Autism

http://aut.sagepub.com/ The Prevalence of Anxiety and Mood Problems among Children with Autism and Asperger Syndrome
Joseph A. Kim, Peter Szatmari, Susan E. Bryson, David L. Streiner and Freda J. Wilson Autism 2000 4: 117 DOI: 10.1177/1362361300004002002 The online version of this article can be found at: http://aut.sagepub.com/content/4/2/117

Published by:
http://www.sagepublications.com

On behalf of:

The National Autistic Society

Additional services and information for Autism can be found at: Email Alerts: http://aut.sagepub.com/cgi/alerts Subscriptions: http://aut.sagepub.com/subscriptions Reprints: http://www.sagepub.com/journalsReprints.nav Permissions: http://www.sagepub.com/journalsPermissions.nav Citations: http://aut.sagepub.com/content/4/2/117.refs.html

Downloaded from aut.sagepub.com by Diana Ababei on September 10, 2010

The prevalence of anxiety and mood problems among children with autism and Asperger syndrome
JOSEPH A. KIM
McMaster University, Canada

autism 2000 SAGE Publications and The National Autistic Society, Vol 4(2) 117132;012608 1362-3613(200006)4:2

P E T E R S Z AT M A R I McMaster University, Canada S U S A N E . B RYS O N York University, Canada DAV I D L . S T R E I N E R University of Toronto, Canada F R E DA J. W I L S O N McMaster University, Canada
a b st r ac t The objective of this study was to report on the prevalence and correlates of anxiety and mood problems among 9- to 14year-old children with Asperger syndrome (AS) and high-functioning autism. Children who received a diagnosis of autism (n 40) or AS (n 19) on a diagnostic interview when they were 4 to 6 years of age were administered a battery of cognitive and behavioural measures. Families were contacted roughly 6 years later (at mean age of 12 years) and assessed for evidence of psychiatric problems including mood and anxiety disorders. Compared with a sample of 1751 community children,AS and autistic children demonstrated a greater rate of anxiety and depression problems.These problems had a signicant impact on their overall adaptation. There were, however, no differences in the number of anxiety and mood problems between the AS and autistic children within this high-functioning cohort. The number of psychiatric problems was not correlated with early autistic symptoms but was predicted to a small extent by early verbal/non-verbal IQ discrepancy scores. These data indicate that high-functioning PDD children are at greater risk for mood and anxiety problems than the general population but the correlates and risk factors for these comorbid problems remain unclear. a d d r e s s Correspondence should be addressed to: pe t e r s zat m a r i , MD, Department of Psychiatry and Behavioural Neurosciences, McMaster University, Hamilton Health Sciences Corporation, Chedoke Campus, PO Box 2000, Station A, Hamilton, Ontario L8N 3Z5, Canada k ey wo r d s

adaptation; anxiety; Asperger syndrome; autism; mood

Introduction
Children with pervasive developmental disorders (PDDs) demonstrate qualitative impairments in reciprocal social interaction and in verbal and 117

Downloaded from aut.sagepub.com by Diana Ababei on September 10, 2010

au t i s m 4(2) non-verbal communication, and engage in a pattern of repetitive stereotypic activities (American Psychiatric Association, 1994). Although autism is the best known example of a PDD, a number of other subtypes have been suggested as being clinically meaningful including atypical autism (or PDD-NOS), Asperger syndrome (AS) and disintegrative disorder (Gillberg and Steffenburg, 1987; Wing, 1981). The diagnostic boundaries between these subtypes are unclear and there is considerable debate as to whether or not these are valid categories of disorder (Szatmari, 1992). One of the most contentious subtypes is AS. This disorder is dened by serious difculties in reciprocal social interaction, uent but pragmatically impaired speech and the presence of bizarre preoccupations and obsessions (Asperger, 1944). Since the reintroduction of AS into the clinical literature (Wing, 1981) and its subsequent addition to DSM-IV and ICD-10 (World Health Organization, 1992), the number of children receiving this diagnosis is steadily increasing. An important issue with respect to the clinical utility of a diagnosis like AS is the extent to which children with AS have an outcome different from those with high-functioning autism. In a 2 year follow-up study, we found that AS children had better social and communication skills, better language abilities, and fewer autistic symptoms compared with higherfunctioning children with autism (Szatmari et al., 1997). We have suggested that these changes are associated with the earlier appearance of useful language in the AS group. In other words, although there are differences at any single point in time between children with AS and highfunctioning autism, children with autism who develop uent language eventually come to resemble the AS children but at a younger chronological age. Another important outcome to investigate is the emergence of comorbid problems in anxiety and mood among PDD children (Ghaziuddin et al., 1998; Kobayashi and Murata, 1998; Realmuto and Ruble, 1999). It is possible that these psychiatric problems distinguish the outcome of AS and autism. We conducted a literature search to identify studies that compared the prevalence and characteristics of anxiety and mood problems among AS and autistic children. There was only one study that involved a direct comparison (Szatmari et al., 1989) so the search was broadened to include a case series of any PDD subtype.We only wished to review studies that systematically assessed a sample of PDD individuals using diagnostic categories or symptom counts of anxiety and mood problems. Individual case reports were not considered informative.We were able to identify only four studies of PDD children using a categorical diagnosis of depression (Ghaziuddin et al., 1998; Rumsey et al., 1985; Szatmari et al., 1989; Wing, 1981), although we considered studies by Tantam (1988) and Wolff and 118

