- Email: [email protected]
Comorbidity in Childhood Anxiety Disorders and Treatment Outcome
Comorbidity in Childhood Anxiety Disorders and Treatment Outcome PHILIP C. KENDALL, PH.D., ERIKA U. BRADY, PH.D., AND TIMOTHY L. VERDUIN, B.A.
ABSTRACT Objective: Psychiatric comorbidity is common in anxious children. The purpose of this study was to investigate the impact of comorbidity on treatment outcome in anxious children. Method: Participants were 173 children between the ages of 8 and 13 years who met primary DSM-III-R/DSM-IV diagnoses of separation anxiety disorder, overanxious disorder/generalized anxiety disorder, or avoidant disorder/social phobia assessed by the Anxiety Disorders Interview Schedule for Children (ADIS-C). The majority (79%) had at least one comorbid diagnosis. Participants were randomly assigned to cognitive-behavioral therapy or waitlist. Group differences in ADIS-C diagnoses were compared after treatment. Multiple parent and child self-report measures were used to measure symptoms as well. Results: Pretreatment comorbidity was not associated with differences in treatment outcome: 68.4% of noncomorbid participants and 70.6% of comorbid participants were free of their primary diagnosis after treatment. Regarding parent and child self-report symptoms, multivariate analyses of variance revealed significant time (treatment) main effects, but no significant main effect for group (comorbid status) or time/group interaction. Conclusions: The cognitive-behavioral treatment program was similarly effective in anxious children with and without comorbid disorders; both groups showed clinically significant reductions in pretreatment diagnoses and symptoms. J. Am. Acad. Child Adolesc. Psychiatry, 2001, 40(7):787–794. Key Words: childhood, anxiety, comorbidity, treatment outcome.
Numerous studies document the high rate of comorbidity for both children and adults presenting with an anxiety disorder (Bernstein, 1991; Brown and Barlow, 1992; Deas-Nesmith et al., 1998; Kendall, 1994; Kessler et al., 1994; Masi et al., 1999; Merikangas et al., 1998). Studies such as these have led researchers to question the current diagnostic categories (e.g., Achenbach, 1995; Caron and Rutter, 1991) and to investigate the role of comorbidity in treatment outcomes (Kendall and Clarkin, 1992). Effective medications and psychosocial treatments have been identified for anxiety disorders in adults. Research on the psychosocial treatment of adult anxiety disorders has indicated effective cognitive-behavioral treatments for specific anxiety disorders (Barlow et al., 1989, 1992; Heimberg et al., 1993). Until recently, research on anxiety Accepted January 23, 2001. From the Child and Adolescent Anxiety Disorders Clinic, Department of Psychology, Temple University, Philadelphia. Funding was provided by two research grants (e.g., no. 44042) awarded to the first author. Reprint requests to Dr. Kendall, Department of Psychology, Temple University, 1701 North 13th Street, Weiss Hall, Room 478, Philadelphia, PA 19122. 0890-8567/01/4007–0787䉷2001 by the American Academy of Child and Adolescent Psychiatry.
disorders in children has lagged behind that of adultfocused research, has emphasized specific fears and phobias, and has consisted largely of case studies (e.g., Albano and Chorpita, 1995; Francis and Beidel, 1995; Kendall et al., 1992). Recent reports of randomized, controlled trials document the efficacy of individual and familybased cognitive-behavioral treatment approaches for childhood anxiety disorders (Barrett et al., 1996; Kendall, 1994; Kendall and Southam-Gerow, 1996). This report will focus on comorbidity and psychosocial treatments for childhood anxiety disorders. Investigating comorbidity is important for several reasons. First, comorbid disorders may moderate treatment outcome. Second, the impact of therapy on comorbid diagnoses may help to elucidate the relationship between the comorbid disorders and thus help to clarify issues related to diagnostic validity. Finally, the effect of treatment on comorbid conditions may indicate the generalizability of certain treatment strategies (Brown et al., 1995). Despite the widespread recognition of significant rates of comorbidity in subjects with anxiety disorders, few studies have addressed the impact of comorbidity on treatment efficacy. There are several possible reasons for this. Brown
J . A M . A C A D . C H I L D A D O L E S C . P S YC H I AT RY, 4 0 : 7 , J U LY 2 0 0 1
787
KENDALL ET AL.
