Cognitive strengths and deficits in schoolchildren with ADHD

Transcription

1 Acta Pædiatrica ISSN REGULAR ARTICLE Cognitive strengths and deficits in schoolchildren with ADHD Ulla Ek 1, Elisabeth Fernell 2, Joakim Westerlund 1, Kirsten Holmberg 2, Per-Olof Olsson 1, Christopher Gillberg 3 1.Department of Psychology, University of Stockholm, Sweden 2.Department of Neuropediatrics, Astrid Lindgren Children s Hospital, Karolinska Hospital, Stockholm, Sweden 3.Department of Child and Adolescent Psychiatry, University of Gothenburg, Sweden Keywords ADHD, WISC, ACID profile Correspondence Ulla Ek, Department of Psychology, Stockholm University, Stockholm, Sweden. Tel: Fax: uek@psychology.su.se Received 26 October 2006; revised 18 January 2007; accepted 27 February DOI: /j x Abstract Background: Few studies provide detailed analyses of the various aspects of the entire cognitive profile of children with ADHD. Material and methods: Cognitive test data were analysed for 10- to 11-year-old children with (1) ADHD, (2) subthreshold ADHD and (3) milder attention and/or learning problems, and compared with normative data. Results: Thirty-two had ADHD and 10 met the criteria for subthreshold ADHD, prevalence rates of 5.4% and 1.6%, respectively. On a group level, children with ADHD/subthreshold ADHD, and those with milder attention and/or learning problems had almost identical cognitive profiles for the 13 subtests comprising the WISC III, with particularly low results on the arithmetic, coding, information and digit span subtests (ACID profile). When analyzed individually, a complete or incomplete ACID profile (three of four subtests) was equally common in children with ADHD/subthreshold ADHD and in children with milder problems, found in about 1/5. The relative strengths of both groups were in areas demanding logical thinking, reasoning and common sense. Conclusion: The specific ACID profile is as common in children with ADHD as in those with minor attention and/or learning problems. The cognitive weaknesses reflected in the ACID profile might play a role as an underlying factor in various developmental disorders. INTRODUCTION Recent research suggests that ADHD (attention deficit/ hyperactivity disorder) (1) is on a spectrum with milder attention deficits/hyperactivity problems and that boundaries vis-à-vis normality are blurred (2). Cognitive functions, including general intellectual levels and executive functions, have been studied in children with ADHD, DAMP [deficits in attention, motor control and perception, which is equivalent to ADHD + DCD (developmental coordination disorder)], and in children with learning disabilities (3 8). Landgren et al. (9) reported that a group of children with DAMP was characterised by a (generally) low IQ, with the majority having an IQ of below 90. In several studies, significantly lower cognitive levels have been reported in children with ADHD compared to controls (10,11). Low factor scores for processing speed and working memory are commonly considered to be of importance among individuals with ADHD (4,12,13). ADHD is considered to be a heterogeneous disorder overlapping other disorders such as dyslexia, communication and language disorders (14,15). The ACID profile is based on four subtests of the WISC (Wechsler intelligence scales for children) arithmetic, coding, information, and digit span. This profile is considered to be present when the scores of all these four subtests are equal to or lower than the lowest scores on the other nine subtests (16). The ACID profile has been shown to distinguish children with certain learning or executive dysfunctions from children without such problems (17), and has been reported, for example, in children with dyslexia (18). However, it should be noted that the applicability of the ACID profile in children with ADHD is disputed (19). Despite the average low IQ and the inability of children with ADHD to concentrate that have been reported, certain children with ADHD are gifted. Lovecky (20) found that such children had their strengths in verbal areas of the WISC. However, by and large, there have been very few studies providing detailed analyses of the various aspects of the entire cognitive profile in children with ADHD. In a previous study (21), we reported cognitive test data for a 4th grade cohort of schoolchildren who were reported by parents and teachers to have behavioural and/or learning problems. The aim of the present study was to compare the cognitive deficits and assets in three groups of children selected from that population-based study with those of same aged children from the Swedish WISC III normative population (16). The children who had cognitive deficits were those who met the criteria for (1) ADHD, (2) subthreshold ADHD and (3) attention/behaviour/learning problems, but not the criteria for ADHD or subthreshold ADHD. Our hypotheses were that (1) children with ADHD would display the specific ACID cognitive profile and (2) children with attention/behaviour/learning problems, not meeting criteria for ADHD, would show a similar profile. SUBJECTS AND METHODS Thirty-two children with ADHD and 10 meeting subthreshold criteria for ADHD (see below) were contrasted with 102 children with attention, behaviour and/or learning problems that had been reported by parents or teachers, but who did 756 C 2007 The Author(s)/Journal Compilation C 2007 Foundation Acta Pædiatrica/Acta Pædiatrica , pp

