Child Adolesc Psychiatric Clin N Am 14 (2005) 631 648

The Developmental Epidemiology of Anxiety Disorders: Phenomenology, Prevalence, and Comorbidity

Anxiety has been one of the most difficult areas of child psychopathology to study in representative population samples. The main reasons for this are clinical uncertainty about the boundaries of the various anxiety disorders and the rarity of several of the disorders in population-based samples. Although the taxonomic problems are far from resolved, there has been considerable progress in the past decade. First, longitudinal and laboratory-based studies have made it clear that different types of anxiety have different correlates, predictors, and courses across childhood and adolescence [1]. Second, although there are still many problems with assessment, the situation is improving. Direct assessment of young children is always difficult because they often lack the cognitive abilities needed to talk about worry, fear, and panic [2]. However, parents have been shown to be reliable reporters about their young childrens anxieties (Angold, submitted for publication, 2005). In addition, most current assessment instruments incorporate measures of functioning so that researchers can decide what level of impairment is required to make a diagnosis. When functional impairment is required, the

This work is based in part on Costello EJ, Egger HL, Angold A. The developmental epidemiology of anxiety disorders. In: Ollendick T, March J, editors. Phobic and anxiety disorders in children and adolescents. New York: Oxford University Press; 2004. p. 6191; copyright 2004, Oxford University Press; with permission. T Corresponding author. E-mail address: [email protected] (E.J. Costello). 1056-4993/05/$ see front matter D 2005 Elsevier Inc. All rights reserved. doi:10.1016/j.chc.2005.06.003 childpsych.theclinics.com

632

costello et al

prevalence of some anxiety disorders such as simple phobias falls dramatically [3] and rates become much more consistent across studies [4]. A third issue for assessment is the overlap of depression and anxiety. The two types of disorders predict one another developmentally [5,6] and often respond to the same treatments [7], which has led some clinicians to treat them as part of the same syndrome [8]. However, a closer look suggests that the overlap of anxiety and depression applies only to some anxiety disorders [5]. It would be premature to change the taxonomy at this stage; we need to know a lot more about the developmental pathways of the various types of anxiety and depression, before that point is reached.

Prevalence and comorbidity This article reviews the epidemiologic literature on anxiety disorder in general and, when they are specified, on separation anxiety disorder (SAD), generalized anxiety disorder (GAD), overanxious disorder (OAD), specific phobias, panic, agoraphobia, social phobia, post-traumatic stress disorder (PTSD), and obsessive-compulsive disorder (OCD). However, many epidemiologic studies have reported on anxiety in general, without distinguishing among the specific categories set out in, for example, the Diagnostic and Statistical Manual of Mental Disorders (DSM), Third Edition, Revised (III-R) or DSM-IV, and it is often unclear how many different diagnoses have been included in the research protocol. Prevalence of anxiety disorders in preschool children Most of the research on anxiety and fear in young children has been conducted from the perspective of temperament and normal development, not psychopathology. In these approaches, anxiety or fear in young children is seen either as a normative phase of development or, in a subset of children, a risk factor for anxiety disorders. Between the age of 7 and 12 months, most infants develop a fear of strangers and express distress when they are separated from their primary caregivers. These fears peak between 9 and 18 months of age and decrease for most children by age 2.5 [9]. Approximately 15% of young children display more intense and persistent fear, shyness, and social withdrawal in response to unfamiliar people, situations, or objects than other children do [1012]. Behaviorally inhibited young children display characteristic patterns of physiology (high heart rate, low heart rate variability, high baseline levels of morning cortisol, and elevated startle responses) [13] and are more likely to develop an anxiety disorder later in childhood or adolescence or to have first-degree relatives with anxiety disorders [12,1418]. Recent advances in the nosology and diagnosis of psychiatric symptoms and disorders in preschool children [19,20] have made it possible to begin to define the boundaries between normative

