Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 7-Day Trial for You or Your Team.

Learn More →

Prevalence of the ADHD phenotype in 7- to 9-year-old children: effects of informant, gender and non-participation

Prevalence of the ADHD phenotype in 7- to 9-year-old children: effects of informant, gender and... Soc Psychiatry Psychiatr Epidemiol (2012) 47:763–769 DOI 10.1007/s00127-011-0379-3 OR IGINAL PAPER Prevalence of the ADHD phenotype in 7- to 9-year-old children: effects of informant, gender and non-participation • • Anne Karin Ullebø Maj-Britt Posserud • • Einar Heiervang Carsten Obel Christopher Gillberg Received: 13 December 2009 / Accepted: 31 March 2011 / Published online: 16 April 2011 The Author(s) 2011. This article is published with open access at Springerlink.com Abstract 6,237 children (66%) whose parents agreed to participate in Purpose To estimate the prevalence of the ADHD phe- the study. notype based on parent and teacher reports in a general Results The prevalence of the ADHD phenotype based on population sample of 7- to 9-year-old Norwegian children the combination of parent and teacher reports was 5.2% and evaluate the effect of parent attrition, gender and among participants. Teacher ratings of non-participants had informant on the prevalence estimate. a doubled rate of ADHD high scorers with an OR of 2.1 Methods The population consisted of all children (95% CI, 1.9–2.4). The non-participant ADHD high scorers (N = 9,430) attending 2nd–4th grade in the City of Ber- had more inattentive and fewer hyperactive/impulsive gen, Norway. The 18 symptoms of ADHD corresponding symptoms as compared to participating ADHD high scor- to the SNAP-IV and DSM-IV were included in the Bergen ers. Teachers reported high scores of hyperactivity/impul- Child Study questionnaire to teachers and parents. Teacher sivity and the combined symptom constellation much more information was available for 9,137 children (97%) and frequently in boys than girls, while the difference between information from both informants was available for the genders was less marked according to parent reports. Conclusions The ADHD phenotype was twice as pre- valent among non-participants as among participants. Reported prevalences in population studies are therefore A. K. Ullebø  M.-B. Posserud likely to be underestimates, if such attrition bias is not Centre for Child and Adolescent Mental Health, accounted for. Choice of informant, criteria for symptom Unifob Health, Bergen, Norway count, definitions of subtypes and gender differences E. Heiervang influence the prevalence estimates of the ADHD phenotype. Institute of Clinical Medicine, University of Oslo, Oslo, Norway Keywords Attention-deficit/hyperactivity disorder C. Obel Child psychiatry  Epidemiology  Attrition  Gender Department of Epidemiology, Institute of Public Health, Aarhus University, Aarhus, Denmark Introduction C. Obel Department of General Practice, Institute of Public Health, Aarhus University, Aarhus, Denmark In spite of decades of research, the prevalence of attention- deficit/hyperactivity disorder (ADHD) has been difficult to C. Gillberg estimate and it is still a matter of controversy how frequent Gillberg Neuropsychiatry Centre, Sahlgrenska Academy, University of Gothenburg, Gothenburg, Sweden this phenotype is in a general population setting [6]. Some of the discrepancies may be caused by cultural and social A. K. Ullebø (&) differences, acting on both the prevalence directly and on RBUP Vest, Uni Helse, PO Box 7810, the reporting style. This may be the reason for the somewhat 5020 Bergen, Norway lower prevalence rates of ADHD found in the Scandinavian e-mail: [email protected] 123 764 Soc Psychiatry Psychiatr Epidemiol (2012) 47:763–769 countries [12]. There is also a wide variation as regards reproducible method to measure the ADHD phenotype in measures and sample characteristics [6]. Optimally, the the community. population prevalence should reflect the total population, but in practice it has been difficult to establish a level of study participation that makes the sample representative. Materials and methods Those who participate do not represent a random sample and this differentiated attrition biases the prevalence esti- The Bergen Child Study mates of child psychiatric disorders such as ADHD. Parents of children rated as deviant by teachers have been found to All data came from the first (screening) stage of the first be less likely to consent to research on child psychiatric wave of the Bergen Child Study (BCS) [13]. The target disorders compared to parents of children rated within the population comprised all 9,430 children in the 2nd–4th normal range [16]. In a previous publication from the grade of all schools in the City of Bergen, Norway, in Bergen Child Study (BCS), the impact of non-responder October 2002. An informed consent form and a detailed bias on the prevalence of several different child mental four-page questionnaire were sent to parents through the health problems was explored and an important finding was schools, and similar questionnaires were distributed to that teachers rated non-responders higher on all symptom teachers. Teachers were asked to complete the questionnaire scales, except tics, and as more impaired than responders for every child in every class. If the parent consent form was [18]. Teacher high scores (75, 90 and 95th percentiles) on returned to the school, teachers identified the corresponding inattention and/or hyperactivity had significantly increased teacher questionnaire through the identification code (ID relative risk for parental non-response. Yet we know little number) provided on the parent consent form. If no parent about the quantitative effect this would have on the esti- consent was provided, the completed teacher questionnaire mation of ADHD prevalence. Another important issue of was returned without any personal identification, other than non-response is whether high scorers in the non-partici- child’s gender and grade. No information about school or pating group might be qualitatively different from high teacher was given, making the children untraceable. For scorers in the participating group with respect to symptom 9,137 children (96.9%), full teacher information on ADHD constellation and/or severity. Such bias could lead to symptoms was obtained. For 6,237 children (66.1%), we had important misinterpretation of results in the further stages of full information from teachers and parents (Fig. 1). The the study where clinical measures are applied and one seeks study was approved by the Western Norway Regional knowledge about clinical conditions in a representative Committee for Medical and Health Research Ethics and the sample from the general population. Few previous studies Norwegian Data Inspectorate. In the present paper, the ‘‘Full Data group’’ refers to have had access to data for non-participants, and if such data have been available, it has included only demographics participants, i.e., children with parent consent for whom such as living area, ethnicity, age and gender. both parent and teacher information on ADHD symptoms Other important factors that influence the prevalence was available. The ‘‘Anonymous Data group’’ refers to estimate in ADHD include the definition applied, symptom non-participants, i.e., children for whom only anonymous count, use of impairment, cross-situational criteria and teacher information was available. Lacking a teacher choice of informant. As there is a wide variety of defini- questionnaire was mainly due to long term sick-leave of the tions, measures, informants and samples [6], a better teacher or missing data on ADHD items (N = 3.1%). The understanding of the factors that influence prevalence children lacking teacher questionnaires were excluded estimates is important when interpreting differences from further analyses. Thus, the non-participants referred between studies. to in this paper are defined as the children in the Anony- The aims of the present study were (1) to estimate the mous Data group. Also, among the participants there was a prevalence of the ADHD phenotype in a general child group