavior disorder24 (58)18 (75)6 (35)4.930.026
  Oppositional defiant disorder15 (37)11 (46)4 (24)  
  Conduct disorder9 (22)7 (29)2 (12)  
 Any anxiety disorder21 (51)10 (42)11 (65)1.290.256
  Specific phobia13 (32)7 (29)6 (35)0.0060.946
  Separation anxiety disorder6 (15)4 (17)2 (12) 1.00
  Generalized anxiety disorder4 (10)0 (0)4 (24) 0.024
  Obsessive-compulsive disorder4 (10)1 (4)3 (18) 0.290
  Social phobia4 (10)2 (8)2 (12) 1.00
 Other disorder
  Tic disorders9 (22)4 (17)5 (29)0.3460.556
  Enuresis8 (20)7 (29)1 (6) 0.110
  Encopresis1 (2)1 (4)0 (0) 1.00
  Adjustment disorders1 (2)0 (0)1 (6) 0.415

Twenty-four subjects (58%) had DBD. The frequency of DBD in the HG was significantly higher than in the non-HG (P = 0.026).

The most prevalent anxiety disorder was specific phobia (n = 13). Subtypes in this disorder were darkness phobia (n = 6), height phobia (n = 4) and others. In addition, separation anxiety disorder (SAD), social phobia, generalized anxiety disorder (GAD) and obsessive-compulsive disorder were diagnosed. Although the frequency of anxiety disorders as a whole in the non-HG did not differ significantly from that in the HG, the frequency of GAD in the non-HG was significantly higher than in the HG (P = 0.024).

The other disorders diagnosed in the present study were tic disorders (motor tic disorder, vocal tic disorder and Tourette's disorder), enuresis, encopresis and adjustment disorders with depressed mood. In the present study we obtained no comorbid diagnoses such as mood disorders, psychotic disorders, eating disorders, alcohol abuse and substance abuse.

Table 3 shows the ADHD RS-IV-J scores rated by parents and teachers. total score, inattention score and hyperactive/impulsive score by rating of both parents and teachers in the HG were significantly higher than those in the non-HG (parents: P < 0.001, P < 0.001, P = 0.044; teachers: P = 0.001, P = 0.001, P = 0.015, respectively). When we compared teachers' ratings with parents' ratings, total score, inattention score and hyperactive/impulsive score by parents' rating were significantly higher than those by teachers in the HG (P = 0.003; P = 0.020; P = 0.003, respectively), and also in the non-HG (P = 0.001; P = 0.002; P = 0.004, respectively).

Table 3.  ADHD RS-IV-J score rated by parents and teachers
 TotalADHD subgroupsAnalysis
ADHD
Mean ± SD
HG
Mean ± SD
non-HG
Mean ± SD
tP
  1. ADHD, attention-deficit/hyperactivity disorder; ADHD RS-IV-J, Japanese version of Attention-Deficit/Hyperactivity Disorder Rating Scale−4th edition; HG, hyperactivity group; non-HG, non-hyperactivity group.

Parentn = 40n = 24n = 16  
 Total score32.9 ± 9.1637.2 ± 6.826.5 ± 8.64.39<0.001
 Inattention score18.9 ± 4.620.0 ± 3.617.1 ± 5.52.08<0.001
 Hyperactive/impulsive score14.1 ± 5.617.2 ± 3.99.4 ± 4.55.750.044
Teachern = 38n = 23n = 15  
 Total score22.9 ± 14.028.4 ± 13.914.4 ± 9.33.440.001
 Inattention score13.0 ± 6.715.1 ± 6.69.8 ± 5.82.560.001
 Hyperactive/impulsive score9.9 ± 8.013.3 ± 8.14.6 ± 4.14.360.015

Table 4 shows T-scores of CBCL and TRF. Borderline clinical ranges of total, internalizing and externalizing are defined by T-scores from 60 to 63, and borderline clinical ranges of withdrawn, somatic complaints, anxious/depressed, social problems, thought problems, attention problems, delinquent behavior and aggressive behavior are defined by T-scores from 67 to 70. According to these cut-off points, clinically significant higher scores were obtained on CBCL for total, internalizing, externalizing and attention problems of total ADHD (the total of two ADHD subgroups) and both ADHD subgroups, and delinquent behavior of total ADHD, and aggressive behavior of total ADHD and the HG. Also on TRF, clinically significant higher scores were obtained for total and externalizing of total ADHD and the HG.