and Barlow (1992) postulated that comorbidity has been understudied because of small sample sizes (inadequate to analyze the impact of specific patterns of comorbidity), the neglect of assessment for the presence of comorbidity, and the relative infancy of many treatments for certain disorders. Kearney and Silverman (1998) suggested that some studies do not address comorbidity because the presence of comorbid disorders was treated as an exclusionary criterion. The findings from the few studies of comorbidity and psychosocial treatments for adult anxiety disorders (e.g., Barlow, 1994; Noyes et al., 1990) have not been consistent. Certain comorbid conditions were associated with poorer outcome (Chambless et al., 1992; Turner, 1987), but other studies did not replicate these findings (Basoglu et al., 1988) and others reported that anxiety-focused treatments reduced comorbidity from pretreatment to posttreatment (Borkovec et al., 1995; Brown et al., 1995). With regard to treatments for anxious youths, few studies have examined the influence of comorbidity. Barrett and colleagues (1996) reported a significant reduction in self-reported depressive symptoms, and similar results were found by Mendlowitz et al. (1999) and Barrett (1998). Other studies have investigated the relationship of comorbid anxiety symptoms and treatment for other primary disorders. Studies by Reynolds and colleagues examined the effect of treatment for depression on anxiety symptoms in children (Reynolds and Coats, 1986; Stark et al., 1987) and found that several depressionfocused treatments reduced both depressive and anxious symptoms. The MTA Cooperative Group (1999) found that children with attention-deficit/hyperactivity disorder (ADHD) and comorbid anxiety who were receiving behavioral, psychopharmacological, or combined treatment showed reduced parent-reported ADHD and internalizing symptoms, although treatment did not specifically target anxiety. Behavior therapy was more effective than community care in anxious participants, but not in nonanxious participants. A later MTA report (March et al., 2000) suggested that parent-reported child anxiety in this sample reflected greater levels of negative affectivity and social problems and lower levels of anxious symptoms than those common to children with primary anxiety disorders. In addition, children with comorbid anxiety, conduct disorder, and ADHD were found to respond less favorably to psychosocial treatment than did anxious children with ADHD but not conduct disorder. The purpose of the present study is to examine, in youths with anxiety disorder, (1) the nature and occur788
rence of comorbidity in youths presenting for treatment of an anxiety disorder, (2) the effect of comorbidity on treatment outcome, and (3) the impact of treatment on comorbid diagnoses. Three main questions will be addressed: (1) Are patterns of comorbidity associated with differences in patient characteristics? (2) Does the presence of a comorbid disorder lessen treatment response? (3) Does treatment for the primary anxiety disorder lead to improvement in comorbid conditions? METHOD Participants Participants were 173 children between the ages of 8 and 13 years (mean = 11.2). Some of these participants were included in two previously reported studies (Kendall, 1994; Kendall et al., 1997), but the present sample also contains participants not included in the previous published reports of outcome. Although there was overlap in participants, this study was not a replication of prior studies; rather, the present study is a more comprehensive examination of the role of comorbidity. Children were referred from multiple community sources to the Child and Adolescent Anxiety Disorders Clinic and were included if they met diagnostic criteria for a primary DSM-III-R or DSM-IV diagnosis of separation anxiety disorder, overanxious disorder or generalized anxiety disorder (GAD), and avoidant disorder or social phobia. Kendall and Warman (1996) found that DSM-III-R and DSM-IV diagnostic categories resulted in similar cases; they found that children with a diagnosis of overanxious disorder were indistinguishable from children with a diagnosis of GAD, so these two diagnostic categories were combined. Likewise, children with avoidant disorder were similar to children with social phobia, and these categories were combined. The child and his or her parent(s) were interviewed separately. For this study, the parent interview was used to determine the presence of a primary anxiety disorder. Although the current Anxiety Disorders Interview Schedule for Children (ADIS-C) (Silverman and Albano, 1997) provides for a composite diagnosis, the ADIS-C used at the time of this study did not. Therefore, it was determined that the parent diagnosis was closest to this criterion. Exclusion criteria were few: the presence of a disabling physical condition, psychotic symptoms, or current use of antianxiety/antidepressant medication. Table 1 presents the distribution of primary diagnoses by parent report. Boys outnumbered girls (107:66), and the majority of participants were white (85%). Although our samples have a balance of males and females, the present sample probably resulted from the selection definition for the group with comorbid externalizing disorders. TABLE 1 Distribution of Primary Diagnoses (N = 173) Diagnosis GAD SAD SP
%
n
58.4 22.2 18.8
101 39 33
Note: GAD = generalized anxiety disorder; SAD = separation anxiety disorder; SP = social phobia.