2 Ek et al. Cognitive profile of children with ADHD not meet the criteria for ADHD and with same-age children from the Swedish normative study of the WISC III. The study is part of a larger population-based survey of all 577 (295 boys, 282 girls) children attending the 4th grade of primary school, plus those 14 children (12 boys, 2 girls) who had repeated one grade and, thus, were in the 3rd grade of primary school in one municipality outside Stockholm during the academic year (child age years). The study was carried out in one municipality with approximately inhabitants in the county of Stockholm. The municipality has a relatively low socio-economic status (SES); just 12% of the adult population had higher education, whilst the corresponding rate for Sweden as a whole was 15%. Approximately 20% of the population had their roots in foreign countries, which is comparable to that of the county of Stockholm as a whole. Of these 591 (307 boys, 284 girls) children, 20 were born in 1990, 553 in 1991 and 18 in In connection with the regular school health examination, the parents were invited to participate in the study with their child. The Conners (22) 10-item-questionnaire, focusing on hyperactivity and attention problems, and a questionnaire containing a set of questions about executive functions and, specifically, reflecting passive and slow behaviour (Executive Functions Screening Scale) (21) were distributed to the parents and teachers of the 591 children who were asked to complete them and return them to the school doctor. A score of at least 10 (from a possible score of 0 30) on the parent or teacher Conners scale, or of at least 17 (of a possible score of 0 51) on the Executive Functions Screening Scale was defined as screen positivity. Each teacher was interviewed by the school doctor (KH) and requested to complete the DSM-IV ADHD symptom scale. Complete screening data were obtained for 542 children. The response rate was 91.7% for parents and 100% for teachers. Children were considered for inclusion in the present study according to the following criteria: (1) screen positivity on at least two questionnaires, with at least one of the questionnaires coming from the parents and at least one from the teachers, or (2) screen positivity on one questionnaire in combination with a report of a documented significant sensory or academic problem (such as a diagnosed developmental disorder or the fact that the child had repeated a grade). In the cohort, 10.9% and 13.6% scored above the cut-off point on the Conners scale according to parents and teachers, respectively, and 11.2% and 17.2%, respectively scored above the cut-off on the Executive Functions Screening Scale. One-hundred and sixty children in 4th grade (109 boys, 51 girls, representing 27% of all the children in 4th grade) and all 14 children in 3rd grade (12 boys, 2 girls) met the inclusion criteria (n = 174), either 1 or 2, and were considered for clinical assessment, including a cognitive assessment, performed by a psychologist. If a psychological examination had been performed during the 2 years prior to the study, the results of this were scrutinized, pending receipt of the parents approval. A structured interview with the parents was conducted by the school doctor (KH), a trained paediatrician. This interview focused on the general health of the child concerned, adjustment to school, developmental screening focusing on speech, language, behaviour, social interaction, motor abilities and current functioning in these areas. The interview was complemented with the DSM-IV criteria regarding ADHD. Seven additional children (4 boys, 3 girls) from the cohort were targeted for inclusion based on substantial concerns of their teachers concerning learning and/or behaviour problems in non-screen positive children at the time of interview. Thus, a total of 181 children (31% of the whole cohort of 591) were targeted for inclusion in the study. Fifteen of these children did not visit the doctor or psychologist. The parents of 22 children refused to allow their child to undergo psychological assessment either because they were not interested or because of major parent language problems. Thus, medical and neuropsychological assessments were performed using the WISC III (16) in 144 children (100 boys, 44 girls, corresponding to a cohort adjusted boy:girl ratio of 2:1). Once the assessments had been completed, there was an individual follow-up meeting to which