Table 4.  T-score of CBCL and TRF between ADHD subgroups
 TotalADHD subgroupsAnalysis
ADHD
Mean ± SD
HG
Mean ± SD
non-HG
Mean ± SD
tP
  1. ADHD, attention-deficit/hyperactivity disorder; CBCL, Child Behavior Checklist; HG, hyperactivity group; non-HG, non-hyperactivity group; TRF, Teacher's Report Form.

CBCLn = 40n = 24n = 16  
 Total74.6 ± 7.774.9 ± 7.274.1 ± 8.50.320.748
 Internalizing68.1 ± 8.367.2 ± 8.169.5 ± 8.6−0.870.390
 Externalizing76.2 ± 11.079.4 ± 11.071.4 ± 9.42.380.023
 Withdrawn67.1 ± 8.266.3 ± 8.368.3 ± 8.1−0.760.451
 Somatic complaints59.1 ± 8.958.3 ± 9.060.4 ± 8.9−0.760.454
 Anxious/depressed67.0 ± 9.466.1 ± 9.168.3 ± 10.0−0.730.471
 Social problems68.4 ± 8.968.2 ± 8.868.8 ± 9.5−0.220.826
 Thought problems64.1 ± 9.762.5 ± 9.966.3 ± 9.3−1.210.235
 Attention problems72.9 ± 7.072.8 ± 7.673.1 ± 6.2−0.160.870
 Delinquent behavior70.0 ± 7.471.1 ± 7.868.3 ± 6.81.190.244
 Aggressive behavior75.1 ± 11.478.7 ± 11.069.8 ± 10.02.610.013
TRFn = 39n = 23n = 16  
 Total65.5 ± 8.468.4 ± 9.561.5 ± 4.12.700.010
 Internalizing58.9 ± 6.461.1 ± 6.555.7 ± 4.92.830.007
 Externalizing66.6 ± 8.669.9 ± 9.361.9 ± 4.33.560.001
 Withdrawn58.7 ± 66.160.6 ± 6.456.1 ± 6.12.190.035
 Somatic complaints53.1 ± 6.454.2 ± 7.451.6 ± 4.41.340.188
 Anxious/depressed58.6 ± 7.660.8 ± 7.855.3 ± 6.12.370.023
 Social problems61.7 ± 6.064.0 ± 6.258.5 ± 3.93.120.003
 Thought problems56.9 ± 9.659.8 ± 10.152.6 ± 7.22.600.013
 Attention problems65.2 ± 8.067.2 ± 8.862.4 ± 5.71.910.063
 Delinquent behavior63.1 ± 8.366.0 ± 7.558.8 ± 7.62.970.005
 Aggressive behavior67.2 ± 10.566.0 ± 7.561.1 ± 5.43.800.001

When we analyzed the differences between the HG and the non-HG, the HG had statistically higher T-scores than the non-HG on externalizing andaggressive behavior on the CBCL (P = 0.023; P = 0.013, respectively), and on total, internalizing, externalizing, withdrawn, anxious/depressed, social problems, thought problems, delinquent behavior and aggressive behavior on the TRF (P = 0.010; P = 0.007; P = 0.001; P = 0.035; P = 0.023; P = 0.003; P = 0.013; P = 0.005; P = 0.001, respectively).

Comparing TRF with CBCL on total ADHD, total, internalizing, externalizing, withdrawn, somatic complaints, anxious/depressed, social problems, thought problems, attention problems, delinquent behavior and aggressive behavior, the T-scores on the CBCL were significantly higher than those on the TRF (P < 0.001; except thought problems, P = 0.003). Comparing TRF with CBCL on the non-HG, total, internalizing, externalizing, withdrawn, somatic complaints, anxious/depressed, social problems, thought problems, attention problems, delinquent behavior and aggressive behavior, the T-scores on the CBCL were also significantly higher than those of the TRF (these 11 scores, P < 0.004). However, when we compared TRF with CBCL on the HG, total, internalizing, externalizing, withdrawn, somatic complaints, anxious/depressed, attention problems, delinquent behavior and aggressive behavior, the T-scores on the CBCL were also significantly higher than those of TRF (these 10 scores, P < 0.04), but the social problems and thought problems T-scores on the CBCL were not significantly higher than those of the TRF (P = 0.070; P = 0.391, respectively).