J . A M . A C A D . C H I L D A D O L E S C . P S YC H I AT RY, 4 0 : 7 , J U LY 2 0 0 1
COMORBIDITY IN CHILDHOOD ANXIETY DISORDERS
Measures Anxiety Disorders Interview Schedule for Children. The revised ADISC/P (Silverman, 1991; Silverman and Albano, 1997) is a structured diagnostic interview that assesses anxiety disorders and other diagnoses. Interviewers were trained to reliability criterion (minimum κ = 0.80). Adequate interrater reliability has been reported (Barrett et al., 1996; Silverman and Nelles, 1988). Revised Children’s Manifest Anxiety Scale. The Revised Children’s Manifest Anxiety Scale is a 37-item scale that measures trait anxiety. It has a nine-item Lie scale and three anxiety factors: Physiological, Worry and Oversensitivity, and Concentration (Reynolds and Richmond, 1978, 1985). This measure has high internal consistency (Total Anxiety α coefficient = .83) and retest reliability (r = 0.68 after 9 months), and normative data are available (Reynolds and Richmond, 1985). Children’s Depression Inventory. The Children’s Depression Inventory (Kovacs, 1992) includes 27 items assessing the cognitive, affective, and behavioral signs of depression. The scale has high internal consistency (α coefficients = .71–.89), and item-total correlations were reported to be acceptable (Kovacs, 1992). Numerous studies have reported retest reliability; results indicate moderate stability. The scale correlates in the expected direction with measures of related constructs (Kendall et al., 1989; Kovacs, 1992), and normative data are available (Kovacs, 1992). State-Trait Anxiety Inventory for Children. The State-Trait Anxiety Inventory for Children has 20 items that assess how the child usually feels (A-Trait) and 20 assessing how the child currently feels (A-State). The A-Trait scale has adequate retest reliability (coefficients of 0.65 for males, 0.71 for females after 6 weeks). The stability of the correlation coefficients for the A-State scale was significantly lower (0.31 and 0.47), as expected given that it targets situational anxiety. The internal consistency of both scales is acceptable. Concurrent validity for the A-Trait scale is adequate, and normative data are available (Spielberger, 1973). Child Behavior Checklist. The 118-item Child Behavior Checklist (CBCL) assesses behavioral problems and social competencies. The mean retest reliability for the problems scale scores was 0.89 over 1 week and 0.75 over 1 year. The CBCL scale scores were favorably correlated with similar scales. Normative data are available (Achenbach and Edelbrock, 1991). Teacher’s Report Form. The 118-item Teacher’s Report Form (TRF), which is similar to the CBCL, provides information on the child’s classroom behavior. The mean retest reliability for the problems scale scores was 0.92, 0.75, and 0.66 after 2 weeks, 2 months, and 4 months, respectively. TRF scales are favorably correlated with similar scales. Normative data are available (Achenbach, 1991). Procedure Children and their parents were initially screened over the telephone. Eligible families were invited for the diagnostic intake. Parent self-report forms, the TRF, and a demographic information sheet were sent home to be completed and returned. The 3- to 4-hour intake consisted of separate diagnostic interviews with the child and the parents. Children completed self-report forms while parents were interviewed, and parents completed forms during the child interview. Children who met criteria were assigned randomly to the treatment or waitlist condition. Participants with ADHD taking medications at the start of their participation were not asked to discontinue medication. Treatment-condition children were randomly assigned to a therapist. Waitlist-condition children were assessed again after the 8-week waitlist and then were assigned randomly to a therapist. Treatment consisted of 16 to 20 weekly hour-long sessions. The first half of the cognitive-behavioral manual-based intervention (Kendall