of children with missing parent information on the population, based on parent and teacher reports, and (2) to ADHD items (N = 361), thus not contributing to the Full analyze the effect of parental attrition, informant and Data group (Fig. 1). gender on ADHD prevalence. We report the prevalence of the ADHD phenotype based The questionnaire on reported symptoms from questionnaires and making no correction for level of impairment, while acknowledging The BCS screening questionnaire included several mea- that a clinical diagnosis cannot be based on questionnaire sures of child mental health [13]. The present data analyses data only. For clinical purposes, the impairment of the were based on the 18 ADHD symptoms specified in the symptoms is crucial, but for epidemiological purposes and DSM-IV criteria for ADHD [3]. The wording of the items comparison with other studies we rely on this readily was consistent with the SNAP-IV [19], but each item was 123 Soc Psychiatry Psychiatr Epidemiol (2012) 47:763–769 765 Fig. 1 Flowchart of the sample in contrast to the original version of SNAP-IV scored by questionnaire [13]. The impact score applied here was parents and teachers on a 3-point Likert scale: 0 for ‘‘Not based on three teacher report items relating to distress, true’’, 1 for ‘‘Somewhat true’’ and 2 for ‘‘Certainly true’’. A impairment and burden with scores ranging from 0 to 6. score of 1 or 2 was defined as the presence of a symptom. In agreement with the symptom count according to DSM- Statistics IV, the threshold for the definition of ADHD high scorers was set at 6/9 symptoms (‘‘somewhat true’’ or ‘‘certainly To estimate the full population prevalence, we assumed the true’’) on at least one subscale. According to our definition, same ratio between teacher and parent high scorers in the the ADHD combined high scorers had C6 symptoms on Anonymous group as in the Full Data group. We then each of the two subscales. The inattention ADHD high multiplied the ADHD prevalence from the Full Data group scorers had C6 symptoms on the inattention subscale only, with the ratio ADHD high scorers teacher non-participants/ and the ADHD hyperactivity/impulsivity high scorers had ADHD high scorers teacher participants. The confidence six or more symptoms on the hyperactivity/impulsivity interval (CI) of the prevalence estimate was calculated subscale only. The ADHD phenotype was defined as being according to the formula for the 95% CI for a proportion a high scorer according to both informants. High scorers on p (95% CI, p ± 1.96 9 SE(p) where SE(p) = p 9 (1 - inattention according to both informants were defined as p)/Hn). having the inattentive subtype (ADHD-I). High scorers on High scorer prevalence in the Full Data group versus the hyperactivity/impulsivity according to both informants Anonymous Data group was assessed with odds ratio (OR) were defined as having the hyperactive–impulsive subtype estimate with 95% confidence interval (95% CI). Gender (ADHD-H). Children defined as having ADHD combined difference in high scorer prevalence according to each subtype (ADHD-Co) were either high scorers on each of informant was evaluated comparing the ORs. Mean group the two subscales according to both informants, or high differences were assessed with two-tailed t tests and dif- scorers on different subscales according to the two infor- ferences in proportions of ADHD symptom subtypes with mants. Thus, it was possible to fulfill our criteria for the Chi-square (v ) analyses. Agreement across informants was combined phenotype with, i.e., 6/9 (or more) symptoms on evaluated using Cohen’s kappa (j). The level of statistical the inattention subscale on parent report and 6/9 symptoms significance was set at 0.05. We used the software package on the hyperactivity/impulsivity subscale on teacher report. SPSS 15 [17]. We found this definition of ADHD-Co to be appropriate given that these children were high scorers according to both informants and they had a symptom count above the Results threshold within each of the domains. For the comparison of participant teacher ADHD high Prevalence of the ADHD phenotype scorers and non-participant teacher ADHD high scorers, we also applied the impact supplement part of the SDQ In the Full Data group, the prevalence of the ADHD phe- (http://www.SDQinfo.org), which was included in the BCS notype was 5.2% (95% CI, 5.1–5.3%). For all subtypes, 123 766 Soc Psychiatry Psychiatr Epidemiol (2012) 47:763–769 Table 1 Prevalence of the ADHD phenotype according to subtype The Anonymous Data group had both higher mean and informant symptom scores and a higher frequency of high scorers compared to the Full Data group (Table 2). ADHD-Co ADHD-I ADHD-H Any ADHD (%) (%) (%) (%) When analyses reported in Table 2 were repeated for boys only and girls only, respectively, the pattern of dif- Parents 4.1 6.5 2.8 13.4 ferences between the two groups remained the same, Teachers 3.1* 5.4* 1.9* 10.4* except the difference in frequency of combined high Both 1.2 1.6 0.3 5.2 scorers between the participants and non-participants, * p \ 0.001, for difference in prevalence between parent and teacher which was no longer significant for girls (v = 4.37, reports p = 0.37). There were no significant differences in the proportion of girls relative to boys among the participant ADHD high parents reported significantly more children as ADHD high scorers compared to the non-participant high scorers 2 2 scorers than did teachers (Table 1). Note that the ADHD- (v = 1.7, p = 0.19) or any age differences (v = 0.91, Co phenotype comprises not only the 1.2% high scorers p = 0.32). The non-participant ADHD high scorers shown in Table 1 having the full combined symptom showed more inattention symptoms and less hyperactivity/ constellation from both informants, but also the 2.1% of inattention symptoms, both as sum scores and as symptom children rated as high scorers in different symptom counts, than the participants. There was no significant domains by parents and teachers (e.g., as inattentive by one difference between the two groups on the sum score of the informant and hyperactive by the other). impact measure (Table 3). Informant agreement Gender by informant effects Agreement between parents and teachers was r = 0.32 for Boys had significantly higher mean scores than girls on all ADHD combined high scorers, r = 0.22 for the ADHD subscales (p \ 0.001 for all comparisons) (Table 4). inattention only high scorers, and r = 0.13 for ADHD The gender (boy:girl) OR (95% CI) for ADHD high hyperactivity/impulsivity only high scorers. Agreement on scorers according to each informant and ADHD symptom ADHD high scorers for any subscale was r = 0.37. domain are summarized in Table 5. The ORs show increased risk for boys among high scorers for all ADHD Effects of attrition symptom domains and according to each informant. For ADHD combined high scorers and ADHD hyperactivity/ The frequency of teacher ADHD high scorers in the impulsivity high scorers, teacher reports yielded higher Anonymous Data group was 19.9% compared to 10.4% in gender (boy:girl) ORs than parent reports (CIs not the Full Data group alone. Combining the Anonymous overlapping). Data group and the Full Data group (N = 9 137), 13.1% were teacher high scorers. Assuming the ratio between teacher reported ADHD high scorers and the ADHD phe- Discussion notype (which is defined based on high scores from both informants) in the whole sample to be the same as in the We found twice as many children with the ADHD phe- Full Data group (5.2:10.4%), the estimate for the ADHD notype among children in the Anonymous Data group phenotype in the total population would be 6.6% (95% CI, versus the Full Data group, which demonstrated that 6.0–7.2%). Table 2 Mean teacher scores Full Data Anonymous Difference (95% CI) and frequency of the teacher ADHD high scorer (6/9) ADHD symptoms subtypes in the Full Data group Mean sum score 2.7 4.2 1.5 (1.25–1.75)* (N = 6,237) and the Mean inattention subscale 1.6 2.6 1.0 (0.85–1.15)* Anonymous Data group (N = 2,539) Mean hyperactivity subscale 1.1 1.6 0.5 (0.37–0.62)* High scorers v values (p value) Combined 3.1% 