developmental epidemiology of anxiety disorders

633

anxiety, temperament variation, and clinically significant anxiety disorders in very young children. Until 5 years ago, there were only three studies that could approximate community-based estimates of the prevalence of DSM anxiety disorders in preschool-aged children. The 1982 study by Earls [21] was ahead of its time, using questionnaires followed by clinical judgment, and applying DSM-III criteria to all of the 3-year-old children on Marthas Vineyard (Massachusetts). Fifteen years later, Keenan and colleagues [22] studied another small sample of children in poverty, who were assessed with a structured clinical interview. Lavigne and colleagues [23] used a combination of the Child Behavior Checklist [8], observational assessments, and measures of adaptive behaviors to make clinical consensus diagnoses of the preschoolers in a pediatric primary care setting. Recently, the Preschool Age Psychiatric Assessment (PAPA) [19,24] was developed for use with parents of children ages 2 through 5 years old. Table 1 [2123,25] shows the prevalence of anxiety disorders from these four studies of preschoolers in nonpsychiatric settings, providing an approximation to expected general population rates. The PAPA study, for which information was available by gender, found no significant gender differences for anxiety disorders overall or for specific anxiety disorders. Four- and 5-year-old children were significantly more likely than 2- and 3-year-old children were to have any anxiety disorder (11.9% versus 7.7%, respectively) or PTSD (1.3% versus 0.0%, respectively). African-American children were less likely to meet criteria for any anxiety disorder (6.4% versus 14.0%, respectively) or social phobia (0.6% versus 4.3%, respectively) than were non-African-American children. Comorbidity with other psychiatric disorder was common, ranging from 53% of cases of generalized anxiety disorder to 100% of cases of specific phobia. The most common type of comorbidity with nonanxiety disorders was with depression. Prevalence of anxiety disorders in school-aged children and adolescents Table 2 [3,2650] summarizes information on prevalence from recent epidemiologic studies of older children and adolescents. It includes all published studies using DSM-III-R (the earliest published in 1992) or DSM-IV (1996 onward). Studies are listed in order of their period of reference (current, 3-, 6-, or 12-month and lifetime.) Any anxiety disorder Studies with a short assessment interval and a single data wave had the lowest prevalence; for example, the current prevalence of one or more anxiety disorders was 2.8% in the Oregon Adolescent Depression Project [26]. Three-month estimates ranged from 2.2% to 8.6%; 6-month estimates ranged from 5.5% to 17.7%; 12-month estimates ranged from 8.6% to 20.9%; and lifetime estimates ranged from 8.3% to 27.0%. Not surprisingly, using a lifetime criterion on the oldest samples generated the highest estimates.

634

Table 1 Prevalence of anxiety disorders in community studies of preschoolers Study [reference] Earls et al, 1982 [21] Diagnostic criteria Questionnaire and clinical interview DSM-III Modified K-SADS DSM-III-R Clinical consensus DSM-III-R DISC DSM-III-R PAPA DSM-IV Age (y) 3 N 100 Any anxiety disorder (%) NR SAD (%) 5.0 GAD (%) NR OAD (%) NR Specific phobia (%) 0.0 Social phobia (%) 2.0 Selective mutism (%) NR costello et al

Keenan et al, 1997 [22] Lavigne et al, 1996 [23] Briggs-Gowan et al, 2000 [25] Angold et al, submitted for publication, 2005

5 25 46 25

104 510 516a 307b

NR NR 6.1 9.5

11.5 0.5 3.6 2.4

NR NR NR 6.5

NR NR 0.5 0.0

4.6 0.6 3.7 2.3

2.3 0.7 NR 2.2

NR NR NR 0.6

Abbreviations: DISC, Diagnostic Interview Schedule for Children; K-SADS, Kiddie-Schedule for Affective Disorders and Schizophrenia; NR, not reported. a Total sample of 1060. b Data weighted back to screening population of 1073.

developmental epidemiology of anxiety disorders

635

Specific anxiety disorders Recent studies have provided new information about the prevalence of specific anxiety disorders. They show that DSM-III-R OAD and DSM-IV GAD are the most common anxiety diagnoses and that panic disorder and agoraphobia (separately or together) are the least common. In the first prevalence studies, the reported rates of specific phobias (simple phobias in DSM-III-R) were extremely high, but most diagnostic instruments for children have now resolved this problem by taking disability into account in making the diagnosis. Two studies using adult instruments (the Diagnostic Interview Schedule [DIS] and the Composite International Diagnostic Interview [CIDI]) continued to report high rates of specific and social phobias and agoraphobia. This suggests that attention needs to be paid to the use of adult measures when assessing phobias in children. In contrast, the Bremen study of adolescents [51], which also used the CIDI, reported rates of specific and social phobia well within the range found in other studies of children and adolescents.

Anxiety and disability One of the most hotly debated areas in the past few years has been the relationship between psychiatric diagnosis and the level of functioning. When the first versions of the DIS for Children (DISC) were introduced in the 1980s, they were found to generate extremely high prevalence rates for some disorders, among which were some anxiety disorders [3,52]; for example, according to data from the four-site Methods for the Epidemiology of Child and Adolescent (MECA) mental disorders study, 39.5% of the children had at least one anxiety diagnosis in the previous 12 months [3]. At the same time, health maintenance organization (HMO) insurance companies and governmental agencies were concerned about whether all these children really needed treatment [53,54]. One solution to both problems was to require that, to receive a diagnosis, a child should show a significant degree of functional impairment or disability (to use the World Health Organizations preferred term). In 1993, the Federal Register defined a new class of psychiatric disorders, called Serious Emotional Disturbance (SED), which required significantly impaired functioning or disability in addition to a diagnosis [55]. SED was to be used as the criterion for assessing the prevalence of child psychiatric disorder in each state for the purpose of allocating federal block grants, and disability criteria were added to psychiatric diagnoses. Disability can be measured at several different levels. Each symptom can require impaired functioning; disability can be evaluated at the level of the syndrome or diagnosis or in the presence of any diagnosis, irrespective of which one causes impaired functioning; the interviewer could rate the childs level of functioning without making a diagnosis, using a separate measure [55,56]; or, of course, more than one method can be used.