Downloaded from aut.sagepub.com by Diana Ababei on September 10, 2010

k i m e t a l . : a n x i e t y a n d m o o d p ro b l e m s Chick (1980) that sampled non-PDD but conceptually similar individuals as well (i.e. those with schizoid disorder). There is general agreement that depression can be described in people with PDD but the presentation is often atypical (Lainhart and Folstein, 1994). Although most of these studies reported a higher than expected rate of mood disorder (up to 13 percent), several methodological constraints limit the interpretation of these results. For example, no study had a control group of non-PDD children, no study employed systematic sampling procedures, and only two studies (Rumsey et al., 1985; Szatmari et al., 1989) used structured psychiatric interviews. Thus it is difcult to know whether the reported rates are greater than expected compared with the general population and whether the estimates are biased by the use of convenience samples. There are only three studies of anxiety disorders, again uncontrolled (Rumsey et al., 1985; Szatmari et al., 1989; Tantam, 1988), and these suffer from the same methodological limitations noted above. Several reports (Rescorla, 1986;Vandergaag et al., 1995;Volkmar et al., 1988) have measured symptoms of anxiety and mood among PDD children and also found higher than expected scores compared with other groups. A signicant difculty with this literature is that no study attempted to differentiate anxiety/mood problems from PDD symptoms. Admittedly, anxiety/mood symptoms are hard to measure in PDD adolescents, particularly those with very limited verbal skills. It can also be difcult to tell whether a particular symptom (such as repetitive questioning) is part of an anxiety disorder or the PDD itself. Thus, it is important to carefully assess changes in behaviour (i.e. disturbances in sleep, changes in appetite and energy level) and obtain as much qualitative information as possible. However, this is often very difcult using questionnaires as measurement tools. Given these inconsistencies and methodological concerns, it is not surprising that there are virtually no data on the correlates and risk factors of anxiety and mood symptoms in the PDD population. One obvious possibility is that anxiety and mood problems are more common among verbal or higher-functioning PDD children simply because they are better able to articulate their feelings. Indeed, some investigators (Ghazziuddin et al., 1995) have cautioned that diagnosing affective disorder in this population may be difcult owing to a limited ability to communicate emotions both verbally and through facial expression. On the other hand, Lainhart and Folstein (1994) report that individuals with mood disorder tended to be female and lower functioning. A simple way to test this hypothesis is to examine the correlation between language ability and the number of anxiety and mood problems among PDD children, an analysis which has not yet been reported. Another possibility is that mood and anxiety problems in this popu119

Downloaded from aut.sagepub.com by Diana Ababei on September 10, 2010

au t i s m 4(2) lation are normal manifestations of environmental stressors. Ghaziuddin et al. (1995) reported that children with PDD suffering from depression experienced signicantly more life events in the preceding 12 months (i.e. change in group homes, change of education programmes, family sickness, bereavement etc.) than those who did not experience depression. The authors suggest that children with PDD react no differently than the general population when faced with negatively impacting life events and so share many similarities with non-autistic depressed patients. Consistent with this hypothesis, Ghaziuddin and Greden (1998) also reported that PDD children who suffer from depression are more likely to have a family history of depression than non-depressed children with PDD. There are no data on the extent to which early symptoms of autism might be correlated with later anxiety and mood problems. It is certainly possible that rituals and resistance to change reect a temperamental disposition towards anxiety that might manifest itself in more traditional form later on. In that case, there should be high correlations observed between early measures of repetitive stereotyped behaviours and later, apparently comorbid, anxiety disorders. There are also no data on the impact of these symptoms/disorders on the general outcome of PDD children. For example, if anxiety and mood disorders are prevalent in this population, problems associated with these internalizing symptoms (such as aggression and disruptive behaviour) may affect family life and social relationships. This could lead to other problems in adaptation, as the ability to be socially integrated into the community is largely dependent on learning how to cope in that environment. The purpose of the following study was to systematically investigate the prevalence of comorbid psychiatric problems, to identify potential correlates and risk factors of these problems, and to assess the impact of these additional handicaps on overall outcome. Our goal was to address four questions: (1) do high-functioning PDD children have more problems with anxiety and depression than the general population; (2) what is the impact of these problems on overall adaptation; (3) do children with AS have more of these problems than children with autism; and (4) what are the risk factors for anxiety and mood problems in terms of early cognitive skills and autistic symptoms? Questionnaire data were used although it was recognized that such data may not be as reliable or valid as using clinical interviews. However, the questionnaires were standardized on a community sample, allowing us to estimate the extent to which the rate of anxiety and mood symptoms in this population was greater than expected. 120