5.7% 32.5* Inattention only 5.4% 11.9% 111.4* Hyperactive/impulsive only 1.9% 2.3% 1.70 (p = 0.192) * p \ 0.001 for difference 123 Soc Psychiatry Psychiatr Epidemiol (2012) 47:763–769 767 Table 3 Mean sum scores and number of symptoms on teacher reports for ADHD high scorers in the Full Data group (N = 648) and the Anonymous Data group (N = 504) Impact sum Hyperactivity Inattention Hyperactivity Inattention score sum score sum score symptoms symptoms Full Data 1.45 6.12 8.46 4.63 6.53 Anonymous Data 1.54 (p = 0.450) 5.50 (p = 0.028) 8.98 (p = 0.020) 4.14 (p = 0.007) 6.90 (p = 0.002) For eight children in each group impact score was missing prevalence rate in the BCS was considerably lower than in Table 4 Mean ADHD symptom sum scores for each gender and a comparable British survey in a head-to-head comparison informant of the two samples with similar age groups and informants Inattentive Hyperactive/impulsive [12]. Parents Teachers Parents Teachers Our prevalence estimate of 5.2% for the ADHD pheno- type is comparable to that found in a recent German study Boys 2.75 2.34 2.03 1.70 [5] reporting a prevalence of 6.4% in the same age group. Girls 1.66* 0.88* 1.26* 0.45* Our prevalence estimate relied on two informants (which N = 6,237 (boys, 3,107; girls, 3,130) led to a decrease in prevalence), whereas the German study * p \ 0.001 for difference between boys and girls mean scores only included parent reports. On the other hand, the study included a 4-point response scale and the two most deviant responses were regarded as indicating the presence of ‘‘symptom’’. This is probably a more conservative symptom Table 5 Gender (boy:girl) ORs (95% CI) for the ADHD high scorer definition than ours, given that we had only three response subtypes for each informant categories and defined the two most deviant as indicating Combined Inattention Hyperactive/impulsive symptom. Observing the behavior in different settings Parent 2.9 (2.2–3.9) 2.1 (1.7–2.6) 1.7 (1.3–2.4) diminishes the likeliness of mixing it up with other behavioral disorders. The German study included no Teacher 6.2 (4.2–9.3) 3.1 (2.4–3.9) 5.4 (3.3–8.8) adjustment for non-responders, meaning that their preva- N = 6,237 (boys, 3,107; girls, 3,130) lence rate was probably also an underestimate. Given these important methodological differences, it is somewhat sur- prising that the prevalence estimates are in the same range. This is not to be taken as support for a more solid evidence attrition in studies with a typical attrition rate underesti- mated the ADHD phenotype prevalence. We estimated the basis—that at the end of the day the reported prevalence rates were very similar. This may rather reflect that the prevalence of the ADHD phenotype to 5.2% (parent and teacher reports) among children whose parents consented choices made in a study may be influenced by previously to participate in the study, but 6.6% in the total population. reported results. This also demonstrates the liability of Both parents and teachers reported more ADHD symptoms prevalence estimates to definition and the importance of in boys than in girls, but the gender difference was greater thorough characterization of the methodology applied when referring to any reported prevalence of ADHD. according to teacher reports. The excess proportion of boys with hyperactivity/impulsivity and the combined symptom The access to anonymous teacher questionnaires for most of the non-participants was a special asset of our constellation high score was higher according to teacher reports than parent reports. Informant agreement was low study. Comparing participants to non-participants, a much higher level of ADHD symptoms was found in the latter to fair. The estimated ADHD phenotype prevalence of 5.2% in group and this finding is relevant for all population-based epidemiological studies independent of their definition of our study was considerably higher than the DAWBA-based ADHD prevalence of 1.3% from a second study phase in ADHD. Similar trends have been reported for autistic the same population based on the Development and Well- symptoms in the same cohort [15]. Teacher reports show- Being Assessment (DAWBA) [13]. This is not unexpected ing a prevalence of 19.9% ADHD high scorers in the given that a DAWBA diagnosis requires the impairment Anonymous Data group compared to 10.4% in the Full Data group (an OR of 2.1) clearly illustrate the very criteria to be fulfilled and is therefore more comparable to a clinical diagnosis. Interestingly, our prevalence estimate important effect of non-participation in population studies of ADHD symptomatology. The non-participant ADHD for the ADHD phenotype was in the range of that reported from similar studies, while the DAWBA ADHD high scorers did not significantly differ from the participant 123 768 Soc Psychiatry Psychiatr Epidemiol (2012) 47:763–769 ADHD high scorers in boy:girl ratio, age or on the impact overidentification by parents. Boys with ADHD are measure. Thus, we did not get any support for the reported to engage in more rule breaking and externalizing hypothesis that teacher-rated non-participant children with behavior than girls with ADHD [2], and this has been ADHD symptoms would be more impaired than partici- found to affect teacher ratings of ADHD [1]. Some authors pants. Teachers completed the questionnaires without any have found support for the hypothesis that the difference in knowledge of who would later belong to the non-partici- symptom ratings across informants could be due to real pant group. However, one could suspect that they might situational differences [7]. Although the cause of the dif- have had a pre-conceived idea of who was going to par- ference in parent and teacher reports on ADHD symptoms ticipate or not. Interestingly, the non-participant ADHD in girls remains unresolved, it is important to bear this in high scorers had higher inattention scores and less hyper- mind and to explore the issue further in future studies. activity/impulsivity than the participant ADHD high scor- The BCS is unique in that teacher questionnaires cover ers. The explanation for this finding is speculative as we 97% of the total population. The current study focused on lack comparable reports from other studies. The finding symptoms of ADHD as reported on questionnaires. The underscores the importance of trying to assess non- validity of such reports may be questioned, since infor- responder bias in epidemiology in general and in psychi- mants may misunderstand items, and may also have rea- atric research specifically. Though generally assumed that sons for over- and underreporting problems in the child. the non-participants are at higher risk for mental disorders Also, the DSM-IV diagnostic criteria require an early onset and less privileged socially, few studies have explored the of the disorder (before age 7) and pervasive impairment non-participation in sufficient detail to characterize the from the symptoms. Thus, the phenotype and subtypes possible heterogeneity of non-participation. Investigating referred to here only indicate symptom constellations as selective participation in the British Child and Adolescent specified in the diagnostic criteria and are not comparable Mental Health Surveys, Goodman and Gatward reported to a clinical diagnosis. However, the symptom count important heterogeneity in the effect of deprivation on approach may be more readily reproducible than clinical parental non-participation [9]. Thus, it is important to note diagnoses in epidemiological research. that the process of non-participation is probably compli- The use of only three response categories represented a cated with a heterogeneous set of reasons, which give rise problem in the current study. It is not clear whether the to diverse effects on the non-participating group. middle category should be regarded as having the symptom We reported an estimate of the influence of attrition on or not. Many DSM-IV ADHD rating scales have used the ADHD phenotype prevalence estimate by assuming 4-point scales, where the two highest scores have been that the hypothetical parent reports of the children in the interpreted