J . A M . A C A D . C H I L D A D O L E S C . P S YC H I AT RY, 4 0 : 7 , J U LY 2 0 0 1
et al., 1990) focused on the recognition of anxious symptoms as cues and consisted of four steps: (1) recognizing anxious feelings and somatic reactions, (2) recognizing and modifying anxious cognition, (3) developing a plan to cope with the anxiety, and (4) evaluating performance and administering self-reward as appropriate. The second half of the program involved in vivo exposure to anxious situations. RESULTS
Although the results presented here are based on diagnoses derived from parent interviews, comparable results were found when the child interview data were analyzed. Nature and Characteristics of Pretreatment Comorbidity
The most frequent comorbid diagnoses for children with a primary anxiety disorder were simple phobia, social phobia, and GAD (Table 2). The overall rate and pattern of comorbidity was similar for children with each of the three primary anxiety disorders. To examine the role of comorbidity, children were divided into three groups according to parent-reported diagnoses: (1) those with only a primary anxiety disorder (ANX-only), (2) those with two or more anxiety disorders (ANX/ANX), and (3) those with comorbid externalizing disorders (ANX/EXT). There were only eight children with a comorbid diagnosis of major depression or dysthymia, a number too small to permit meaningful analysis of this group. These groups were compared with respect to age, gender, family income, TABLE 2 Distribution of Comorbid Diagnoses at Pretreatment Diagnosis GAD SAD SP SIM PTSD AG OCD PD SCH DEP ADHD ODD CD
%
n
29.0 17.0 33.5 46.2 0.0 0.0 1.2 1.7 20.8 4.6 15.0 9.2 1.2
50 29 58 80 0 0 2 3 36 8 26 16 2
Note: GAD = generalized anxiety disorder; SAD = separation anxiety disorder; SP = social phobia; SIM = simple phobia; PTSD = posttraumatic stress disorder; AG = agoraphobia; OCD = obsessivecompulsive disorder; PD = panic disorder; SCH = school phobia; DEP = major depression/dysthymia; ADHD = attention-deficit/ hyperactivity disorder; ODD = oppositional defiant disorder; CD = conduct disorder.
789
KENDALL ET AL.
parental education level, and race via analyses of variance (ANOVAs) or χ2 tests. Table 3 displays the demographic information grouped by diagnosis and were found not to differ (all p values > .05). However, χ2 analysis revealed a significant difference with regard to gender (χ2 [2, N = 173] = 8.2, p < .05). The ANX/EXT group contained more males, consistent with the greater male-to-female ratio expected in these disorders. Multivariate analysis of variance (MANOVA) indicated that the groups differed significantly with respect to pretreatment internalizing symptom severity as assessed by child self-report questionnaires (F2,171 = 0.88, p < .05). For parent report of internalizing symptom severity, ANOVAs indicated that the groups differed significantly (F2,172 = 5.1, p < .05). Scheffé post hoc analyses indicated that the ANX/ANX group and the ANX/EXT group displayed more severe internalizing symptoms than the ANX-only group, but did not differ from each other. For parent report of externalizing symptom severity, ANOVAs indicated that the groups differed significantly (F2,172 = 25.2, p < .05). Scheffé post hoc analyses indicated that the ANX/ EXT group displayed more severe externalizing symptoms than the ANX-only and ANX/ANX groups, but the latter two groups did not differ significantly from each other. According to teacher reports of internalizing symptoms, the groups did not differ significantly (p = .99). However, teacher reports of externalizing symptoms indicated that the groups differed significantly (F2,167 = 3.9, p < .05). Scheffé post hoc analyses indicated that the ANX/EXT group displayed more severe externalizing symptoms than the ANX-only group and the ANX/ANX group, but the latter two groups did not differ significantly from each other. Occasional missing forms resulted in missing data, but analyses revealed no pattern to the missing scores. Table 4 presents the means for the dependent measures.