Downloaded from aut.sagepub.com by Diana Ababei on September 10, 2010

k i m e t a l . : a n x i e t y a n d m o o d p ro b l e m s

Method
Subjects All children 46 years of age, either coming for assessment, or currently in treatment, at a PDD service of six different centres which serve preschool children with developmental disabilities in southern Ontario, were identied. If the child received a clinical diagnosis of PDD but was untestable, or received a mental age score less than half their chronological age on psychometric testing, they were dropped from further consideration. The remaining children, including those without psychometric data, were asked to participate in the study and informed consent was obtained from the parents. Children who received a diagnosis of autism or Asperger syndrome using data from the Autism Diagnostic Interview (ADI: Le Couteur et al., 1989), and who had either a Leiter IQ score above 68 or a StanfordBinet IQ score above 70, were included in the nal cohort. During the enrolment phase, 164 children, 46 years of age, were screened at one of the six PDD centres. Of these, 80 children were dropped either because they were not PDD, or because their behaviour was too low functioning to get an IQ estimate, or because previous psychometric testing revealed they were functioning below the mental age criterion. The remaining 84 PDD children underwent the full psychometric battery. A further 16 were dropped because their IQ on both the Leiter Scales and the revised StanfordBinet was below the IQ cut-off listed above. Thus, the sample size at enrolment was 68 non-retarded preschool PDD children. To qualify for a diagnosis of autism or AS, all children had to have at least one example of an impairment in reciprocal social interaction (section A of the DSM-IV criteria), at least one example of an impairment in verbal or non-verbal communication (section B of DSM-IV) and at least one example of a repetitive, stereotyped behaviour (section C) as specied on the ADI. Unfortunately, a diagnostic algorithm from the ADI has not been developed for Asperger syndrome. As a result, data from the interview were used to classify children into this category based on our previous work and a review of the literature (Szatmari, 1992; Szatmari et al., 1989). For AS, there had to be (in addition to at least one PDD symptom in all three domains) an absence of clinically signicant language delay (i.e. spontaneous phrase speech must be achieved by 36 months of age), and an absence of persistent (greater than 3 months) deviant language development such as delayed echolalia, pronoun reversal and neologisms. Autistic children met ICD-10 criteria for autism and spoke after 36 months or had evidence of deviant language as dened above. Using these criteria, 47 children met criteria for autism and 21 for Asperger syndrome. This denition is consistent with, though not identical to, the DSM-IV
121

Downloaded from aut.sagepub.com by Diana Ababei on September 10, 2010

au t i s m 4(2) criteria for AS (American Psychiatric Association, 1994) which only appeared after data collection at enrolment was completed. DSM-IV stipulates that if a child meets criteria for both autism and Asperger disorder, he/she is preferentially given a diagnosis of autism. As previously reported (Miller and Ozonoff, 1997; Szatmari et al., 1995), however, strictly applying the DSM-IV criteria would have resulted in few children in our sample receiving a diagnosis of Asperger disorder. Almost all AS children also met the ADI algorithm diagnosis for autism (Szatmari et al., 1995). As a result, we stipulated that if a child met criteria for both, he/she was instead given a diagnosis of AS. In essence, then, the criteria for Asperger syndrome in this study included some children who have enough symptoms to meet the ADI criteria for autism but who also have an absence of delayed and deviant language development. To keep this distinction clear, children in our sample are said to have Asperger syndrome, not Asperger disorder. The mean age of the sample at inception was 66 11 months. There were only seven girls in the sample, consistent with the high malefemale ratio found in high-functioning autistic children. A full description of the characteristics of the sample at inception is available in a previous publication (Szatmari et al., 1995). For the present report, families were contacted roughly 6 years after their enrolment in the study when the children were between 9 and 14 years of age. The mean age was 12 years (SD 1.2). The attrition rate was kept low; data were available for 19 children in the AS group, and 40 in the autism group.

Procedure The assessments conducted at enrolment that were used in this analysis consisted of a full psychometric evaluation, including two IQ tests, and a semi-structured interview (the Autism Diagnostic Interview). During the follow-up 6 years later, we administered a questionnaire to parents to measure the prevalence of psychiatric problems (OCHSR). Instruments at enrolment Autism Diagnostic Interview (ADI) This semi-structured interview for parents (Le Couteur et al., 1989), administered to parents, was designed to make a diagnosis of autism according to both draft ICD-10 and DSM-III-R criteria. The ADI has been shown to have excellent reliability (intra-class correlation for multiple raters ranges between 0.94 and 0.97) and validity in a study discriminating autistic and mentally retarded children (Le Couteur et al., 1989). The interview items can be summed to produce scores in the domains of reciprocal social interaction, communication and repetitive stereotypic activities.
122