all the parents were invited, and at which the results were reported and discussed. The relationship (Pearson correlation) between the parents and teachers ratings, respectively, of the child s behaviour, according to Conners and the Executive Functions Screening Scale was computed. There was a very strong positive correlation for the parents ratings across these two screening instruments for the children with ADHD (clinical plus subthreshold) r 36 =.70, p <.001. A somewhat lower, but nevertheless significant, correlation was obtained for the parents ratings of the non-adhd group, r 93 =.46, p <.001. On the other hand, the teachers ratings of the ADHD group did not correlate strongly r 39 =.23, p =.14. A relatively low, but significant, correlation was found for the teachers ratings of the non-adhd children, r 96 =.32, p =.001. Diagnosis of ADHD The inattention and hyperactivity impulsivity related symptoms of the children, as reported by their parents and teachers at the interview, were used to classify the children according to the DSM-IV criteria (1). Children meeting at least six criteria for inattention and/or at least six for hyperactivity impulsivity, according to one or other of the informants, and who had reported symptoms present both at home and school, were considered to have ADHD (of the combined, inattentive or hyperactive-impulsive type). Children fulfilling four or five of the criteria indicative of inattention and/or at least four or five criteria indicative of hyperactivity impulsivity and who had symptoms both at home and school were considered to have subthreshold ADHD. Children who met fewer criteria, below these levels, were categorised as not having ADHD. A child s behavioural problems had to be excessive for his or her mental age to be recorded as symptoms indicative of ADHD/subthreshold ADHD. For some analyses, the groups of children classified as having ADHD or subthreshold ADHD were collapsed into one category, a group composed of all children with ADHD C 2007 The Author(s)/Journal Compilation C 2007 Foundation Acta Pædiatrica/Acta Pædiatrica , pp

3 Cognitive profile of children with ADHD Ek et al. and referred to as all those with ADHD. All diagnostic uncertainties were resolved within a group comprising four experienced professionals (paediatricians and clinical psychologists). A total of 42 children (32 boys and 10 girls, adjusted boy/girl ratio 3:1) together comprised the group of those children determined to have some degree of ADHD, of these children,10 fulfilled the criteria for subthreshold ADHD (6 boys and 4 girls). The remaining 102 children (68 boys, 34 girls; adjusted boy/girl ratio 2:1) had also screened positive, but did not meet our criteria for inclusion in the ADHD group. This group composed of those determined not to satisfy the criteria for ADHD included children with milder attention problems, and/or varying degrees of behavioural and learning difficulties. The WISC III results for the two groups (ADHD and non-adhd group) were contrasted and compared to those of a reference group of 10- to 11-year-old children from the Swedish normative study conducted for the WISC III (16). Analyses of cognitive data Full Scale (FSIQ), verbal (VIQ) and performance (PIQ) IQs were obtained, as were the Kaufman factor scores, namely, the verbal comprehension index (VCI), perceptual organisation index (POI), freedom from distractibility index (FDI) and the processing speed index (PSI). The mean is 100 and the standard deviation (SD) is 15 for all of these WISC III scales. For all 13 WISC III subtests, scaled scores (mean 10, SD 3) were calculated and analysed. The ACID index (arithmetic, coding, information and digit span; mean = 40) was also calculated. Indices for crystallized intelligence (referring to the possession of a general fund of knowledge that can be used when solving problems based on information, vocabulary and comprehension) and fluid intelligence (by which it is meant an ability to conduct logical reasoning when solving problems, based on arithmetic, picture completion and object assembly) were also calculated. An level of 0.05 was used for all statistical tests. ETHICS The study was approved by the Ethics Committee at the Karolinska Hospital, Stockholm. RESULTS Prevalence of ADHD The prevalence of ADHD was calculated on the basis of the 553 children born in Thirty of these (5.4%) met the criteria for ADHD and nine (1.6%) met the criteria for subthreshold ADHD. Cognitive data Two of the 42 children in the group composed of those with ADHD and 16 of the 102 children in the group composed of those determined not to have ADHD had total IQs of below 70. Detailed data including factor scores for the two groups are presented in Table 1. Table 1 Full scale IQ (FSIQ), verbal IQ (VIQ), performance IQ (PIQ), four factor scores: verbal comprehension index (VCI), perceptual organization index (POI), freedom from distractibility index (FDI) and processing