as indicating a symptom [4, 8, 21]. However, our prevalence of the ADHD phenotype according to tea- Anonymous group would have related to teacher reports at the same high scorer ratio as in the Full Data group. More cher reports is comparable to figures reported in previous sophisticated methods taking account of the differential studies of teacher-reported DSM-IV ADHD [4, 8, 15, 21]. parent–teacher agreement across number of symptoms for Similarly, the frequency of parent-reported ADHD symp- the high scorers or bootstrap methods might have been tom subtypes was comparable to that found by other used to estimate the effect of attrition on the total popu- studies using parent information [10]. Our use of strict lation prevalence. However, as discussed above, there are cross informant criteria compensated for a somewhat less several different uncertainties and limitations attached to restrictive individual symptom definition in estimating the the prevalence estimate (such as the differential use of prevalence based on both informants. A more conservative impact, etc.) that in the end we opted for illustrating the definition of symptom presence would have been to count non-response effect by this simple method as the inter- only ‘‘Certainly true’’ answers as symptom present. We pretation of this estimate is straightforward. We underline considered that the somewhat more inclusive symptom the importance of evaluating each aspect of the various criteria were suitable for the epidemiological consider- methodological influences rather than taking any one ations in this general child population study. prevalence estimate as reflective of the ‘‘true’’ rate. Our reported boy:girl ratios for ADHD high scorers on DSM-IV symptoms are in the range of earlier studies in Conclusion community samples [4, 8, 10, 11, 14, 20, 21]. Parents identified more girls than teachers, a finding that has been The prevalence of the ADHD phenotype based on teacher reported for the hyperactive/impulsive and for the com- and parent-reported symptomatology was clearly influ- bined subtypes in a previous study [8], but it is not clear enced by non-participation. The non-participation not only whether the higher number of girls identified by parents led to an underestimation of the prevalence, but also represent an underidentification by teachers or an affected the rates of inattention and hyperactivity/ 123 Soc Psychiatry Psychiatr Epidemiol (2012) 47:763–769 769 8. Gomez R, Harvey J, Quick C, Scharer I, Harris G (1999) DSM- impulsivity. The definition of the phenotype, gender and IV AD/HD: confirmatory factor models, prevalence, and gender choice of informant also influenced the detailed epidemio- and age differences based on parent and teacher ratings of Aus- logy of the ADHD phenotype in the present study. tralian primary school children. J Child Psychol Psychiatry 40:265–274 Acknowledgments We are grateful for the participation of children, 9. Goodman A, Gatward R (2008) Who are we missing? Area parents and teachers in the study. We thank Stein Atle Lie and Tore deprivation and survey participation. Eur J Epidemiol Wentzel-Larsen for statistical advice and to Jim Stevenson and Astri 23:379–387 J. Lundervold for important comments on the manuscript. This study 10. Graetz B, Sawyer M, Baghurst P (2005) Gender differences was supported by the Norwegian Research Council, the Norwegian among children with DSM-IV ADHD in Australia. J Am Acad Directorate of Health, the Western Regional Health Authority and the Child Adolesc Psychiatry 44:159–168 Centre for Child and Adolescent Mental Health, Uni Health, Bergen. 11. Graetz B, Sawyer M, Hazell P, Arney F, Baghurst P (2001) Christopher Gillberg was funded by the Swedish Medical Research Validity of DSM-IV ADHD subtypes in a nationally represen- Council. The work of Carsten Obel was funded by the Nordic Council tative sample of Australian children and adolescents. J Am Acad of Ministers research program ‘Longitudinal Epidemiology’ Child Adolesc Psychiatry 40:1410–1417 (020056). 12. Heiervang E, Goodman A, Goodman R (2008) The Nordic advantage in child mental health: separating health differences Open Access This article is distributed under the terms of the from reporting style in a cross-cultural comparison of psycho- Creative Commons Attribution Noncommercial License which per- pathology. J Child Psychol Psychiatry 49:678–685 mits any noncommercial use, distribution, and reproduction in any 13. Heiervang E, Stormark KM, Lundervold AJ, Heimann M, medium, provided the original author(s) and source are credited. Goodman R, Posserud MB, Ullebo AK, Plessen KJ, Bjelland I, Lie SA, Gillberg C (2007) Psychiatric disorders in Norwegian 8- to 10-year-olds: an epidemiological survey of prevalence, risk factors, and service use. J Am Acad Child Adolesc Psychiatry 46:438–447 References 14. Nolan E, Gadow K, Sprafkin J (2001) Teacher reports of DSM- IV ADHD, ODD, and CD symptoms in schoolchildren. J Am 1. Abikoff HB, Courtney M, Pelham W, Koplewicz H (1993) Acad Child Adolesc Psychiatry 40:241–249 Teachers’ ratings of disruptive behaviors: the influence of halo 15. Posserud M, Lundervold A, Gillberg C (2006) Autistic features in effects. J Abnorm Child Psychol 21:519–533 a total population of 7–9-year-old children assessed by the ASSQ 2. Abikoff HB, Jensen PS, Arnold L, Hoza B, Hechtman L, Pollack (Autism Spectrum Screening Questionnaire). J Child Psychol S, Martin D, Alvir J, March JS, Hinshaw S, Vitiello B, Newcorn Psychiatry 47:167–175 J, Greiner A, Cantwell DP, Conners CK, Elliott G, Greenhill LL, 16. Rutter M, Cox A, Tupling C, Berger M, Yule W (1975) Attain- Kraemer H, Pelham WE, Severe JB, Swanson JM, Wells K, ment and adjustment in two geographical areas. I. The prevalence Wigal T (2002) Observed classroom behavior of children with of psychiatric disorder. Br J Psychiatry 126:493–509 ADHD: relationship to gender and comorbidity. J Abnorm Child 17. SPSS I (1993) SPSS Advanced Statistics. SPSS, Chicago Psychol 30:349–359 18. Stormark KM, Heiervang E, Heimann M, Lundervold A, Gillberg 3. American Psychiatric Association (1994) Diagnostic and statis- C (2008) Predicting nonresponse bias from teacher ratings of tical manual of mental disorders: DSM-IV. American Psychiatric mental health problems in primary school children. J Abnorm Association, Washington Child Psychol 36:411–419 4. Carlson C, Tamm L, Gaub M (1997) Gender differences in 19. Swanson JM, Kraemer HC, Hinshaw SP, Arnold LE, Conners children with ADHD, ODD, and co-occurring ADHD/ODD CK, Abikoff HB, Clevenger W, Davies M, Elliott GR, Greenhill identified in a school population. J Am Acad Child Adolesc LL, Hechtman L, Hoza B, Jensen PS, March JS, Newcorn JH, Psychiatry 36:1706–1714 Owens EB, Pelham WE, Schiller E, Severe JB, Simpson S, Vi- 5. Dopfner M, Breuer D, Wille N, Erhart M, Ravens-Sieberer U tiello B, Wells K, Wigal T, Wu M (2001) Clinical relevance of (2008) How often do children meet ICD-10/DSM-IV criteria of the primary findings of the MTA: success rates based on severity attention deficit-/hyperactivity disorder and hyperkinetic disor- of ADHD and ODD symptoms at the end of treatment. J Am der? Parent-based prevalence rates in a national sample—results Acad Child Adolesc Psychiatry 40:168–179 of the BELLA study. Eur Child Adolesc Psychiatry 17(Suppl 20. Szatmari P, Offord D, Boyle M (1989) Ontario Child Health 1):59–70 Study: prevalence of attention deficit disorder with hyperactivity. 6. Faraone SV, Sergeant J, Gillberg C, Biederman J (2003) The J Child Psychol Psychiatry 30:219–230 worldwide prevalence of ADHD: is it an American condition? 21. Wolraich M, Hannah J, Pinnock T, Baumgaertel A, Brown J World Psychiatry 2:104–113 (1996) Comparison of diagnostic criteria for attention-deficit 7. Gomez R (2007) Australian parent and teacher ratings of the hyperactivity disorder in a county-wide sample. J Am Acad Child DSM-IV ADHD symptoms: differential symptom functioning Adolesc Psychiatry 35:319–324 and parent–teacher agreement and differences. J Atten Disord 11:17–27 http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Social Psychiatry and Psychiatric Epidemiology Pubmed Central