Downloaded from aut.sagepub.com by Diana Ababei on September 10, 2010

k i m e t a l . : a n x i e t y a n d m o o d p ro b l e m s Arthur Adaptation of the Leiter Performance Scales (Levine, 1986) This is a standard measure of non-verbal problem solving and learning ability. The Leiter is widely used with PDD and other language impaired children. It is especially appropriate to the population under study because it does not require verbal instructions for administration, and correlates highly with WISCR IQs (Levine, 1986). StanfordBinet Intelligence Scale, fourth edition (Thorndike et al., 1985) The StanfordBinet measures overall cognitive development as well as four different cognitive domains: verbal reasoning, quantitative reasoning, abstract/visual reasoning and short-term memory skills (mean 100, SD 15). The verbal reasoning score was used to measure language competence. If the child was mute, or was unable to reach a basal level on the verbal reasoning subtest, he or she was given a score equal to one below the lowest score obtained by a child in the sample who was able to complete the test. This was done to ensure that missing data were kept to a minimum and provided a reasonably valid, if conservative, estimate of abilities. If a child was uncooperative during testing on several occasions, the score was recorded as missing for that test.

Measure of psychiatric problems at 6 year follow-up At the 6 year follow-up, a measure of psychiatric problems (the OCHSR) was administered to the parents. The Revised OCHS is a parent questionnaire originally used in the 1983 Ontario Child Health Study (OCHS), and is itself a revision of the Child Behaviour Checklist (Achenbach and Edelbrock, 1983). Our group substantially modied the instrument by adding questions and devising a scoring method to measure DSM-III-R disorders of conduct disorder, oppositional disorder, ADHD, overanxious disorder, separation anxiety disorder and depression. The instrument was standardized on a community sample of 1751 children and adolescents and norms are available by age and gender (Boyle et al.,1993). A score two standard deviations above the mean is classied as clinically relevant and the proportion of PDD adolescents above the threshold on the overanxious and depression scales was calculated. Three percent of the general population obtained scores above this cut-off. Parents were asked to answer a variety of questions on their childs emotional and behavioural states within the past 6 months. For example, parents were asked to rate if their child was scared to go to sleep without parents being near. Each question on the checklist was answered never or not true, sometimes or somewhat true, or often or very true. Questions with more than one answer checked were assigned an intermediate score. Upon return of the questionnaires, we calculated T scores for the follow123

Downloaded from aut.sagepub.com by Diana Ababei on September 10, 2010

au t i s m 4(2) ing measures: overanxious scale (OA), separation anxiety scale (SA) and depression scale (DEP). T scores are standardized scores that measure the amount of deviation from a population mean of 50 and a standard deviation of 10. This is done to eliminate differences in scores between disorders that differ in the number of symptoms included. For example, if a child receives a T score of 70, that represents a score two standard deviations above the mean for that age and gender. Testretest reliability for the overanxious scale is 0.73, and for the depression scale is 0.64 (Boyle et al., 1993). The scales are meaningful in the sense that they explained 60 percent of the variance in predicting attending versus non-attending outpatient mental health clinics (Boyle et al., 1993). We also calculated two summary scores. The internalizing score (IS) was the sum of OA, SA and DEP. The externalizing score (ES), a measure of disruptive behaviour, was the sum of scores on three scales: conduct disorder scale (CD), oppositional scale (OPP) and ADHD scale (ADHD). Parents also answered a variety of questions on how their childs behaviour affected overall adaptation in the past 6 months. Some questions concerned practical matters. For example, parents were asked if their childs behaviour had made it difcult or prevented them from taking their child out in public, from going shopping or visiting, leaving the child with a babysitter, forgoing their holidays, being anxious about the future or preventing the parents from having friends over to the house. These were coded in a simple yesno format. We also asked questions about the childs friendships and social activities. For example, parents rated how well their child had managed with friends, teachers and family members during the past 6 months. These questions were scored on a ve-point ordinal scale ranging from very poorly to very good.

Analysis Our hypothesis was that higher-functioning PDD children show more symptoms of anxiety and mood disorders than the standardization sample. A one-sample t-test was conducted to see if mean scores in the PDD cohort were greater than the population mean of 50 (SD 10). A two-sample ttest would have been inappropriate given the unequal cell sizes and the larger number of children in the control group. Children whose parents reported problems in adaptation were compared with children who did not have these difculties on the anxiety/depression scales using a twosample t-test. Correlations were also calculated between current behaviour difculties (the summary externalizing behaviour score) and anxiety and mood problems as well as the impact of these problems on social relationships using the items from the OCHSR questionnaire. A two-sample t-test was also used to see whether AS children had more anxiety and mood
124

Downloaded from aut.sagepub.com by Diana Ababei on September 10, 2010

k i m e t a l . : a n x i e t y a n d m o o d p ro b l e m s problems than children with autism. Finally, risk factors for anxiety/mood problems were identied in two regression analyses using early scores in cognitive functioning and autistic symptoms from the psychometric assessments and the ADI taken at enrolment. All anxiety and mood scores are reported as T scores based on age and gender population means.