636 Table 2 Summary of Diagnostic and Statistical Manual of Mental Disorders (III-R and IV) and International Classification of Diseases (Tenth Revision) studies of anxiety disorder prevalence Age of child/ adolescent (y) N 1418 Agoraphobia Panic Specific with or Social Avoidant disorder OCD phobia without phobia PTSD disorder OAD (%) (%) (%) panic (%) (%) (%) (%) (%) 0.3 1.3 4.4 0.1 1.1 0.9 2.5 0.5 4.4 Any anxiety GAD disorder (%) (%) 2.8 8.3 27.0 8.6

Study [reference] Oregon Adolescent Depression [26,27] Virginia Twin Study of Adolescent Behavioral Development [28] Caring for Children in the Community [29] Great Smoky Mountains [30] Caring for Children in the Community [29]

Diagnostic criteria DSM-III-R, K-SADS

Period of SAD reference (%) 0.2 1.2

costello et al

DSM-III-R, CAPA

817

1709 Current Lifetime Lifetime by age 19 2824 3 mo

DSM-III-R, DSM-IV, CAPA DSM-III-R, DSM-IV, CAPA DSM-III-R, DSM-IV, CAPA

912

388 3 mo

3.6

0.2

0.1

0.3

0.4

0.8

2.6

0.0

1.5

1.4

5.0

912

2709 3 mo

2.1

0.1

0.1

0.1

0.2

0.3

0.5

0.0

0.6

1.4

2.9

1317

532 3 mo

2.6

1.8

0.3

0.5

0.6

1.7

4.0

0.1

3.6

3.9

5.9

Great Smoky Mountains [30] Quebec Child Mental Health Survey [31] Methods for the Epidemiology of Child and Adolescent Mental Disorders [3] Health Maintenance Organization [32] Random sample (The Netherlands) [33]

DSM-III-R, DSM-IV, CAPA DSM-III-R, DISC 2.25

1316

3895 3 mo

0.4

0.3

0.2

0.3

0.3

0.7

1.0

0.1

1.5

2.3

2.2

614

2400 6 mo

DSM-III-R, 917 DISC 2.3 (Dx + CGAS b71)

1285 6 mo

2.6 (child) 1.6 (parent) 3.9

4.9 (child) 11.5 (parent) 2.6 3.3

5.4

3.1 (child) 3.8 (parent) 5.7

9.1 (child) 14.7 (parent) 13.0

developmental epidemiology of anxiety disorders

DSM-III-R, DISC 2.3 DSM-III-R, DISC 2.3

1218 1318

278 6 mo 274 6 mo

3.2 1.8

1.1 0.4

1.0

3.6 12.7

2.2 2.6

5.1 9.2

1.8 4.0

7.1 3.1

4.6 1.3

17.7 23.5 9.7 with CGAS b71 5.3 without CGAS b61 5.5

Northern Plains (child only) [34]

DSM-III-R, DISC 2.1C

1416

109

6 mo

1.9

2.9

2.0

1.9

(continued on next page)

637

638

Table 2 (continued ) Age of child/ adolescent (y) N 15 1424 Agoraphobia Panic Specific with or Social Avoidant disorder OCD phobia without phobia PTSD disorder OAD (%) (%) (%) panic (%) (%) (%) (%) (%) 1.2 0.6 1.8 1.6 (without panic) 2.6 (without panic) 4.0 9.1 2.6 0.7 Any anxiety GAD disorder (%) (%) 0.5 12.8 9.3

costello et al

Study [reference] Christchurch Longitudinal [35] Early Developmental Stages of Psychopathology [36] National Comorbidity Survey [37,38] Dunedin Longitudinal [39,40]

Diagnostic criteria DSM-III-R, DISC 2.3 DSM-IV, CIDI

Period of SAD reference (%)

1000 6 mo 3021 12 mo

Lifetime

1.6

0.7

2.3

3.5

1.3

0.8

14.4

DSM-III-R, CIDI DSM-III-R, DIS, DISC

1517

479 12 mo Lifetime 993 12 mo (DIS)

3.0 3.1

4.0

11.8 12.2

12.4 13.1 11.1

0.3 0.6

20.9 24.7 12.4

18

Dunedin Longitudinal [41] Puerto Rico [42] Iowa Family [43]

DSM-III-R, DIS, DISC DSM-IV, DISC DSM-III-R, UM-CIDI

21

417 Any onsets during 1519 1217 17.5

960 12 mo (DIS, DISC) 1897 12 mo 303

0.6

7.1

8.4

3.8

9.7

1.9

20.3

3.1

0.7

2.6

1.7

2.8 5.0

0.8

2.4

9.5 8.6 developmental epidemiology of anxiety disorders

1.3 (attack)

Essau Bremen [4447] Minnesota ParentChild Project [48] New York State Longitudinal [49] Boston Longitudinal [50]

DSM-IV, CIDI K-SADS, DSM-III-R DSM-III-R, DISC DSM-III-R

1035 Lifetime 172 Lifetime

4.6

0.5 1.7

1.7

3.5

1.6 5.8

1.6

1.7

4.6

15.1

Any onsets by 18 21

551

15.1

384 Lifetime

6.0

Abbreviations: CAPA, Child and Adolescent Psychiatric Assessment; CGAS, Childrens Global Assessment Scale; CIDI, Composite International Diagnostic Interview; DIS, Diagnostic Interview Schedule; DISC, Diagnostic Interview Schedule for Children; Dx, diagnosis; K-SADS, Kiddie-Schedule for Affective Disorders and Schizophrenia; UM-CIDI, University of Michigan Composite International Diagnostic Interview.