Prevalence of the ADHD phenotype in 7- to 9-year-old children: effects of informant, gender and non-participation

Ullebø, Anne Karin; Posserud, Maj-Britt; Heiervang, Einar; Obel, Carsten; Gillberg, Christopher
Social Psychiatry and Psychiatric Epidemiology , Volume 47 (5) – Apr 16, 2011
Download PDF
7 pages

Loading next page...
 
/lp/pubmed-central/prevalence-of-the-adhd-phenotype-in-7-to-9-year-old-children-effects-GMZDxQhvkn

References (24)

Publisher
Pubmed Central
Copyright
© The Author(s) 2011
ISSN
0933-7954
eISSN
1433-9285
DOI
10.1007/s00127-011-0379-3
Publisher site
See Article on Publisher Site

Abstract

Soc Psychiatry Psychiatr Epidemiol (2012) 47:763–769 DOI 10.1007/s00127-011-0379-3 OR IGINAL PAPER Prevalence of the ADHD phenotype in 7- to 9-year-old children: effects of informant, gender and non-participation • • Anne Karin Ullebø Maj-Britt Posserud • • Einar Heiervang Carsten Obel Christopher Gillberg Received: 13 December 2009 / Accepted: 31 March 2011 / Published online: 16 April 2011 The Author(s) 2011. This article is published with open access at Springerlink.com Abstract 6,237 children (66%) whose parents agreed to participate in Purpose To estimate the prevalence of the ADHD phe- the study. notype based on parent and teacher reports in a general Results The prevalence of the ADHD phenotype based on population sample of 7- to 9-year-old Norwegian children the combination of parent and teacher reports was 5.2% and evaluate the effect of parent attrition, gender and among participants. Teacher ratings of non-participants had informant on the prevalence estimate. a doubled rate of ADHD high scorers with an OR of 2.1 Methods The population consisted of all children (95% CI, 1.9–2.4). The non-participant ADHD high scorers (N = 9,430) attending 2nd–4th grade in the City of Ber- had more inattentive and fewer hyperactive/impulsive gen, Norway. The 18 symptoms of ADHD corresponding symptoms as compared to participating ADHD high scor- to the SNAP-IV and DSM-IV were included in the Bergen ers. Teachers reported high scores of hyperactivity/impul- Child Study questionnaire to teachers and parents. Teacher sivity and the combined symptom constellation much more information was available for 9,137 children (97%) and frequently in boys than girls, while the difference between information from both informants was available for the genders was less marked according to parent reports. Conclusions The ADHD phenotype was twice as pre- valent among non-participants as among participants. Reported prevalences in population studies are therefore A. K. Ullebø  M.-B. Posserud likely to be underestimates, if such attrition bias is not Centre for Child and Adolescent Mental Health, accounted for. Choice of informant, criteria for symptom Unifob Health, Bergen, Norway count, definitions of subtypes and gender differences E. Heiervang influence the prevalence estimates of the ADHD phenotype. Institute of Clinical Medicine, University of Oslo, Oslo, Norway Keywords Attention-deficit/hyperactivity disorder C. Obel Child psychiatry  Epidemiology  Attrition  Gender Department of Epidemiology, Institute of Public Health, Aarhus University, Aarhus, Denmark Introduction C. Obel Department of General Practice, Institute of Public Health, Aarhus University, Aarhus, Denmark In spite of decades of research, the prevalence of attention- deficit/hyperactivity disorder (ADHD) has been difficult to C. Gillberg estimate and it is still a matter of controversy how frequent Gillberg Neuropsychiatry Centre, Sahlgrenska Academy, University of Gothenburg, Gothenburg, Sweden this phenotype is in a general population setting [6]. Some of the discrepancies may be caused by cultural and social A. K. Ullebø (&) differences, acting on both the prevalence directly and on RBUP Vest, Uni Helse, PO Box 7810, the reporting style. This may be the reason for the somewhat 5020 Bergen, Norway lower prevalence rates of ADHD found in the Scandinavian e-mail: [email protected] 123 764 Soc Psychiatry Psychiatr Epidemiol (2012) 47:763–769 countries [12]. There is also a wide variation as regards reproducible method to measure the ADHD phenotype in measures and sample characteristics [6]. Optimally, the the community. population prevalence should reflect the total population, but in practice it has been difficult to establish a level of study participation that makes the sample representative. Materials and methods Those who participate do not represent a random sample and this differentiated attrition biases the prevalence esti- The Bergen Child Study mates of child psychiatric disorders such as ADHD. Parents of children rated as deviant by teachers have been found to All data came from the first (screening) stage of the first be less likely to consent to research on child psychiatric wave of the Bergen Child Study (BCS) [13]. The target disorders compared to parents of children rated within the population comprised all 9,430 children in the 2nd–4th normal range [16]. In a previous publication from the grade of all schools in the City of Bergen, Norway, in Bergen Child Study (BCS), the impact of non-responder October 2002. An informed consent form and a detailed bias on the prevalence of several different child mental four-page questionnaire were sent to parents through the health problems was explored and an important finding was schools, and similar questionnaires were distributed to that teachers rated non-responders higher on all symptom teachers. Teachers were asked to complete the questionnaire scales, except tics, and as more impaired than responders for every child in every class. If the parent consent form was [18]. Teacher high scores (75, 90 and 95th percentiles) on returned to the school, teachers identified the corresponding inattention and/or hyperactivity had significantly increased teacher questionnaire through the identification code (ID relative risk for parental non-response. Yet we know little number) provided on the parent consent form. If no parent about the quantitative effect this would have on the esti- consent was provided, the completed teacher questionnaire mation of ADHD prevalence. Another important issue of was returned without any personal identification, other than non-response is whether high scorers in the non-partici- child’s gender and grade. No information about school or pating group might be qualitatively different from high teacher was given, making the children untraceable. For scorers in the participating group with respect to symptom 9,137 children (96.9%), full teacher information on ADHD constellation and/or severity. Such bias could lead to symptoms was obtained. For 6,237 children (66.1%), we had important misinterpretation of results in the further stages of full information from teachers and parents (Fig. 1). The the study where clinical measures are applied and one seeks study was approved by the Western Norway Regional knowledge about clinical conditions in a representative Committee for Medical and Health Research Ethics and the sample from the general population. Few previous studies Norwegian Data Inspectorate. In the present paper, the ‘‘Full Data group’’ refers to have had access to data for non-participants, and if such data have been available, it has included only demographics participants, i.e., children with parent consent for whom such as living area, ethnicity, age and gender. both parent and teacher information on ADHD symptoms Other important factors that influence the prevalence was available. The ‘‘Anonymous Data group’’ refers to estimate in ADHD include the definition applied, symptom non-participants, i.e., children for whom only anonymous count, use of impairment, cross-situational criteria and teacher information was available. Lacking a teacher choice of informant. As there is a wide variety of defini- questionnaire was mainly due to long term sick-leave of the tions, measures, informants and samples [6], a better teacher or missing data on ADHD items (N = 3.1%). The understanding of the factors that influence prevalence children lacking teacher questionnaires were excluded estimates is important when interpreting differences from further analyses. Thus, the non-participants referred between studies. to