speed index (PSI) for two groups, composed of those with and those without ADHD ADHD Non-ADHD Mean SD Range Mean SD Range FSIQ VIQ PIQ VCI POI FDI PSI When analysing the 13 subscales of the WISC III for the entire group of 144 children (42/102) on a group basis, the lowest scores were obtained with the arithmetic, coding, information, digit span (ACID), and symbol search indicating that the screen positive children in the population, regardless of whether or not they had actually received a diagnosis of ADHD or not, have specific problems with attention, processing speed, auditory sequencing and working memory. When analysing the two groups separately, the mean ACID index was 30.1 and 29.2 for the groups composed of those determined to have ADHD and those who were determined not to have ADHD in this investigation, respectively (n.s.). When the profiles were analyzed individually, only two of the 42 children comprising the group of those determined to have ADHD in this study had a complete ACID profile, both were boys, and an incomplete ACID profile (three of four subtests) was found in another 6/42, 2 girls and 4 boys. In the non-adhd group, a complete ACID-profile was found in 6 children, 2 girls and 4 boys, and an incomplete ACID profile was present in 13 children, 2 girls, and 11 boys. Thus, a complete or incomplete ACID profile was equally common in the two groups, being found in about 1/5 of the children, in the two groups composed of those who had screened positive. Only 6 girls had complete or incomplete ACID profiles, in contrast to 21 boys. Ten of the 27 children had repeated 1 year at school. The results obtained with the WISC III were analyzed with a 2 13 (Group Subscale) analysis of variance with diagnostic group (ADHD and without) as the betweensubject factor, and subscale as the within-subject factor. We found almost identical profiles in the two groups (Fig. 1). There was no main effect of group (F 1,126 < 1) and no interaction effect of group subscale (F 12,1512 < 1). A significant main effect of subscale (F 12,1512 = 7.84, p <.001) indicated that the subscales were of varying degree of difficulty. The highest scores in both diagnostic groups were found on the subscales reflecting verbal comprehension and common sense (similarities, comprehension and picture arrangement) and the lowest scores in both groups were found on subscales reflecting working memory, planning and 758 C 2007 The Author(s)/Journal Compilation C 2007 Foundation Acta Pædiatrica/Acta Pædiatrica , pp

4 Ek et al. Cognitive profile of children with ADHD Figure 1 Mean performance on WISC III as a function of group (i.e., whether ADHD or non-adhd) and subscale. The abbreviations for the subscale are as follows: Inf = Information, Sim = Similarities, Arit = Arithmetic, Voc = Vocabulary, Comp = Comprehension, DS = Digit span, PC = Picture completion, Cod = Coding, PA = Picture Arrangement, BD = Block Design, OA = Object Assembly, SS = Symbol Search, Maz = Mazes. fluency (information, arithmetic, digit span, coding and symbol search). Using one-sample t-tests comparing the mean values on each subscale for the ADHD group with the Swedish normative sample, all differences were found to be statistically significant with the exception being for similarities, comprehension and picture arrangement. On these three subscales, the children in the group composed of those with ADHD scored at a level equivalent to that of the normative sample. When comparing the non-adhd group with the Swedish normative sample, significant differences were obtained on all subscales, except for picture arrangement, where they scored at the same level as the normative group. The mean indices for crystallized intelligence were 25.4 for the group composed of those determined to have ADHD and 24.5 for those without ADHD (n.s.). The mean indices for fluid intelligence were 24.2 and 24.1, respectively (n.s.). DISCUSSION Our hypotheses that children with ADHD would display the specific ACID cognitive profile and that children with attention/behaviour/learning problems, but not meeting the criteria for ADHD, would show a similar profile were confirmed, but only on a group basis. The lowest scores throughout the whole screen positive group were obtained on the arithmetic, coding, information and digit span subtests (and symbol search), corresponding to the ACID profile. This profile is known to characterise the cognitive pattern of children with dyslexia (18), a condition which is commonly combined with ADHD. Interestingly, this profile was also evident in the non-adhd group, composed of children with milder attention, behaviour and learning difficulties. Thus, the cognitive weakness reflected in this specific (and very common) cognitive profile seems to play a role in various developmental disorders, as an underlying factor. However, there are different opinions about the diagnostic utility of the ACID profile in children with attention deficit