DISCUSSION

This is the first study to investigate the relationship between the hyperactivity and comorbidity of ADHD children comprehensively using a standardized semistructured interview in Japan. In the present study we divided 41 strictly diagnosed ADHD subjects into HG (n = 24) and non-HG (n = 17), and compared the two groups for comorbidities and psychopathologies. The results indicated that HG subjects had a higher comorbidity rate with DBD and more serious psychopathologies both at home and at school than the non-HG subjects.

Exclusion of PDD and MR

In recent years, the following findings have been noted. Even in children for whom a experienced child psychiatrist has diagnosed as ADHD, the diagnosis is changed later to PDD in 16%.17 A wrong diagnosis of ADHD can be made when we apply operational criteria to high-functioning PDD (HFPDD).18 Differential diagnosis of ADHD and HFPDD by cross-sectional evaluation is difficult in later childhood. The differential diagnosis requires an infant development history investigation and longitudinal evaluation because it is impossible to diagnose PDD by using only a structured diagnosis interview.17 These reports suggest that, because the differential diagnosis was insufficient, the previous ADHD studies might have included PDD children with inattention or hyperactivity by mistake, and the results might thus be distorted. We therefore strictly excluded PDD. In addition, it is known that the intelligence of children affects their inattention, hyperactivity and comorbid disorders. Therefore WISC-III was used for all subjects. Many children had been suspected to have ADHD because of their inattention and hyperactivity, and were referred to the Osaka City University Hospital. But consequently 54 PDD and 18 MR children were excluded from the present study. In terms of these strict exclusions, we can regard the remaining subjects as having pure ADHD.

Comorbidity

The rate of subjects having comorbid diagnoses was 87.8%, higher than the 47.1% reported by Ishii et al. in Japan.9 This may be related to differences in evaluation methods (use of a semistructured diagnosis interview). The ratio of ADHD children having comorbidities usually rises if a structured diagnosis interview is used. The high comorbidity rate in the present study suggests that Japanese ADHD children have many comorbid impairments, just like western ADHD children.1

The rate of subjects having DBD in the present study was 58%, similar to the result (64%) reported by Saito,10 who used a semistructured interview. These results are consistent with prevalence rates reported in Western studies;19 DBD was found to be present in 40–70% of children with ADHD in several epidemiological studies, and the prevalence of comorbid DBD in clinical population is probably even higher than in community samples. In the present study the frequency of DBD in the HG subjects was significantly higher than in the non-HG subjects. Table 5 shows the proportion of subjects with comorbidity in each ADHD subgroup on the previous clinical studies. We redivided the ADHD subjects of each report into HG or non-HG according to the criterion of the present study. Because the studies of Morgan et al.5 and Weiss et al.6 include no ADHD/HI, the HG of those studies consist of only ADHD/C. Despite the different cultural background, as with the present study, these studies found higher comorbid rates with DBD in HG than in non-HG subjects.5,6,9,20,21 This finding indicates a universal relationship between hyperactivity and DBD, which suggests that hyperactivity may represent a risk factor for the subsequent development of DBD, or hyperactivity and DBD may share the same or related risk factors.

Table 5.  Proportion of subjects with comorbidity in each ADHD subgroups on clinical samples
ComorbidityADHD subgroups 
HG
(%)
non-HG
(%)
  1. ADHD, attention-deficit/hyperactivity disorder; HG, hyperactivity group; non-HG, non-hyperactivity group.

Any disruptive behavior disorder
 42.36.7Morgan et al. (1996)5
87.652.2Faraone et al. (1998)20
48.118.3Weiss et al. (2003)6
62.927.3Byun et al. (2006)21
13.70Ishii et al. (2003)9
Any anxiety disorder
 11.513.3Morgan et al. (1996)5
03.5Weiss et al. (2003)6
31.431.8Byun et al. (2006)21
5.916.7Ishii et al. (2003)9
Any tic disorder
 32.417.4Faraone et al. (1998)20
17.19.1Byun et al. (2006)21
5.98.3Ishii et al. (2003)9
Any elimination disorder
 41.432.6Faraone et al. (1998)20
14.318.2Byun et al. (2006)21
7.825Ishii et al. (2003)9

Several previous studies demonstrated that the prevalence of ADHD/I tended to be higher than ADHD/C in community samples, but was lower in clinical samples. The present results are consistent with these findings. These facts suggest the following: whereas non-HG is hard to recognize as a problem for children in daily life because the inattention symptom is not outstanding, children with hyperactivity are easily recognized and referred to a hospital because comorbid behavioral problems bring serious dysfunctions.