Impact of Comorbidity on Treatment Outcome
Note: ANX = primary anxiety disorder only; ANX/ANX = primary anxiety disorder and comorbid anxiety; ANX/EXT = primary anxiety disorder and a comorbid externalizing disorder.
To examine whether pretreatment comorbidity moderated treatment response, a χ2 analysis was conducted with a categorical measure of treatment outcome (presence or absence of pretreatment primary diagnosis at posttreatment). These analyses indicated that individuals with comorbid diagnoses at pretreatment were not significantly less likely to respond to treatment as indicated by remission of the primary diagnosis (all χ2 values < 1). The results of these analyses are depicted in Figure 1. For example, 68.4% of patients with no comorbid diagnoses and 70.6% of patients with one or more comorbid diagnoses were free of their primary diagnosis at the conclusion of treatment. To examine the impact of comorbidity on treatment outcome as indicated by changes on dependent measures (child self-report measures, parent report measures, and teacher report measure), 3 (groups, between subjects) ⫻ 2 (assessment periods, within subjects) MANOVAs or ANOVAs were conducted. The MANOVA revealed significant differences for the assessment period main effect (F1,148 = 3.90, p < .05), indicating that treatment produced gains from pretreatment to posttreatment. However, there was no main effect for group (p = .19) and no interaction between group and time (p = .21), indicating that the presence of comorbidity was not associated with significant differential treatment effects. Analyses yielded significant main effects for time (preto posttreatment) on the parent measures of internalizing and externalizing symptom severity and the teacher measure of internalizing symptom severity (all p values < .05). No significant association was found between treatment and change in severity of teacher-reported externalizing symptoms. Comorbidity was not significantly associated with pre- to posttreatment changes in severity of parentreported internalizing symptoms (p = .57) or externalizing symptoms (p = .94) on the CBCL. Similarly, nonsignificant interactions were found between comorbidity and pre- to posttreatment changes in teacher-reported internalizing symptoms (p = .19) and externalizing symptoms (p = .22). Table 4 presents means for dependent measures from pretreatment to posttreatment and follow-up. To examine the relationship of continued comorbidity (comorbid diagnoses at posttreatment) to outcome, with absence of primary anxiety disorder diagnosis as the outcome measure, a χ2 analysis was conducted. This analysis indicated that patients with continued presence of comorbidity were less likely to show remission of
790
J . A M . A C A D . C H I L D A D O L E S C . P S YC H I AT RY, 4 0 : 7 , J U LY 2 0 0 1
TABLE 3 Demographic Characteristics of Diagnostic Groups Variable Gender (n) Male Female Age (M) Race (n) White African American
ANX
ANX/ANX
ANX/EXT
20 15 11.2
48 38 11.3
35 9 10.9
31 4
74 12
35 9
COMORBIDITY IN CHILDHOOD ANXIETY DISORDERS
TABLE 4 Group Means and Standard Deviations: Pretreatment to Follow-up Pretreatment Scale RCMAS M SD A-Trait M SD A-State M SD CDI M SD CBCL-I M SD CBCL-E M SD TRF-I M SD TRF-E M SD
Posttreatment
Follow-up
ANX
ANX/ANX
ANX/EXT
ANX
ANX/ANX
ANX/EXT
ANX
ANX/ANX
ANX/EXT
51.2 12.3
53.1 10.3
54.3 11.5
40.7 12.4
42.7 10.5
42.2 10.6
41.9 9.1
41.4 12.0
41.9 8.4
47.9 13.9
51.3 12.7
47.5 11.9
35.9 13.0
51.9 12.4
48.1 12.3
37.3 10.5
38.5 13.2
37.6 12.5
52.8 9.6
52.9 11.0
50.1 11.5
42.5 8.5
45.2 10.2
45.9 12.9
45.4 10.2
47.7 11.2
48.2 8.8
7.4 6.4
11.2 8.5
11.2 9.3
4.3 4.8
5.3 4.9
6.1 7.4
4.2 3.6
5.4 7.5
5.8 7.5
66.6 8.0
71.1 7.8
70.9 6.8
57.5 9.1
60.0 11.8
64.2 9.8
56.3 8.0
58.6 12.4
60.2 9.8
47.6 8.7
50.8 8.7
59.6 7.2
44.6 9.4
45.7 9.0
56.1 8.2
46.0 8.5
46.5 10.9
56.5 9.9
61.7 12.2
61.7 11.7
61.5 11.0
58.5 10.2
56.0 11.0
57.6 11.5
61.4 11.3
56.8 9.7
51.4 10.1
49.6 7.6
49.5 7.0
53.5 10.2
47.9 6.5
46.4 7.0
52.6 12.2
48.1 8.6
47.7 7.8
48.6 10.1
Note: ANX = Primary anxiety disorder only; ANX/ANX = primary anxiety disorder and comorbid anxiety; ANX/EXT = primary anxiety disorder and a comorbid externalizing disorder. RCMAS = Revised Children’s Manifest Anxiety Scale; A-Trait = State-Trait Anxiety Inventory for Children, Trait subscale; A-State = State-Trait Anxiety Inventory for Children, State subscale; CDI = Children’s Depression Inventory; CBCLI = Child Behavior Checklist, Internalizing scale; CBCL-E = Child Behavior Checklist, Externalizing scale; TRF-I = Teacher’s Report Form, Internalizing scale; TRF-E = Teacher’s Report Form, Externalizing scale.