639

640

costello et al

The effects on the prevalence of anxiety disorders of assessing disability in different ways can be seen in the four-site MECA study using the DISC version 2.3. The study used two kinds of disability assessment: one type attached to each symptom cluster such that the interviewer asked about disability if the child or parent endorsed half plus one symptoms (ie, one more than half the symptoms needed for the diagnosis) and one that required the interviewer to rate the child on a scale of 0 to 100 on a level of functioning using the Childrens Global Assessment Scale (CGAS) [56] after the interview was ended. Adding either diagnosis-specific impairment or mild impairment (70 or less) on the CGAS halved the prevalence rate; adding both reduced it by two-thirds. A requirement of both diagnosis-specific impairment and severe (50 or below) impairment on the CGAS reduced it by almost 90% [3]. Anxiety was of all diagnoses the area most severely affected by requiring impairment, and among the anxiety disorders, simple phobia was the most affected; the prevalence estimate fell from 21.6% (no impairment requirements) to 0.7% (diagnosis-specific plus CGAS 50). Requiring disability as a criterion for making the diagnosis brings the rates down to levels that certainly make provider institutions more comfortable. However, there is growing evidence that disability can be associated with anxiety symptoms that do not reach the threshold for a diagnosis [57] and that even controlling for comorbidity with other psychiatric disorders anxiety disorders are associated with a high degree of disability [58]. The true burden to children, families, and society associated with these conditions is still unclear and needs further longitudinal research.

Sex and age differences in the prevalence of anxiety disorders Girls are somewhat more likely than boys are to report an anxiety disorder of some sort. However, at the level of individual diagnoses, few of the gender differences are large. If we assume that the difference is likely to be clinically and statistically meaningful if twice as many girls as boys reported a diagnosis, then only the eight studies cited in Table 2 reported any meaningful gender differences. Three studies reported more specific phobias in girls; two studies reported more panic disorder; two studies reported more agoraphobia; and one study reported more separation anxiety disorder and OAD. Lewinsohn and colleagues [26], in one of the few studies to examine the effects of potentially confounding factors associated with both gender and anxiety, found that controlling for 15 such factors did not eliminate the excess of anxiety disorders in girls. It is difficult to draw conclusions about age trends from this review because, in many cases, the age of subjects was confounded with the time frame of the interview. Thus, the 3-month studies had both the lowest prevalence rates and the youngest subjects, whereas the 12-month studies tended to have the highest prevalence as well as the oldest subjects. It is worth noting that Lewinsohn and

developmental epidemiology of anxiety disorders

641

Fig. 1. Mean age of onset by age 16, and interquartile range, anxiety disorders. (From Phobic and anxiety disorders in children and adolescents: a clinicians guide to effective psychosocial and pharmacological interventions by Thomas Ollendick, edited by John S. March, copyright Oxford University Press, Inc.; with permission.)

colleagues [26], using retrospective data, identified the female preponderance in anxiety disorders as emerging by age 6 years. Fig. 1 summarizes the ages of onset of different anxiety disorders. It represents the range and median ages for the studies from Table 2 for which this information is available. It shows that GAD began earliest, with a median age of around 6 years, whereas panic disorders rarely began before midadolescence. It is important to note that the range of estimates is very wide for some anxiety disorders.

Comorbidity among anxiety disorders Comorbidity among anxiety disorders has historically been a problem, not only for nosology and epidemiology but also for diagnosis and treatment. This is an area in which the high level of comorbidity found in clinical samples is mirrored in community samples. A review of published studies yields inconclusive results because (1) not all diagnoses were included in every study, and the number of anxiety disorders included in the analyses of comorbidity varies from study to study; (2) there is a lack of consensus about whether to