Results
The prevalence of mood and anxiety problems is presented in Table 1. Compared with a random sample of 1751 community children (Boyle et al., 1993), 16.9 percent of the PDD sample scored at least two standard deviations above the population mean on a measure of ADHD, 16.9 percent scored at this level on the measure of depression (DEP), and 13.6 percent scored at this level on a measure of generalized anxiety (OA). On the measure of separation anxiety (SA), 8.5 percent of the PDD sample scored at least two standard deviations above the normal population mean. On the cumulative internalizing score (IS, the sum of overanxious, depression and separation anxiety measures), 13.6 percent of the PDD sample were classied as clinically relevant. On a summary measure of disruptive behaviour (ES), a substantial portion of the PDD sample (8.5 percent) scored at
Table 1 Mean and prevalence of mood and anxiety problems Percentage of clinically relevant scoresa IS OA SA DEP ES CD ADHD OPP
a Percentage b p-value

Mean of PDD sample 58.43 57.09 53.82 60.12 57.46 51.23 60.00 53.48

SD

t-test

p-valueb

13.6 13.6 8.5 16.9 8.5 3.4 16.9 6.8

10.5 11.7 10.8 11.1 10.1 10.5 11.1 10.5

6.16 4.67 2.73 7.03 5.65 0.90 7.03 2.55

0.001 0.001 0.009 0.001 0.001 0.373 0.001 0.013

of scores two standard deviations above the normal population mean of 50. for one-sample t-test (two-tailed).

Parents IS OA SA DEP ES CD ADHD OPP

report of symptoms in PDD subjects internalizing score (OA, SA, DEP) overanxious separation anxiety depression externalizing score (CD, ADHD, OPP) conduct disorder ADHD oppositional

125

Downloaded from aut.sagepub.com by Diana Ababei on September 10, 2010

au t i s m 4(2) least two standard deviations above the population mean. Table 1 also shows the mean scores of the cohort. A one-sample t-test comparing the scores of the PDD cohort with the population mean of 50 showed that the PDD group had higher scores on all measures except CD. There were particularly large differences between the PDD group and the population mean on the measures of DEP and OA. There was also a high correlation between anxiety and mood problems and difculties with disruptive behaviour as measured by the composite externalizing behaviour score (ES) from the OCHSR. The OA and DEP scales were highly correlated with each other (r 0.51, p 0.001) and both were correlated with the externalizing score (r 0.67 with DEP, p 0.001; and r 0.50 with OA, p 0.001). In other words, children with anxiety and mood problems had more aggressive and oppositional behaviour than children with low scores on these measures. Anxiety and mood problems also had signicant effects on both the parents and the childrens adaptation. Parents of PDD children who were worried about their childs chances in the future reported more anxiety and depression in the children than parents who were not so worried (Table 2). Similarly, parents who had to forgo holidays because of the childs behaviour and could not invite friends over to the house reported more depression in the children (but not anxiety problems) than parents without these concerns. Children with high scores on the DEP scale had poorer relationships with teachers (r 0.27; p 0.05), peers (r 0.28; p 0.05) and family members (r 0.43; p 0.05) compared with PDD children without these psychiatric problems.There were no signicant correlations with the anxiety scale. Thus mood problems, more than anxiety, seemed to have a signicant impact on the family and on social relationships.
Table 2 Mean anxiety and depression scores (SDs) Anxiety yes no yes no yes no yes no yes no (9) (50) (6) (52) (10) (49) (26) (33) (6) (53) 57.6 54.3 57.4 55.1 59.1 56.7 60.8 54.2 59.4 56.8 (11.4) (13.5) (11.9) (11.6) (11.5) (11.9) (10.2)a (12.6) (15.5) (11.3) Depression 65.5 59.1 66.4 59.4 68.2 58.5 64.8 56.4 69.9 59.0 (11.6) (10.0) (9.6) (11.3) (8.8)a (10.8) (10.6)a (10.1) (10.6)a (10.6)

Variable (N) Prevented from going in public Prevented from going out Forgoing holidays Anxious about the future No friends to the house
ap

0.05 by t-test (two-tailed).

126

Downloaded from aut.sagepub.com by Diana Ababei on September 10, 2010

k i m e t a l . : a n x i e t y a n d m o o d p ro b l e m s
Table 3 Comparison of psychiatric problems between AS and autistic children N DEP: OA:
a

Mean 59.4 61.7 56.5 58.4

SD 9.6 13.8 11.8 11.7

t (d.f.) 0.73 (57) 0.60 (57)

p-valuea 0.47 0.55

autism AS autism AS

40 19 40 19

p-value for t-test (two-tailed).