hyperactivity disorder. Snow and Sapp (23), among others, have demonstrated the value of the ACID profile or the SCAD index (symbol search, coding, arithmetic and digit span) in the diagnostic procedure, whereas in their study of Greek children, Filippatou and Livaniou (19) found no such evidence. In this study, we found an equally high ratio of complete or incomplete ACID profiles in the group composed of those with ADHD and in those with milder problems but no ADHD. These findings suggest that the ACID profile should be considered as a marker of potential attention problems in children. The FSIQ, VIQ, PIQ and the four Kaufman indices were almost identical for the children in the groups composed of those determined to have ADHD and those without. The highest scores for both groups were obtained for the verbal comprehension index, reflecting certain vocabulary-related tasks, reasoning, problem-solving and abstract verbal thinking and the perceptual organization index, reflecting nonverbal abilities and a capacity for spatial reasoning. This finding is in accordance with those of Lovecky (20), who emphasised the relative talents of children with ADHD, specifically some verbal abilities, capacities that are frequently overlooked. The lowest scores found were for the freedom from distractibility and processing speed indices, reflecting working memory and cognitive speed, this, too, is in line with other reports (4,13). Interestingly, we found no clearcut differences between the cognitive profiles of the children fulfilling the criteria for ADHD and those who did not in this sample comprised solely of those who had screened positive. On the contrary, the two groups had almost identical profiles on the 13 subtests of the WISC III, indicating that attention deficits might occur on a spectrum from milder variants to severe, regardless of ADHD-diagnostic status (2). A striking clinical observation from the study was how well the children adapted to the test situation when they received continuous support and feedback in a well-structured environment, on a one to one basis in an environment almost completely without distractions. However, such an optimal situation may contrast strongly with the child s everyday school and home environment. Indeed, the behaviour of children with attention problems is known to be highly context-bound (24). Little interest has been expended on the cognitive strengths of children with problems within the attention spectrum (20), studies usually only focus on deficits. However, our findings suggest that cognitive assets in children with attention, learning and behaviour problems may provide information about the best approach to adopt when planning interventions and in relation to the provision of education. The cognitive assets of both groups involved in the study were found in areas demanding logical thinking, reasoning and common sense, but it is known that such a cognitive pattern in children with some relatively good verbal abilities might easily mask a child s specific difficulty in the area of executive functions (3). According to Volkmar C 2007 The Author(s)/Journal Compilation C 2007 Foundation Acta Pædiatrica/Acta Pædiatrica , pp

5 Cognitive profile of children with ADHD Ek et al. and Klin (25), a higher crystallized or verbal intelligence may create unreasonable expectations of a very competent student. But when this verbal asset coexists with a low processing speed and/or poor working memory, the demands of everyday life imposed by teachers, parents and friends can create problems in school and at home. In this study, a representative group of almost to 11-year-old Swedish children in primary school were screened for attention, behavioural and learning problems. The two screening instruments used resulted in a large group of children being targeted. The attrition was rather limited during the initial screening procedure, and in the psychological assessments conducted later (<15% in total). According to parents and teachers reports, about 30% of all children in the population exhibited some type of attention, behavioural and/or learning problem. The prevalence of ADHD (5.4%), calculated for one birth cohort, accorded well with that of other published studies (9,26). The two groups in this study were cognitively very similar. This study confirms the statement that: most neuropsychological functions are usually viewed as a continuously varying trait (7) (p. 757). The cognitive similarities between the groups was striking, yet there appears to be a difference in their adaptation to behavioural challenges and everyday demands, as shown by their variable ADHD-group status. In this respect, it should be noted that behaviour and learning problems may have different levels of severity despite similar cognitive measures being obtained from tests. Despite almost identical cognitive deficits and assets, some children had more pervasive attention problems and met