642

costello et al

control for comorbidity with other anxiety disorders or with other diagnoses when examining the strength of a particular association; and (3) concurrent and sequential comorbidity are not always distinguished clearly. The two published studies that have explored the issue of comorbidity among anxiety disorders [27,28] used bivariate analyses (corrected for gender and age in the latter case), so it is hard to interpret the finding that the majority of comparisons yielded a significant odds ratio. The present authors attempted to conduct a meta-analysis of the available data sets along the lines of work on psychiatric comorbidity that we have published previously [59]. However, for many of the diagnostic comparisons, there were too few data sets for such analyses to be feasible. Therefore, we can only draw some very tentative conclusions based mainly on studies for which we had direct access to the data: the Great Smoky Mountains Study (GSMS) [30], the Caring for Children in the Community (CCC) study [29], the Virginia Twin Study of Adolescent Behavioral Development [28], the HMO study [32], and the National Comorbidity Survey [37,38] (see Table 2). Generalized anxiety disorder and overanxious disorder A question of nosologic interest is the extent to which the older overanxious disorder category overlaps with the DSM-IV generalized anxiety diagnosis. The intention was that children who would formerly have received a diagnosis of overanxious disorder of childhood would be subsumed into the new GAD category. The criteria for GAD were loosened for children, who could receive the diagnosis if they had only one of the six symptoms of criterion C (restlessness, fatigue, difficulty concentrating, irritability, muscle tension, and sleep disturbance). However, with one exception, these symptom classes are very different from those defined for overanxious disorder (worries about the past or future, concerns about ones competence, need for reassurance, somatic symptoms, selfconsciousness, and muscle tension). Although it is mentioned briefly in the description of criterion A (excessive anxiety or worry), the latter symptoms are not set out in the new formal diagnostic criteria. On the other hand, five of the six new criterion C symptoms are very similar to symptoms of major depressive episode; it is very difficult to write diagnostic questions that reliably capture the subtle differences between, for example, the fatigue associated with depression and that associated with GAD. Thus, any examination of the overlap between OAD and GAD should take into account the possibility of their overlap with depression. Only three data sets (GSMS, CCC, and HMO) permitted a comparison of GAD, OAD, and depression in the same children. Here we use GSMS data to examine concurrent comorbidity among OAD using DSM-III-R criteria, GAD, using DSM-IV criteria, and DSM-IV depression. Over the course of the study, 182 children (11.6% of the sample) had one or more of the three diagnoses by the age of 16 years. Of those who were comorbid (5.4% of the sample or 47% of those with any of the three diagnoses) more than half (52%) had all three

developmental epidemiology of anxiety disorders

643

disorders. Because GAD was supposed to subsume OAD, one might expect this combination to be quite common. In fact, only 12 children (weighted 16% of those with either GAD or OAD) had both disorders without depression over the course of the study. Of the children with OAD without GAD, 36 of 88 (weighted 42%) also had a depressive disorder, not far from the 135 of 296 (weighted 34%) children with GAD but not OAD. There is a great deal of similarity between many of the symptoms of depression and GAD in DSM-IV. Therefore, one might have expected more comorbidity between depression and GAD than between depression and OAD, but this did not occur. In summary, although there is evidence for considerable comorbidity among GAD, OAD, and depression, tracing the extent to which this degree of comorbidity is real rather than methodological will require detailed longitudinal investigation. Comorbidity among the phobias and separation and anxiety disorders Almost all the studies confirmed significant comorbidities among the phobias: specific, social, and agoraphobia. The concurrent association between panic disorder and separation anxiety was nonsignificant in three out of the four studies that measured it. Absence of comorbidity Evidence for the lack of comorbidity among disorders generally lumped together under the label anxiety is as interesting as evidence for comorbidity. Little connection was found between separation anxiety and the group of phobias or between separation anxiety and overanxious disorder. GAD and OAD were unrelated to simple or specific phobias. There was, however, a consistent pattern of significant association between OAD and social phobia. Interestingly, in light of the clinical data suggesting a developmental link, there was no evidence of a cross-sectional association between separation anxiety and panic disorder. However, it must be emphasized that the evidence is often patchy: some associations could only be examined in two or three studies. Also, most studies examined were cross-sectional and could not test for possible sequential or developmental relationships.

Comorbidity with other disorders A review of comorbidity with anxiety disorders published in 1999 [59] showed that, controlling for other comorbid conditions, the highest level of anxious comorbidity was with depression, with a median odds ratio of 8.2 (95% CI, 5.812.0). This means that across all available studies, depression was 8.2 times as likely in children with anxiety disorders as in children without anxiety disorders and that 95 of 100 times the increase in likelihood of depression in the presence of anxiety would lie between 5.8 and 12 times. The odds ratio for

644

costello et al

comorbidity with conduct disorder or oppositional defiant disorder (ODD) was 3.1 (95% CI, 2.24.6) and that with attention deficit-hyperactivity disorder (ADHD) was 3.0 (95% CI, 2.14.3). These confidence intervals all exclude 1, indicating a statistically and substantively significant degree of comorbidity. In the case of substance use or abuse, although the bivariate odds ratios were significant in some studies, the association disappeared once comorbidity between anxiety and other psychiatric disorders was controlled [60]. There are few published reports that permit a review of comorbidity between specific anxiety disorders and other psychiatric diagnoses. Comorbidity analyses of the Oregon Adolescent Depression Study data set [27] looking at lifetime diagnoses showed that depression was significantly associated with each of the anxiety disorders except OCD, controlling for other disorders. Other lifetime associations found were ADHD with simple phobia, ODD with OCD, bipolar disorder with separation anxiety (in males), and alcohol abuse or dependence with OAD. The importance of a more detailed approach is shown by Kaplow and colleagues [61] reanalysis of the data from GSMS. This found that different anxiety disorders had different relationships to the risk of beginning substance use. Children with separation anxiety symptoms were less likely than other children to begin drinking alcohol and did so later than others, whereas those with generalized anxiety symptoms were more likely than other children to begin drinking and did so earlier.