Measures of depression and generalized anxiety did not differ signicantly between children initially diagnosed with AS or autism (see Table 3). In fact, the mean scores between the groups were very similar.There was also no association between early autistic symptoms as measured by the ADI at enrolment and anxiety/mood problems 6 years later. None of the three independent variables (impairments in reciprocal social interaction, impairments in verbal and non-verbal communication, and repetitive stereotypic activities) was signicantly associated with later anxiety (F 0.57; d.f. 3, 53; p 0.64) or mood problems (F 0.73; d.f. 3, 53; p 0.54). In particular, the correlation between the repetitive activities (which includes rituals and resistance to change) and anxiety and depression scores was low and non-signicant (r 0.08 and 0.17 respectively). We also examined the extent to which early cognitive variables measured at the time of diagnosis could predict later emotional problems.Three cognitive variables taken at enrolment were entered into a multiple regression model: Leiter non-verbal IQ, the verbal reasoning score from the StanfordBinet and a measure of verbal/non-verbal discrepancy (verbal reasoning minus non-verbal IQ). A backwards selection procedure was used for both anxiety and mood problems as the dependent variable. We hypothesized that children with better language scores (verbal reasoning) would score higher on anxiety and mood, in large part due to a better ability to express themselves. In both regressions, however, the only variable to remain in the model was the discrepancy score (for OA, F 9.83; d.f. 1, 57; p 0.003; for DEP, F 4.06; d.f. 1, 57; p 0.049). In other words, children with higher verbal than non-verbal abilities had more anxiety and mood problems. The amount of variance explained was very small in both cases (13 percent for OA and 5 percent for DEP) even though the models were signicant.

Discussion
Six years following the inception of the study, we sent the OCHSR questionnaire to parents to measure the prevalence of psychiatric problems, 127

Downloaded from aut.sagepub.com by Diana Ababei on September 10, 2010

au t i s m 4(2) according to DSM-III-R criteria, in a cohort of high-functioning children with autism and AS. Our goal was to address four questions: (1) do highfunctioning PDD children have more problems with anxiety and depression than the general population; (2) what is the impact of these problems on overall adaptation; (3) do children with AS have more of these problems than children with autism; and (4) what are the risk factors for anxiety and mood problems in terms of early cognitive skills and autistic symptoms? A substantial proportion of the cohort of high-functioning PDD children described here scored at clinically relevant levels on several scales including depression and generalized anxiety.Thus it was evident that these problems were more common in this high-functioning PDD group than in the general community of children of the same age. Internalizing problems were more common than externalizing problems. Although many have assumed higher rates of psychiatric disorder, the present report is the rst demonstration that comorbid psychiatric problems in PDD children are more common than in a community sample.These problems were not trivial; almost a fth (17 percent) of the children scored at clinically relevant levels of depression, which was the most common problem in this group. Moreover, these psychiatric problems had an important impact on the parents and childrens lives. For example, those children with anxiety and mood problems were more aggressive, limited their parents social activities and had poorer relationships with teachers, peers and family members. These data support the need for close attention to the presence of anxiety and mood problems among high-functioning PDD children. It may be that treating these comorbid problems is an effective way of decreasing levels of aggression. We were unable to identify strong risk factors for these problems in anxiety and mood. It is well known that emotional and behavioural problems are more common in learning disabled (LD) children as a whole than in the general population. The emotional and behaviour problems in the PDD children reported here may simply reect that fact. A control group of LD adolescents would have been helpful in sorting this out, though one could still make the argument that even if the rates of anxiety and mood problems were similar in the LD and PDD children, the mechanisms may be quite different. It was interesting to note that there were no differences on the anxiety and mood measures between children with a diagnosis of autism or AS. This was somewhat surprising in terms of the available literature and in view of the better langauge abilities of the AS adolescents. Indeed, early language skills did not predict later anxiety or mood problems. In other words, the prevalence of comorbid problems was not simply a reection 128

Downloaded from aut.sagepub.com by Diana Ababei on September 10, 2010

k i m e t a l . : a n x i e t y a n d m o o d p ro b l e m s of the ability to express feelings of anxiety and sadness, though it is important to note that the reports were by parents who inferred these symptoms based on the childs behaviour and comments. We also hypothesized that children with more repetitive stereotypic activities would show more anxiety symptoms. We felt that there would be homology between these early autistic symptoms and later anxiety. However, no such correlation between these variables was observed. The only cognitive variable to be associated with later psychiatric problems was a score reecting the discrepancy between early verbal and non-verbal abilities; i.e. children with lower non-verbal and higher verbal skills at enrolment showed more anxiety and mood problems 6 years later. The mechanism for this association is unclear but it is reminiscent of the prole of non-verbal learning disability seen in some children with AS as reported by Klin et al. (1995). Perhaps this prole reects greater right hemisphere dysfunction and so places these children at risk of anxiety and mood disorders as suggested by Rourke and Ozols (1985). In general, the amount of variance explained was small, however. Other variables such as family history, life events, theory of mind and ones awareness of disability may also be relevant to understanding the potential risk factors for the development of mood and anxiety problems in this PDD population and should be explored in future. This study has several strengths and unique features: (1) it is the only outcome study of an inception cohort of children with autism and AS; (2) sample loss between enrolment and follow-up assessment 6 years later was minimal; and (3) it is the only study to systematically investigate the prevalence of comorbid psychiatric problems compared with the general population. The study also has several limitations. It would certainly have been better to use interview data from both the child and the parents to make an assessment of comorbidity. We may have underestimated the prevalence of these problems by not interviewing the children themselves; and with questionnaire data, we cannot be certain that the problems reported by parents are true symptoms of anxiety and depression rather than variable expressions of PDD symptoms. We would argue though that even with interview data, such a differentiation would be difcult as there are no clear and simple rules to determine whether repetitive questioning, for example, is sign of anxiety, a verbal ritual or a communication impairment. There is an urgent need for measurement studies of this sort. Another potential concern is the difference between our denition of AS and that of DSM-IV. Our argument is that it is very difcult to identify preschool children with AS by applying DSM-IV criteria. In particular, the hierarchy rule that a diagnosis of autism takes precedence over a diagnosis of AS is very problematic. Indeed Miller and Ozonoff (1997) have shown 129