in this paper are defined as the children in the Anony- The aims of the present study were (1) to estimate the mous Data group. Also, among the participants there was a prevalence of the ADHD phenotype in a general child group of children with missing parent information on the population, based on parent and teacher reports, and (2) to ADHD items (N = 361), thus not contributing to the Full analyze the effect of parental attrition, informant and Data group (Fig. 1). gender on ADHD prevalence. We report the prevalence of the ADHD phenotype based The questionnaire on reported symptoms from questionnaires and making no correction for level of impairment, while acknowledging The BCS screening questionnaire included several mea- that a clinical diagnosis cannot be based on questionnaire sures of child mental health [13]. The present data analyses data only. For clinical purposes, the impairment of the were based on the 18 ADHD symptoms specified in the symptoms is crucial, but for epidemiological purposes and DSM-IV criteria for ADHD [3]. The wording of the items comparison with other studies we rely on this readily was consistent with the SNAP-IV [19], but each item was 123 Soc Psychiatry Psychiatr Epidemiol (2012) 47:763–769 765 Fig. 1 Flowchart of the sample in contrast to the original version of SNAP-IV scored by questionnaire [13]. The impact score applied here was parents and teachers on a 3-point Likert scale: 0 for ‘‘Not based on three teacher report items relating to distress, true’’, 1 for ‘‘Somewhat true’’ and 2 for ‘‘Certainly true’’. A impairment and burden with scores ranging from 0 to 6. score of 1 or 2 was defined as the presence of a symptom. In agreement with the symptom count according to DSM- Statistics IV, the threshold for the definition of ADHD high scorers was set at 6/9 symptoms (‘‘somewhat true’’ or ‘‘certainly To estimate the full population prevalence, we assumed the true’’) on at least one subscale. According to our definition, same ratio between teacher and parent high scorers in the the ADHD combined high scorers had C6 symptoms on Anonymous group as in the Full Data group. We then each of the two subscales. The inattention ADHD high multiplied the ADHD prevalence from the Full Data group scorers had C6 symptoms on the inattention subscale only, with the ratio ADHD high scorers teacher non-participants/ and the ADHD hyperactivity/impulsivity high scorers had ADHD high scorers teacher participants. The confidence six or more symptoms on the hyperactivity/impulsivity interval (CI) of the prevalence estimate was calculated subscale only. The ADHD phenotype was defined as being according to the formula for the 95% CI for a proportion a high scorer according to both informants. High scorers on p (95% CI, p ± 1.96 9 SE(p) where SE(p) = p 9 (1 - inattention according to both informants were defined as p)/Hn). having the inattentive subtype (ADHD-I). High scorers on High scorer prevalence in the Full Data group versus the hyperactivity/impulsivity according to both informants Anonymous Data group was assessed with odds ratio (OR) were defined as having the hyperactive–impulsive subtype estimate with 95% confidence interval (95% CI). Gender (ADHD-H). Children defined as having ADHD combined difference in high scorer prevalence according to each subtype (ADHD-Co) were either high scorers on each of informant was evaluated comparing the ORs. Mean group the two subscales according to both informants, or high differences were assessed with two-tailed t tests and dif- scorers on different subscales according to the two infor- ferences in proportions of ADHD symptom subtypes with mants. Thus, it was possible to fulfill our criteria for the Chi-square (v ) analyses. Agreement across informants was combined phenotype with, i.e., 6/9 (or more) symptoms on evaluated using Cohen’s kappa (j). The level of statistical the inattention subscale on parent report and 6/9 symptoms significance was set at 0.05. We used the software package on the hyperactivity/impulsivity subscale on teacher report. SPSS 15 [17]. We found this definition of ADHD-Co to be appropriate given that these children were high scorers according to both informants and they had a symptom count above the Results threshold within each of the domains. For the comparison of participant teacher ADHD high Prevalence of the ADHD phenotype scorers and non-participant teacher ADHD high scorers, we also applied the impact supplement part of the SDQ In the Full Data group, the prevalence of the ADHD phe- (http://www.SDQinfo.org), which was included in the BCS notype was 5.2% (95% CI, 5.1–5.3%). For all subtypes, 123 766 Soc Psychiatry Psychiatr Epidemiol (2012) 47:763–769 Table 1 Prevalence of the ADHD phenotype according to subtype The Anonymous Data group had both higher mean and informant symptom scores and a higher frequency of high scorers compared to the Full Data group (Table 2). ADHD-Co ADHD-I ADHD-H Any ADHD (%) (%) (%) (%) When analyses reported in Table 2 were repeated for boys only and girls only, respectively, the pattern of dif- Parents 4.1 6.5 2.8 13.4 ferences between the two groups remained the same, Teachers 3.1* 5.4* 1.9* 10.4* except the difference in frequency of combined high Both 1.2 1.6 0.3 5.2 scorers between the participants and non-participants, * p \ 0.001, for difference in prevalence between parent and teacher which was no longer significant for girls (v = 4.37, reports p = 0.37). There were no significant differences in the proportion of girls relative to boys among the participant ADHD high parents reported significantly more children as ADHD high scorers compared to the non-participant high scorers 2 2 scorers than did teachers (Table 1). Note that the ADHD- (v = 1.7, p = 0.19) or any age differences (v = 0.91, Co phenotype comprises not only the 1.2% high scorers p = 0.32). The non-participant ADHD high scorers shown in Table 1 having the full combined symptom showed more inattention symptoms and less hyperactivity/ constellation from both informants, but also the 2.1% of inattention symptoms, both as sum scores and as symptom children rated as high scorers in different symptom counts, than the participants. There was no significant domains by parents and teachers (e.g., as inattentive by one difference between the two groups on the sum score of the informant and hyperactive by the other). impact measure (Table 3). Informant agreement Gender by informant effects Agreement between parents and teachers was r = 0.32 for Boys had significantly higher mean scores than girls on all ADHD combined high scorers, r = 0.22 for the ADHD subscales (p \ 0.001 for all comparisons) (Table 4). inattention only high scorers, and r = 0.13 for ADHD The gender (boy:girl) OR (95% CI) for ADHD high hyperactivity/impulsivity only high scorers. Agreement on scorers according to each informant and ADHD symptom ADHD high scorers for any subscale was r = 0.37. domain are summarized in Table 5. The ORs show increased risk for boys among high scorers for all ADHD Effects of attrition symptom domains and according to each informant. For ADHD combined high scorers and ADHD hyperactivity/ The frequency of teacher ADHD high scorers in the impulsivity high scorers, teacher reports yielded higher Anonymous Data group was 19.9% compared to 10.4% in gender (boy:girl) ORs than parent reports (CIs not the Full Data group alone. Combining the Anonymous overlapping). Data group and the Full Data group (N = 9 137), 13.1% were teacher high scorers. Assuming the ratio between teacher reported ADHD high scorers and the ADHD phe- Discussion notype (which is defined based on high scores from both informants) in the whole sample to be the same as in the We found twice as many children with the ADHD phe- Full Data group (5.2:10.4%), the estimate for the ADHD notype among children in the Anonymous Data group phenotype in the total population would be 6.6% (95% CI, versus the Full Data group, which demonstrated that 6.0–7.2%). Table 2 Mean teacher scores Full Data Anonymous Difference (95% CI) and frequency of the teacher ADHD high scorer (6/9) ADHD symptoms subtypes in the Full Data group Mean sum score 2.7 4.2 1.5 (1.25–1.75)* (N = 6,237) and the Mean inattention subscale 1.6 2.6 1.0 (0.85–1.15)* Anonymous Data group (N = 2,539) Mean hyperactivity subscale 1.1 1.6 0.5 (0.37–0.62)* High scorers v values (p value) Combined 3.1% 5.7% 32.5* Inattention only 5.4% 11.9% 111.4* Hyperactive/impulsive only 1.9% 2.3% 1.70 (p = 0.192) * p \ 0.001 for difference 123 Soc Psychiatry Psychiatr Epidemiol (2012) 47:763–769 767 Table 3 Mean sum scores and number of symptoms on teacher reports for ADHD high scorers in the Full Data group (N = 648) and the Anonymous Data group (N = 504) Impact sum Hyperactivity Inattention Hyperactivity Inattention score sum score sum score symptoms symptoms Full Data 1.45 6.12 8.46 4.63 6.53 Anonymous Data 1.54 (p = 0.450) 5.50 (p = 0.028) 8.98 (p = 0.020) 4.14 (p = 0.007) 6.90 (p = 0.002) For eight children in each group impact score was missing prevalence rate in the BCS was considerably lower than in Table 4 Mean ADHD symptom sum scores for each gender and a comparable British survey in a head-to-head comparison informant of the two samples with similar age groups and informants Inattentive Hyperactive/impulsive [12]. Parents Teachers Parents Teachers Our prevalence estimate of 5.2% for the ADHD pheno- type is comparable to that found in a recent German study Boys 2.75 2.34 2.03 1.70 [5] reporting a prevalence of 6.4% in the same age group. Girls 1.66* 0.88* 1.26* 0.45* Our prevalence estimate relied on two informants (which N = 6,237 (boys, 3,107; girls, 3,130) led to a decrease in prevalence), whereas the German study * p \ 0.001 for difference between boys and girls mean scores only included parent reports. On the other hand, the study included a 4-point response scale and the two most deviant responses were regarded as indicating the presence of ‘‘symptom’’. This is probably a more conservative symptom Table 5 Gender (boy:girl) ORs (95% CI) for the ADHD high scorer definition than ours, given that we had only three response subtypes for each informant categories and defined the two most deviant as indicating Combined Inattention Hyperactive/impulsive symptom. Observing the behavior in different settings Parent 2.9 (2.2–3.9) 2.1 (1.7–2.6) 1.7 (1.3–2.4) diminishes the likeliness of mixing it up with other behavioral disorders. The German study included no Teacher 6.2 (4.2–9.3) 3.1 (2.4–3.9) 5.4 (3.3–8.8) adjustment for non-responders, meaning that their preva- N = 6,237 (boys, 3,107; girls, 3,130) lence rate was probably also an underestimate. Given these important methodological differences, it is somewhat sur- prising that the prevalence estimates are in the same range. This is not to be taken as support for a more solid evidence attrition in studies with a typical attrition rate underesti- mated the ADHD phenotype prevalence. We estimated the basis—that at the end of the day the reported prevalence rates were very similar. This may rather reflect that the prevalence of the ADHD phenotype to 5.2% (parent and teacher reports) among children whose parents consented choices made in a study may be influenced by previously to participate in the study, but 6.6% in the total population. reported results. This also demonstrates the liability of Both parents and teachers reported more ADHD symptoms prevalence estimates to definition and the importance of in boys than in girls, but the gender difference was greater thorough characterization of the methodology applied when referring to any reported prevalence of ADHD. according to teacher reports. The excess proportion of boys with hyperactivity/impulsivity and the combined symptom The access to anonymous teacher questionnaires for most of the non-participants was a special asset of our constellation high score was higher according to teacher reports than parent reports. Informant agreement was low study. Comparing participants to non-participants, a much higher level of ADHD symptoms was found in the latter to fair. The estimated ADHD phenotype prevalence of 5.2% in group and this finding is relevant for all population-based epidemiological studies independent of their definition of our study was considerably higher than the DAWBA-based ADHD prevalence of 1.3% from a second study phase in ADHD. Similar trends have been reported for autistic the same population based on the Development and Well- symptoms in the same cohort [15]. Teacher reports show- Being Assessment (DAWBA) [13]. This is not unexpected ing a prevalence of 19.9% ADHD high scorers in the given that a DAWBA diagnosis requires the impairment Anonymous Data group compared to 10.4% in the Full Data group (an OR of 2.1) clearly illustrate the very criteria to be fulfilled and is therefore more comparable to a clinical diagnosis. Interestingly, our prevalence estimate important effect of non-participation in population studies of ADHD symptomatology. The non-participant ADHD for the ADHD phenotype was in the range of that reported from similar studies, while the DAWBA ADHD high scorers did not significantly differ from the participant 123 768 Soc Psychiatry Psychiatr Epidemiol (2012) 47:763–769 ADHD high scorers in boy:girl ratio, age or on the impact overidentification by parents. Boys with ADHD are measure. Thus, we did not get any support for the reported to engage in more rule breaking and externalizing hypothesis that teacher-rated non-participant children with behavior than girls with ADHD [2], and this has been ADHD symptoms would be more impaired than partici- found to affect teacher ratings of ADHD [1]. Some authors pants. Teachers completed the questionnaires without any have found support for the hypothesis that the difference in knowledge of who would later belong to the non-partici- symptom ratings across informants could be due to real pant group. However, one could suspect that they might situational differences [7]. Although the cause of the dif- have had a pre-conceived idea of who was going to par- ference in parent and teacher reports on ADHD symptoms ticipate or not. Interestingly, the non-participant ADHD in girls remains unresolved, it is important to bear this in high scorers had higher inattention scores and less hyper- mind and to explore the issue further in future studies. activity/impulsivity than the participant ADHD high scor- The BCS is unique in that teacher questionnaires cover ers. The explanation for this finding is speculative as we 97% of the total population. The current study focused on lack comparable reports from other studies. The finding symptoms of ADHD as reported on questionnaires. The underscores the importance of trying to assess non- validity of such reports may be questioned, since infor- responder bias in epidemiology in general and in psychi- mants may misunderstand items, and may also have rea- atric research specifically. Though generally assumed that sons for over- and underreporting problems in the child. the non-participants are at higher risk for mental disorders Also, the DSM-IV diagnostic criteria require an early onset and less privileged socially, few studies have explored the of the disorder (before age 7) and pervasive impairment non-participation in sufficient detail to characterize the from the symptoms. Thus, the phenotype and subtypes possible heterogeneity of non-participation. Investigating referred to here only indicate symptom constellations as selective participation in the British Child and Adolescent specified in the diagnostic criteria and are not comparable Mental Health Surveys, Goodman and Gatward reported to a clinical diagnosis. However, the symptom count important heterogeneity in the effect of deprivation on approach may be more readily reproducible than clinical parental non-participation [9]. Thus, it is important to note diagnoses in epidemiological research. that the process of non-participation is probably compli- The use of only three response categories represented a cated with a heterogeneous set of reasons, which give rise problem in the current study. It is not clear whether the to diverse effects on the non-participating group. middle category should be regarded as having the symptom We reported an estimate of the influence of attrition on or not. Many DSM-IV ADHD rating scales have used the ADHD phenotype prevalence estimate by assuming 4-point scales, where the two highest scores have been that the