Homotypic and heterotypic continuity An important question for clinicians is whether children with anxiety disorders can be expected to have further episodes of the same disorder (homotypic continuity) or to develop other psychiatric conditions (heterotypic continuity). There are few studies that deal thoughtfully with issues of concurrent versus sequential comorbidity [62]. Some studies have suggested that childhood anxiety predicts adolescent depression [5], but there also is evidence that early depression predicts anxiety [6]. Study of GSMS subjects [61] has demonstrated that the relationships among OAD, SAD, and alcohol use changed across development. The confused temporal relationship between anxiety and depression also may need more fine-grained analysis before we understand it properly. There are few epidemiologic studies that provide information about continuity among the anxiety disorders. The clinical literature suggests that separation anxiety is a predictor of later panic disorder [31,6365], for which there is some support in the GSMS (Bittner, submitted for publication, 2005). Controlling for concurrent comorbidity among the anxiety disorders, the GSMS showed a high degree of homotypic continuity of separation anxiety and social phobia. DSM-IIIR overanxious disorder also showed significant continuity. There was relatively little heterotypic continuity, which suggests a level of predictive validity in the diagnostic categories for the anxiety disorders across childhood and adolescence.

developmental epidemiology of anxiety disorders

645

Summary This article argues that the quality (accuracy, reliability, validity) of measures used to measure anxiety disorders in the child and adolescent population have improved enormously in the past few years. As a result, prevalence estimates are less erratic, our understanding of comorbidity is increasing, and the role of impairment as a criterion for caseness is more carefully considered. Several of the instruments developed for epidemiologic research are now being used in clinical settings. The further integration of research methods can be expected in the next few years as, for example, laboratory methods for testing stress response become available for use in the field. The integration of laboratory, clinical, and epidemiologic ideas and methods can only benefit children with anxiety disorders and their families.

References
[1] Ollendick T, March JS, editors. Phobic and anxiety disorders in children and adolescents: a clinicians guide to effective psychosocial and pharmacological interventions. New York7 Oxford University Press; 2004. [2] Dadds MR, James RC, Barrett PM, et al. Diagnostic issues. In: Ollendick TH, March JS, editors. Phobic and anxiety disorders in children and adolescents: a clinicians guide to effective psychosocial and pharmacological interventions. New York7 Oxford University Press; 2004. p. 3 33. [3] Shaffer D, Fisher PW, Dulcan M, et al. The NIMH diagnostic interview schedule for children (DISC 2.3): description, acceptability, prevalences, and performance in the MECA study. J Am Acad Child Adolesc Psychiatry 1996;35:865 77. [4] Costello EJ, Egger HL, Angold A. The developmental epidemiology of anxiety disorders. In: Ollendick T, March J, editors. Phobic and anxiety disorders in children and adolescents: a clinicians guide to effective psychosocial and pharmacological interventions. New York7 Oxford University Press; 2004. p. 61 91. [5] Costello EJ, Mustillo S, Keeler G, et al. Prevalence of psychiatric disorders in childhood and adolescence. In: Lubotsky Levin B, Petrila J, Hennessey K, editors. Mental health services: a public health perspective. New York7 Oxford University Press; 2004. p. 111 28. [6] Silberg J, Rutter M, Eaves L. Genetic and environmental influences on the temporal association between earlier anxiety and later depression in girls. Biol Psychiatry 2001;49:1040 9. [7] Ferdinand R, Barrett J, Dadds MR. Anxiety and depression in childhood: prevention and intervention. In: Ollendick TH, March JS, editors. Phobic and anxiety disorders in children and adolescents: a clinicians guide to effective psychosocial and pharmacological interventions. New York7 Oxford University Press; 2004. p. 459 75. [8] Achenbach TM. Manual for the child behavior checklist 418 and 1991 profile. Burlington (VT)7 University of Vermont, Department of Psychiatry; 1991. [9] Warren SL, Sroufe LA. Developmental issues. In: Ollendick TH, March JS, editors. Phobic and anxiety disorders in children and adolescents: a clinicians guide to effective psychosocial and pharmacological interventions. New York7 Oxford University Press; 2004. p. 92 115. [10] Kagan J, Snidman N. Infant predictors of inhibited and uninhibited profiles. Psychol Sci 1991;2:40 4. [11] Biederman J, Rosenbaum JF, Hirshfeld DR, et al. Psychiatric correlates of behavioral inhibition in young children of parents with and without psychiatric disorders. Arch Gen Psychiatry 1990;47:21 6.