Downloaded from aut.sagepub.com by Diana Ababei on September 10, 2010

au t i s m 4(2) that the children described by Asperger himself would not have met DSMIV criteria for AS.The discrepancy in criteria should not be a problem with the validity of the results but may have some impact on the generalizability of our ndings. For example, it may be easier to identify Asperger disorder children at a later age who do not also meet ADI criteria for autism. By using an inception cohort of preschool children, we may have selected a more severe group of Asperger syndrome children. The problem with not using an inception cohort, however, is that one can never be sure that systematic sample loss has not taken place (Sackett et al., 1997). The clinical implications of these results are clear: a full assessment of the high-functioning PDD adolescent should include a careful investigation of anxiety and mood problems. Behavioural interventions to reduce anxiety and improve mood may be indicated if these problems also affect overall adaptation over and above the PDD. In so far as these difculties are also amenable to pharmacological intervention, overall adaptation may also be improved but randomized controlled trials are needed to demonstrate the benets of pharmacotherapy in this population. Over-medication is already a serious problem among PDD children (Aman et al., 1995) and more work needs to be done at the basic level of psychometrics, as well as on our understanding of possible mechanisms, before the clinical importance of treating comorbid anxiety and mood problems on long-term outcome is established.

Acknowledgements This work was supported by awards from the Ontario Mental Health Foundation, the Vellum Foundation and the National Health Research and Development Program of Health Canada.The authors would like to express their gratitude to the families and clinicians who participated in the project. References
ac h e n bac h , t. m . & e d e l b ro c k , c . s. (1983) Manual for the Child Behavior Checklist and Revised Child Behavior Prole. University of Vermont Department of Psychiatry, Burlington, VT. a m a n, m . g. , va n b o u rg on d i e n, m . e . , wo l f o r d, p. l . & s a r ph a r e , g . (1995) Psychotropic and Anticonvulsant Drugs in Subjects with Autism: Prevalence and Patterns of Use, Journal of American Academy of Child and Adolescent Psychiatry 34 (12): 167281. a m e r i c a n p s yc h i at r i c a s s o c i at i on (1994) Diagnostic and Statistical Manual of Mental Disorders, 4th edn (DSM-IV). Washington, DC: APA. a s pe rg e r , h . (1944) Die autistischen Psychopathen, Kindesalter Archives fur Psychiatri und Nervenkrankenheiten 117: 76136. b oy l e , m . h . , o f f o r d, d. r . , r ac i n e , y. , f l e m i n g, j. e . , s zat m a r i , p. &