hypothetical parent reports of the children in the interpreted as indicating a symptom [4, 8, 21]. However, our prevalence of the ADHD phenotype according to tea- Anonymous group would have related to teacher reports at the same high scorer ratio as in the Full Data group. More cher reports is comparable to figures reported in previous sophisticated methods taking account of the differential studies of teacher-reported DSM-IV ADHD [4, 8, 15, 21]. parent–teacher agreement across number of symptoms for Similarly, the frequency of parent-reported ADHD symp- the high scorers or bootstrap methods might have been tom subtypes was comparable to that found by other used to estimate the effect of attrition on the total popu- studies using parent information [10]. Our use of strict lation prevalence. However, as discussed above, there are cross informant criteria compensated for a somewhat less several different uncertainties and limitations attached to restrictive individual symptom definition in estimating the the prevalence estimate (such as the differential use of prevalence based on both informants. A more conservative impact, etc.) that in the end we opted for illustrating the definition of symptom presence would have been to count non-response effect by this simple method as the inter- only ‘‘Certainly true’’ answers as symptom present. We pretation of this estimate is straightforward. We underline considered that the somewhat more inclusive symptom the importance of evaluating each aspect of the various criteria were suitable for the epidemiological consider- methodological influences rather than taking any one ations in this general child population study. prevalence estimate as reflective of the ‘‘true’’ rate. Our reported boy:girl ratios for ADHD high scorers on DSM-IV symptoms are in the range of earlier studies in Conclusion community samples [4, 8, 10, 11, 14, 20, 21]. Parents identified more girls than teachers, a finding that has been The prevalence of the ADHD phenotype based on teacher reported for the hyperactive/impulsive and for the com- and parent-reported symptomatology was clearly influ- bined subtypes in a previous study [8], but it is not clear enced by non-participation. The non-participation not only whether the higher number of girls identified by parents led to an underestimation of the prevalence, but also represent an underidentification by teachers or an affected the rates of inattention and hyperactivity/ 123 Soc Psychiatry Psychiatr Epidemiol (2012) 47:763–769 769 8. Gomez R, Harvey J, Quick C, Scharer I, Harris G (1999) DSM- impulsivity. The definition of the phenotype, gender and IV AD/HD: confirmatory factor models, prevalence, and gender choice of informant also influenced the detailed epidemio- and age differences based on parent and teacher ratings of Aus- logy of the ADHD phenotype in the present study. tralian primary school children. J Child Psychol Psychiatry 40:265–274 Acknowledgments We are grateful for the participation of children, 9. Goodman A, Gatward R (2008) Who are we missing? Area parents and teachers in the study. We thank Stein Atle Lie and Tore deprivation and survey participation. Eur J Epidemiol Wentzel-Larsen for statistical advice and to Jim Stevenson and Astri 23:379–387 J. Lundervold for important comments on the manuscript. This study 10. Graetz B, Sawyer M, Baghurst P (2005) Gender differences was supported by the Norwegian Research Council, the Norwegian among children with DSM-IV ADHD in Australia. J Am Acad Directorate of Health, the Western Regional Health Authority and the Child Adolesc Psychiatry 44:159–168 Centre for Child and Adolescent Mental Health, Uni Health, Bergen. 11. Graetz B, Sawyer M, Hazell P, Arney F, Baghurst P (2001) Christopher Gillberg was funded by the Swedish Medical Research Validity of DSM-IV ADHD subtypes in a nationally represen- Council. The work of Carsten Obel was funded by the Nordic Council tative sample of Australian children and adolescents. J Am Acad of Ministers research program ‘Longitudinal Epidemiology’ Child Adolesc Psychiatry 40:1410–1417 (020056). 12. Heiervang E, Goodman A, Goodman R (2008) The Nordic advantage in child mental health: separating health differences Open Access This article is distributed under the terms of the from reporting style in a cross-cultural comparison of psycho- Creative Commons Attribution Noncommercial License which per- pathology. J Child Psychol Psychiatry 49:678–685 mits any noncommercial use, distribution, and reproduction in any 13. Heiervang E, Stormark KM, Lundervold AJ, Heimann M, medium, provided the original author(s) and source are credited. Goodman R, Posserud MB, Ullebo AK, Plessen KJ, Bjelland I, Lie SA, Gillberg C (2007) Psychiatric disorders in Norwegian 8- to 10-year-olds: an epidemiological survey of prevalence, risk factors, and service use. J Am Acad Child Adolesc Psychiatry 46:438–447 References 14. Nolan E, Gadow K, Sprafkin J (2001) Teacher reports of DSM- IV ADHD, ODD, and CD symptoms in schoolchildren. J Am 1. Abikoff HB, Courtney M, Pelham W, Koplewicz H (1993) Acad Child Adolesc Psychiatry 40:241–249 Teachers’ ratings of disruptive behaviors: the influence of halo 15. Posserud M, Lundervold A, Gillberg C (2006) Autistic features in effects. J Abnorm Child Psychol 21:519–533 a total population of 7–9-year-old children assessed by the ASSQ 2. Abikoff HB, Jensen PS, Arnold L, Hoza B, Hechtman L, Pollack (Autism Spectrum Screening Questionnaire). J Child Psychol S, Martin D, Alvir J, March JS, Hinshaw S, Vitiello B, Newcorn Psychiatry 47:167–175 J, Greiner A, Cantwell DP, Conners CK, Elliott G, Greenhill LL, 16. Rutter M, Cox A, Tupling C, Berger M, Yule W (1975) Attain- Kraemer H, Pelham WE, Severe JB, Swanson JM, Wells K, ment and adjustment in two geographical areas. I. The prevalence Wigal T (2002) Observed classroom behavior of children with of psychiatric disorder. Br J Psychiatry 126:493–509 ADHD: relationship to gender and comorbidity. J Abnorm Child 17. SPSS I (1993) SPSS Advanced Statistics. SPSS, Chicago Psychol 30:349–359 18. Stormark KM, Heiervang E, Heimann M, Lundervold A, Gillberg 3. American Psychiatric Association (1994) Diagnostic and statis- C (2008) Predicting nonresponse bias from teacher ratings of tical manual of mental disorders: DSM-IV. American Psychiatric mental health problems in primary school children. J Abnorm Association, Washington Child Psychol 36:411–419 4. Carlson C, Tamm L, Gaub M (1997) Gender differences in 19. Swanson JM, Kraemer HC, Hinshaw SP, Arnold LE, Conners children with ADHD, ODD, and co-occurring ADHD/ODD CK, Abikoff HB, Clevenger W, Davies M, Elliott GR, Greenhill identified in a school population. J Am Acad Child Adolesc LL, Hechtman L, Hoza B, Jensen PS, March JS, Newcorn JH, Psychiatry 36:1706–1714 Owens EB, Pelham WE, Schiller E, Severe JB, Simpson S, Vi- 5. Dopfner M, Breuer D, Wille N, Erhart M, Ravens-Sieberer U tiello B, Wells K, Wigal T, Wu M (2001) Clinical relevance of (2008) How often do children meet ICD-10/DSM-IV criteria of the primary findings of the MTA: success rates based on severity attention deficit-/hyperactivity disorder and hyperkinetic disor- of ADHD and ODD symptoms at the end of treatment. J Am der? Parent-based prevalence rates in a national sample—results Acad Child Adolesc Psychiatry 40:168–179 of the BELLA study. Eur Child Adolesc Psychiatry 17(Suppl 20. Szatmari P, Offord D, Boyle M (1989) Ontario Child Health 1):59–70 Study: prevalence of attention deficit disorder with hyperactivity. 6. Faraone SV, Sergeant J, Gillberg C, Biederman J (2003) The J Child Psychol Psychiatry 30:219–230 worldwide prevalence of ADHD: is it an American condition? 21. Wolraich M, Hannah J, Pinnock T, Baumgaertel A, Brown J World Psychiatry 2:104–113 (1996) Comparison of diagnostic criteria for attention-deficit 7. Gomez R (2007) Australian parent and teacher ratings of the hyperactivity disorder in a county-wide sample. J Am Acad Child DSM-IV ADHD symptoms: differential symptom functioning Adolesc Psychiatry 35:319–324 and parent–teacher agreement and differences. J Atten Disord 11:17–27

Journal

Social Psychiatry and Psychiatric Epidemiology – Pubmed Central

Published: Apr 16, 2011

There are no references for this article.