646

costello et al

[12] Hirshfeld DR, Rosenbaum JF, Biederman J, et al. Stable behavioral inhibition and its association with anxiety disorder. J Am Acad Child Adolesc Psychiatry 1992;31:103 11. [13] Kagan J, Reznick JS, Snidman N. The physiology and psychology of behavioral inhibition in young children. Child Dev 1987;58:1459 73. [14] Biederman J, Rosenbaum JF, Bolduc-Murphy EA, et al. A 3-year follow-up of children with and without behavioral inhibition. J Am Acad Child Adolesc Psychiatry 1993;32:814 21. [15] Kagan J, Snidman N. Early childhood predictors of adult anxiety disorders. Biol Psychiatry 1999;46:1536 41. [16] Rosenbaum JF, Biederman J, Hirshfeld DR, et al. Further evidence of an association between behavioral inhibition and anxiety disorders: results from a family study of children from a nonclinical sample. J Psychiatr Res 1991;25:49 65. [17] Rosenbaum JF, Biederman J, Hirshfeld DR, et al. Behavioral inhibition in children: a possible precursor to panic disorder or social phobia. J Clin Psychiatry 1991;52:5 9. [18] Rosenbaum JF, Biederman J, Bolduc EA, et al. Comorbidity of parental anxiety disorders as risk for childhood-onset anxiety in inhibited children. Am J Psychiatry 1992;149:475 81. [19] Egger HL, Angold A. The preschool age psychiatric assessment (PAPA): a structured parent interview for diagnosing psychiatric disorders in preschool children. In: DelCarmen-Wiggins R, Carter A, editors. Handbook of infant, toddler, and preschool mental assessment. New York7 Oxford University Press; 2004. p. 223 43. [20] Angold A, Egger HL. Psychiatric diagnosis in preschool children. In: DelCarmen-Wiggins R, Carter A, editors. Handbook of infant, toddler, and preschool mental health assessment. New York7 Oxford University Press; 2004. p. 123 39. [21] Earls F. Application of DSM-III in an epidemiological study of preschool children. Am J Psychiatry 1982;139:242 3. [22] Keenan K, Shaw DS, Walsh B, et al. DSM-III-R disorders in preschool children from lowincome families. J Am Acad Child Adolesc Psychiatry 1997;36:620 7. [23] Lavigne JV, Gibbons RD, Christoffel KK, et al. Prevalence rates and correlates of psychiatric disorders among preschool children. J Am Acad Child Adolesc Psychiatry 1996;35:204 14. [24] Egger HL, Ascher BH, Angold A. The preschool age psychiatric assessment: version 1.1. Durham (NC)7 Center for Developmental Epidemiology, Department of Psychiatry and Behavioral Sciences, Duke University Medical Center; 1999. [25] Briggs-Gowan MJ, Horwitz SM, Schwab-Stone ME, et al. Mental health in pediatric settings: distributions of disorders and factors related to service juse. J Am Acad Child Adolesc Psychiatry 2000;39:841 9. [26] Lewinsohn PM, Lewinsohn M, Gotlib IH, et al. Gender differences in anxiety disorders and anxiety symptoms in adolescents. J Abnorm Psychol 1998;107:109 17. [27] Lewinsohn P, Zinbarg J, Lewinsohn M, et al. Lifetime comorbidity among anxiety disorders and between anxiety disorders and other mental disorders in adolescents. J Anxiety Disord 1997; 11:377 94. [28] Simonoff E, Pickles A, Meyer JM, et al. The Virginia twin study of adolescent behavioral development: influences of age, sex and impairment on rates of disorder. Arch Gen Psychiatry 1997;54:801 8. [29] Angold A, Erkanli A, Farmer EMZ, et al. Psychiatric disorder, impairment, and service use in rural African American and white youth. Arch Gen Psychiatry 2002;59:893 901. [30] Costello EJ, Angold A, Burns BJ, et al. The Great Smoky Mountains Study Of Youth: goals, designs, methods, and the prevalence of DSM-III-R disorders. Arch Gen Psychiatry 1996; 53:1129 36. [31] Silove D, Manicavasagar V, Curtis J, et al. Is early separation anxiety a risk factor for adult panic disorder? A critical review. Compr Psychiatry 1996;37:167 79. [32] Costello EJ, Angold A, Keeler GP. Adolescent outcomes of childhood disorders: the consequences of severity and impairment. J Am Acad Child Adolesc Psychiatry 1999;38: 121 8. [33] Verhulst FC, van der Ende J, Ferdinand RF, et al. The prevalence of DSM-III-R diagnoses in a national sample of Dutch adolescents. Arch Gen Psychiatry 1997;54:329 36.