The present study showed that the rate of comorbid anxiety disorders was 51%. This was higher than the comorbid rates of ADHD and anxiety disorders in Western reports, which found rates of approximately 25% in both community and clinical samples.19 Moreover, in the present study various comorbid anxiety disorders were apparent, but the number of subjects having anxiety disorders was not significantly different between ADHD subgroups, just as in the study by Eiraldi et al.7 Some studies reported that ADHD/C had more anxiety disorders, GAD in particular, than ADHD/I.22,23 But these studies used community samples, and no difference was found in several studies using clinical samples.5,7,20 These results may suggest the following: ADHD/C subjects come to a hospital for behavioral problems rather than anxiety disorders, although ADHD/C subjects have more anxiety disorders than ADHD/I subjects in community samples. In contrast, ADHD/I subjects may easily come to the hospital when they have anxiety disorders. Therefore the comorbidity rate of ADHD/I and anxiety disorders rose, so that the significant difference between ADHD/C and ADHD/I was lost in clinical samples.

The prevalence of mood disorders in ADHD children was reported to be relatively high (15–75%)1 in Western studies. In contrast, comorbid mood disorders were not found in the present study, and the only disorder that presented depressive symptoms was ‘adjustment disorders with depressed mood.’ One reason for this is that the prevalence of ADHD comorbid with mood disorders is lower in Japanese children than in Western children as indicated by Saito.10 Another reason is that the prevalences of major depressive disorders and dysthymic disorders in many epidemiological studies were lower for childhood than for adolescence.1,24 Namely, the comorbidity rate of mood disorders decreased markedly in the present study because of the lower mean age of the subjects.

ADHD RS-IV-J, CBCL, TRF

Externalizing Score and Aggressive Behavior Score on CBCL were significantly higher in the HG than in the non-HG subjects. This result is natural because HG frequently coexists with DBD. For TRF, almost all scores were significantly higher in the HG than in the non-HG subjects. This indicates that the symptoms of the HG subject at school are manifold and serious, or that teachers easily find other symptoms because they already recognize externalizing and aggressive problems of the HG subject.

On ADHD RS-IV-J, not only hyperactive/impulsive score but also inattention score of the HG was higher than for non-HG. In contrast, no significant difference appeared in attention problems on the CBCL and TRF. The 11 questions related to attention problems on the CBCL and TRF examine not only inattentive symptoms but also other symptoms: only four of the 11 questions relate to inattention symptoms. The difference in results may be due to these facts. We think that the results of ADHD RS-IV-J, which includes many questions specifically about inattention symptoms, are more useful than those of CBCL or TRF. As a result, it is suggested that the HG subjects have more serious symptoms than the non-HG subjects, not only in terms of hyperactive/impulsive problems but also in inattentive problems.

In addition, we found that the scores by parents were significantly higher than those by teachers on most of the subcategories of ADHD RS-IV-J, CBCL and TRF. This finding indicates that the symptoms of ADHD children are expressed more seriously at home than at school. When we assess and treat ADHD patients, it becomes important to keep in mind that there are differences between the evaluations of parents and teachers, and to obtain information from both sides.25

There were three limitations: (i) we did not analyze these data after having distributed data by gender because of the small sample size (some investigators have suggested gender differences in comorbid disorders,26,27 so it was not ideal to analyze data mixing male and female); (ii) our results cannot be generalized in community samples because the present study used clinical samples; and (iii) there is a possibility that the subtypes may change over time: several studies have reported that some ADHD children shifted to another subtype or desisted from ADHD in later years.17,28 We accounted for this as far as possible by narrowly limiting the age range.

CLINICAL IMPLICATION AND CONCLUSIONS

This study demonstrates that Japanese ADHD children with hyperactivity have more serious comorbid psychopathologies than those without hyperactivity. A treatment strategy with due consideration for comorbid disorders is thus required, especially when we treat ADHD children with hyperactivity. And it is necessary to obtain information from multiple sources in order to accurately evaluate the problems of ADHD children. A longitudinal study of ADHD subtypes, to establish the causality between hyperactivity and DBD, and to investigate the shift of comorbidity over time, would be beneficial.

Ancillary