J . A M . A C A D . C H I L D A D O L E S C . P S YC H I AT RY, 4 0 : 7 , J U LY 2 0 0 1
791
% of Patients Improved Fig. 1 Percentage of patients who improved. ANX = primary anxiety disorder only; ANX/ANX = primary anxiety disorder and comorbid anxiety; ANX/ EXT = primary anxiety disorder and a comorbid externalizing disorder.
their primary pretreatment diagnosis (χ2 [2, N = 129] = 7.2, p < .05). The impact of comorbidity on maintenance of treatment gains was examined with separate 3 (groups; between subjects) ⫻ 2 (assessment periods; posttreatment to follow-up, within subjects) MANOVAs. No significant change was found in level of symptoms (p = .07), although the continued decrease in self-reported symptoms approached significance. There was no effect of comorbidity on maintenance of treatment gains (p = .45), and no interaction effect between comorbidity and gains (p = .90). Because attrition between posttreatment and follow-up may have influenced these results, participants who completed the follow-up assessment were compared with participants who declined to complete the follow-up. These groups were compared with t tests or χ2 tests and found not to differ significantly with regard to demographic variables, pre/posttreatment diagnostic status, or pre/posttreatment dependent variables (e.g., self- and parent reports).
KENDALL ET AL.
The present results were consistent with published reports documenting a high rate of comorbidity in children with anxiety disorders (Brady and Kendall, 1992). The majority of children presenting for treatment with a primary anxiety disorder had at least one comorbid diagnosis (79%). The most common additional diagnoses were secondary anxiety disorders including GAD, simple phobia, social phobia, and separation anxiety disorder. Also
consistent with the literature, comorbidity was associated with greater severity of internalizing symptoms (Brown et al., 1995; Manassis and Menna, 1999; Nottelmann and Jensen, 1995). Pretreatment comorbidity was not associated with treatment outcome or maintenance of treatment gains at follow-up, defined broadly. However, those individuals who continued to have comorbid diagnoses at the end of the treatment were significantly less likely to show remission of their primary anxiety diagnosis (see also Borkovec et al., 1995; Brown et al., 1995). The present treatment, designed to address the child’s primary anxiety disorder, was nevertheless associated with a reduction in comorbid diagnoses. The literature suggests that comorbidity is associated with greater severity of symptoms, more persistent difficulties, and a greatly increased likelihood of seeking treatment (Manassis and Menna, 1999; McGee et al., 1990; Nottelmann and Jensen, 1995). The finding of increased severity of symptoms in comorbid groups in this sample is consistent with published reports, yet it raises issues regarding nosology. Is comorbidity within the anxiety disorders simply a proxy for general severity rather than an indication of an overlap of different types of psychopathology? Although the present study does not directly test this notion, the fact that children with multiple disorders exhibited greater levels of internalizing symptoms is consistent with the notion of proxy. The findings that comorbidity did not moderate response to treatment and that the treatment yielded a reduction in multiple anxiety diagnoses also seem to support the notion of a general anxiety syndrome. Brown and Barlow (1992) and Borkovec et al. (1995) articulated three reasons why treatment of a primary anxiety disorder might contribute to the reduction of comorbid diagnoses. First, the patient may generalize treatment skills and apply them to other areas not specifically targeted. Second, many disorders, particularly anxiety disorders, share overlapping features and symptoms, and improvement in some of these overlapping symptoms may account for reduced comorbidity after treatment. Third, there may be basic underlying psychological processes common to anxiety disorders even if secondary differentiating processes and symptoms are present. In this study, presence of disorders other than comorbid anxiety disorders did not moderate response to treatment, although this may have been due to the large treatment effect. Rather, there was a reduction in the rate of ADHD