130

Downloaded from aut.sagepub.com by Diana Ababei on September 10, 2010

k i m e t a l . : a n x i e t y a n d m o o d p ro b l e m s
s a n f o r d, m . (1993) Evaluation of the Revised Ontario Child Health Study Scales, Journal of Child Psychology and Psychiatry 34 (2): 189213. g h a z i u d d i n, m . & g r e d e n, j. f. (1998) Depression in Children with Autism/Pervasive Developmental Disorders: A Case-Control Family History Study, Journal of Autism and Developmental Disorders 28 (2): 11115. g h a z i u d d i n, m . , a l e s s i , n. & g r e d e n, j. f . (1995) Life Events and Depression in Children with Pervasive Developmental Disorders, Journal of Autism and Developmental Disorders 25 (5): 495502. g h a z i u d d i n, m . , w e i d m e r - m i k h a i l , e . & g h a z i u d d i n, n. (1998) Comorbidity of Asperger Syndrome: a Preliminary Report, Journal of Intellectual Disability Research 42 (4): 27983. g i l l b e rg, c . & st e f f e n bu rg, s . (1987) Outcome and Prognostic Factors in Infantile Autism and Similar Conditions: A Population-Based Study of 46 Cases Followed Through Puberty, Journal of Autism and Developmental Disorders 17: 27387. k l i n, a . , vo l k m a r , f. r . , s pa r row, s. s. , c i c c h e t t i , d. v. & ro u r k e , b. p. (1995) Validity and Neuropsychological Characterization of Asperger Syndrome, Journal of Child Psychology and Psychiatry 36: 112740. ko baya s h i , r . & m u r ata , t. ( 1998) Behavioural Characteristics of 187 Young Adults with Autism, Psychiatry & Clinical Neuroscience 52 (4): 38390. la i n h a rt, j. e . & f o lst e i n, s. e . (1994) Affective Disorders in People with Autism: A Review of Published Cases, Journal of Autism and Developmental Disorders 24 (5): 587601. l e c o u t e u r , a . , ru t t e r , m . , lo r d, c . , r i o s, p. , ro b e rt s on, s. , h o l d g r at e r , m . & m c l e n na n, j. d. (1989) Autism Diagnostic Interview, Journal of Autism and Developmental Disorders 19: 36388. l ev i n e , m . n. (1986) Arthur Adaptation of the Leiter International Performance Scale: A Handbook. Montreal Institute of Psychological Research, Montreal. m i l l e r , j. n. & o zon o f f, s. (1997) Did Aspergers Cases Have Asperger Disorder? A Research Note, Journal of Child Psychology and Psychiatry 38 (2): 24751. r e a l m u to, g. m . , & ru b l e , l . a . (1999) Sexual Behaviors in Autism: Problems of Denition and Management, Journal of Autism and Developmental Disorders 29 (2): 1217. r e s c o r la , l . a . (1986) Preschool Psychiatric Disorders: Diagnostic Classication and Symptom Patterns Journal of the American Academy of Child and Adolescent Psychiatry 25 (2): 1629. ro u r k e , b. p. & o zo ls, e . j. (1985) Dimensions of Social Sensitivity in Two Types of Learning-Disabled Children, in B . P. RO U R K E (ed.) Neuropsychology of Learning Disabilities, pp. 281301. New York: Guilford. ru m s ey, j. m . , r a po po rt, j. l . & b e e ry, w. r . (1985) Autistic Children as Adults: Psychiatric, Social and Behavioral Outcomes, Journal of the American Academy of Child and Adolescent Psychiatry 24: 46573. s ac k e t t, d. l . , r i c h a r d s on, s. r . , ro s e n b e rg, w. & h ay n e s, r . b. (1997) Evidence-Based Medicine: How to Practice and Teach EBM. London: Churchill Livingstone. s zat m a r i , p . (1992) The Validity of Autistic Spectrum Disorders: A Literature Review, Journal of Autism and Developmental Disorders 22: 583600. s zat m a r i , p. , ba rto luc c i , g. & b r e m n e r , r . (1989) Aspergers Syndrome and Autism: Comparisons on Early History and Outcome, Developmental Medicine and Child Neurology 31: 70920.

131

Downloaded from aut.sagepub.com by Diana Ababei on September 10, 2010

au t i s m 4(2)
s zat m a r i , p. , a rc h e r , l . , f i s m a n, s. f. , st r e i n e r , d. l . & w i ls on, f. j . (1995) Aspergers Syndrome and Autism: Differences in Behavior, Cognition and Adaptive Functioning, Journal of American Academy of Child and Adolescent Psychiatry 34 (12): 166270. s zat m a r i , p. , b rys on, s. e . , st r e i n e r , d. l . , w i ls on, f. j. , a rc h e r , l . & ry e r s e , c . (1997) The Two-Year Outcome of Preschool Children with Autism and Asperger Syndrome, paper presented at the Annual Meeting of the American Child and Adolescent Academy, Toronto. ta n ta m , d. (1988) Lifelong Eccentricity and Social Isolation. Part 2: Aspergers Syndrome or Schizoid Personality Disorder?, British Journal of Psychiatry 153: 78391. t h o r n d i k e , r . l . , h ag n, e . p. & s u t t e r , j. m . (1985) StanfordBinet Intelligence Scale, 4th edn. Chicago, IL: Riverside. va n d e rga ag e t a l . (1995) A Controlled Multivariate Chart Review of Multiple Complex Developmental Disorder, Journal of American Academy of Child and Adolescent Psychiatry 34 (8): 1096106. vo l k m a r , f. r . , c o h e n, d. j. , h o s h i n o, y. , r e n d e , r . d. & pau l , r . (1988) Phenomenology and Classication of the Childhood Psychoses, Psychological Medicine 18 (1): 191201. w i n g, l . (1981) Aspergers Syndrome: A Clinical Account, Psychological Medicine 11 (1): 11529. wo l f f, s. & c h i c k , j. (1980) Schizoid Personality in Childhood: A Controlled Follow-Up Study, Psychological Medicine 10: 85100. wo r l d h e a lt h o rga n i zat i on (1992) The ICD-10 Classication of Mental and Behavioural Disorders: Clinical Description and Diagnostic Guidelines. Geneva: WHO.

132

Downloaded from aut.sagepub.com by Diana Ababei on September 10, 2010