the criteria for ADHD or subthreshold ADHD. We strongly agree with Rucklidge and Tannock (13) that further research is necessary to continue to investigate how we can measure this construct we know as ADHD. LIMITATIONS In spite of the fact that this was a fairly representative sample of children with attention, behaviour and learning problem (the majority without an obvious learning disability) numbers are relatively small and conclusions can only be tentative. The number of girls included was particularly small, and, even though gender did influence ADHD diagnostic status, the findings in this respect should be interpreted with caution. ACKNOWLEDGEMENTS We gratefully acknowledge the contribution of Bengt Ramund, who was very helpful in providing comparison data from the Swedish WISC III normative sample of children of years of age and who assisted with the data analysis. The study was partly financed by the Centre for Competence in Treatment and Care. References 1. American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders, 4th ed. Washington, DC: American Psychiatric Association, Cabral P. Attention deficit disorders: are we barking up the wrong tree? Eur J Paediatr Neurol 2006; 10: Pennington BF, Ozonoff S. Executive functions and developmental psychopathology. J Child Psychol Psychiatr 1996; 37: Nydén A, Billstedt E, Hjelmquist E, Gillberg C. Neurocognitive stability in Asperger syndrome, ADHD, and reading and writing disorder: a pilot study. Dev Med Child Neurol 2001; 43: Gillberg C. Deficits in attention, motor control, and perception: a brief review. Arch Dis Child 2003; 88: Gillberg C, Söderstrom H. Learning disability. Lancet 2003; 362: Biederman J, Monuteaux MC, Doyle AE, Seidman LJ, Wilens TE, Ferrero F, et al. Impact of executive function deficits and attention-deficit/hyperactivity disorder (ADHD) on academic outcomes in children. J Consult Clin Psychol 2004; 72: Lawrence V, Houghton S, Douglas G, Durkin K, Whiting K, Tannock R. Executive function and ADHD: a comparison of children s performance during neuropsychological testing and real-world activities. J Att Dis 2004; 7: Landgren M, Pettersson R, Kjellman B, Gillberg C. ADHD, DAMP and other neurodevelopmental/psychiatric disorders in 6-year-old children: epidemiology and co-morbidity. Dev Med Child Neurol 1996; 38: Kuntsi J, Eley TC, Taylor A, Hughes C, Asherson P, Caspi A, et al. Co-occurrence of ADHD and low IQ has genetic origins. AmJMedGen2004; 1: Frazier TW, Demaree HA, Youngstrom EA. Meta-analysis of intellectual and neuropsychological test performance in attention-deficit/hyperactivity disorder. Neuropsychol 2004; 18: Mayes SD, Calhoun SL. The Gordon diagnostic system and WISC-III freedom from distractibility index: validity in identifying clinic-referred children with and without ADHD. Psychol Rep 2002; 91: Rucklidge JJ, Tannock R. Neuropsychological profiles of adolescents with ADHD: effects of reading difficulties and gender. J Child Psychol Psychiatr 2002; 43: Purvis KL, Tannock R. Language abilities in children with attention deficit hyperactivity disorder, reading disabilities and normal controls. J Abnorm Child Psychol 1997; 25: Tirosh E, Cohen A. Language deficit with attention-deficit disorder: a prevalent co-morbidity. J Child Neurol 1998; 13: Wechsler D. Wechsler intelligence scale for children. 3rd ed. Revised. New York: Psychological Corporation, Rourke BP, editor. Neuropsychology of learning disabilities. Essentials of subtype analyses. New York: Guilford Press, Thomson M. Monitoring dyslexics intelligence and attainments: a follow-up study. Dyslexia 2003; 9: Filippatou DN, Livaniou EA. Comorbidity and WISC-III profiles of Greek children with attention deficit hyperactivity disorder, learning disabilities, and language disorders. Psychol Rep 2005; 97: Lovecky DV. Different minds. Gifted children with AD/HD, Asperger syndrome, and other learning deficits. London: Jessica Kingsley, Ek U, Holmberg K, de Geer L, Swärd C, Fernell E. Behavioural and learning problems in schoolchildren related to cognitive test data. Acta Paediatr 2004; 93: Conners CK. A teacher rating scale for use in drug studies with children. Am J Psychiatr 1969; 126: C 2007 The Author(s)/Journal Compilation C 2007 Foundation Acta Pædiatrica/Acta Pædiatrica , pp

6 Ek et al. Cognitive profile of children with ADHD 23. Snow JB, Sapp GL. WISC-III subtest patterns of ADHD and normal samples. Psychol Rep 2000; 87: Schachar R, Tannock R. Syndromes of hyperactivity and attention deficits. In: Rutter S, Taylor E, editors. Child and Adolescent Psychiatry. Oxford:Blackwell Publishing, Volkmar FR, Klin A. Behavioral and learning problems in schoolchildren related to cognitive test data. Acta Paediatr 2004; 93: Cuffe SP, Moore CG, McKeown RE. Prevalence and correlates of ADHD symptoms in the national health interview survey. J Att Dis 2006; 9: C 2007 The Author(s)/Journal Compilation C 2007 Foundation Acta Pædiatrica/Acta Pædiatrica , pp