developmental epidemiology of anxiety disorders

647

[34] Beals J, Piasecki J, Nelson S, et al. Psychiatric disorder among American Indian adolescents: prevalence in northern plains youth. J Am Acad Child Adolesc Psychiatry 1997;36:1252 9. [35] Fergusson DM, Horwood LJ, Lynskey MT. Prevalence and comorbidity of DSM-III-R diagnoses in a birth cohort of 15 year olds. J Am Acad Child Adolesc Psychiatry 1993;32:1127 34. [36] Wittchen H-U, Nelson CB, Lachner G. Prevalence of mental disorders and psychosocial impairments in adolescents and young adults. Psychol Med 1998;28:109 26. [37] Kessler RC. The national comorbidity survey of the united states. Int Rev Psychiatry 1994; 6:365 76. [38] Brady KT, Killeen TK, Brewerton T, et al. Comorbidity of psychiatric disorders and posttraumatic stress disorder. J Clin Psychiatry 2000;61(Suppl 7):S22 32. [39] Feehan M, McGee R, Williams SM. Mental health disorders from age 15 to age 18 years. J Am Acad Child Adolesc Psychiatry 1993;32:1118 26. [40] Douglass HM, Moffitt TE, Dar R, et al. Obsessive-compulsive disorder in a birth cohort of 18-year-olds: prevalence and predictors. J Am Acad Child Adolesc Psychiatry 1995;34:1424 31. [41] Newman DL, Moffitt TE, Caspi A, et al. Comorbid mental disorders: implications for treatment and sample selection. J Abnorm Psychol 1998;107:305 11. [42] Canino G, Shrout P, Rubio-Stipec M, et al. The DSM-IV rates of child and adolescent disorders in Puerto Rico. Arch Gen Psychiatry 2004;61:85 93. [43] Rueter MA, Scaramella L, Wallace LE, et al. First onset of depressive or anxiety disorders predicted by the longitudinal course of internalizing symptoms and parent-adolescent disagreements. Arch Gen Psychiatry 1999;56:726 32. [44] Essau CA, Karpinski NA, Petermann F, et al. Frequency and comorbidity of psychological disorders in adolescents: results of the Bremen adolescent study. Z Klin Psychol Psychopathol Psychother 1998;46:105 24. [45] Essau CA, Conradt J, Petermann F. Frequency and comorbidity of social phobia and social fears in adolescents. Behav Res Ther 1999;37:8321 843. [46] Essau CA, Conradt J, Petermann F. Incidence of post-traumatic stress disorder in adolescents: results of the Bremen adolescent study. Z Kinder Jugenpsychiatr 1999;27:37 45. [47] Essau CA, Conradt J, Petermann F. Frequency of panic attacks and panic disorder in adolescents. Depress Anxiety 1999;9:19 26. [48] Warren SL, Huston L, Egeland B, et al. Child and adolescent anxiety disorders and early attachment. J Am Acad Child Adolesc Psychiatry 1997;36:637 44. [49] Kasen S, Cohen P, Skodol AE, et al. Influence of child and adolescent psychiatric disorders on young adult personality disorder. Am J Psychiatry 1999;156:1529 35. [50] Giaconia RM, Reinherz HZ, Silverman AB, et al. Traumas and posttraumatic stress disorder in a community population of older adolescents. J Am Acad Child Adolesc Psychiatry 1995; 34:1369 80. [51] Essau CA, Conradt J, Petermann F. Frequency and comorbidity of social anxiety and social phobia in adolescents: Results of a Bremen adolescent study. Fortschr Neurol Psychiatr 1998; 66:524 30. [52] Costello AJ, Edelbrock CS, Dulcan MK, et al. Development and testing of the NIMH diagnostic interview schedule for children in a clinic population: final report. Contract no. Rfp-db-810027. Rockville (MD)7 NIMH Center for Epidemiologic Studies; 1984. [53] Costello EJ, Burns BJ, Angold A, et al. How can epidemiology improve mental health services for children and adolescents? J Am Acad Child Adolesc Psychiatry 1993;32:1106 13. [54] US Government. Fed Regist 1993;58:29425. [55] Hodges K, Doucette-Gates A, Liao Q. The relationship between the child and adolescent functional assessment scale (CAFAS) and indicators of functioning. J Child Fam Stud 1999;8: 109 22. [56] Bird HR, Andrews H, Schwab-Stone M, et al. Global measures of impairment for epidemiologic and clinical use with children and adolescents. Int J Psych Res 1996;6:295 307. [57] Angold A, Costello EJ, Farmer EMZ, et al. Impaired but undiagnosed. J Am Acad Child Adolesc Psychiatry 1999;38:129 37.

648

costello et al

[58] Ezpeleta L, Keeler G, Erkanli A, et al. Epidemiology of psychiatric disability in childhood and adolescence. J Child Psychol Psychiatry 2001;42:901 14. [59] Angold A, Costello EJ, Erkanli A. Comorbidity. J Child Psychol Psychiatry 1999;40:57 87. [60] Armstrong TD, Costello EJ. Community studies on adolescent substance use, abuse, or dependence and psychiatric comorbidity. J Consult Clin Psychol 2002;70:1224 39. [61] Kaplow JB, Curran PJ, Angold A, et al. The prospective relation between dimensions of anxiety and the initiation of adolescent alcohol use. J Clin Child Psychol 2001;30:316 26. [62] Orvaschel H, Lewinsohn PM, Seeley JR. Continuity of psychopathology in a community sample of adolescents. J Am Acad Child Adolesc Psychiatry 1995;34:1525 35. [63] Perwien AR, Bernstein GA. Separation anxiety disorder. Phobic and Anxiety Disorders in Children and Adolescents 2004;10:272 305. [64] Klein RG. Is panic disorder associated with childhood separation anxiety disorder? Clin Neuropharmacol 1995;18:S7 14. [65] Breton J-J, Bergeron L, Valla J-P, et al. Quebec child mental health survey: prevalence of DSM-III-R mental health disorders. J Child Psychol Psychiatry 1999;40:375 84.