792
J . A M . A C A D . C H I L D A D O L E S C . P S YC H I AT RY, 4 0 : 7 , J U LY 2 0 0 1
Impact of Treatment on Comorbid Conditions
To evaluate whether treatment for a primary anxiety disorder resulted in changes in comorbid diagnoses, a McNemar test was conducted to determine whether the overall rate of comorbidity changed significantly from pretreatment to posttreatment and follow-up. At pretreatment, 79% of the children had at least one additional diagnosis; at posttreatment, this rate had decreased to 47%. At follow-up, only 20% of the children continued to evidence comorbid diagnoses. These reductions in the rate of comorbidity were significant (McNemar test, p values < .05). Additional McNemar tests were conducted to determine whether anxiety treatment was associated with the reduction in specific comorbid diagnoses apart from anxiety diagnoses. At pretreatment, 15% of the children had comorbid ADHD. At posttreatment, this rate was 4.7%, and at follow-up it was 3.6%. These results were significant (McNemar test, p < .05). Similarly, there was a significant reduction in the rate of oppositional defiant disorder (ODD) from pretreatment (9.2%) to follow-up (1.8%), and posttreatment to follow-up, but not from pretreatment to posttreatment (6.8%) (McNemar test, p < .05). Because at least one study in the adult literature (Borkovec et al., 1995) indicated that the effectiveness of treatment for the primary condition affects reduction in comorbidity, differential decrease in comorbidity in treatment responders was evaluated. Participants were categorized into either a treatment success group (no longer met criteria for the pretreatment primary anxiety disorder diagnosis) or a treatment nonsuccess group. Chi-square analyses examined the rate of comorbidity in treatment responders versus treatment nonresponders at posttreatment and at follow-up. Treatment responders were significantly more likely to show a reduction in comorbidity at posttreatment (χ2 [1, N = 129] = 7.2, p < .05) and at follow-up (χ2 [1, N = 126] = 4.0, p < .05). DISCUSSION
COMORBIDITY IN CHILDHOOD ANXIETY DISORDERS
and ODD as well as in the overall rate of comorbidity. These results suggest that the treatment program may have robust effects, even on disorders that are not directly targeted for intervention and that are conceptualized as distinct from anxiety disorders. Reduction in comorbid anxiety diagnoses can be explained in terms of the three hypotheses listed above, as well as the capacity of the therapists to flexibly apply the manual-based treatment to address secondary anxiety-based concerns. However, it is more difficult to account for reduction in disorders that are not theoretically related to anxiety or the targets of an anxiety treatment program. For example, although ADHD and anxiety disorders do share some overlapping features, it is difficult to account for the decrease in diagnoses of ADHD on this basis alone. In some cases, a change in overlapping targeted symptoms may have resulted in a child’s not meeting criteria for ADHD (e.g., relaxation training helping a child to reduce excess fidgeting, a symptom of both anxiety and ADHD). An alternative explanation for the reduction in the rate of ADHD and ODD is that these disorders were causally related to the anxiety disorder or were exacerbated by anxiety. The present results indicate that children who evidenced a clinically significant change in anxiety problems were significantly more likely to show remission of comorbid diagnoses at posttreatment and follow-up. This finding could be interpreted as support for the argument that anxiety was exacerbating symptoms of ADHD or ODD. For example, anxiety might interfere with task-focused attention, and an anxious child’s behavior may appear similar to that of a child with ADHD under conditions of anxious arousal. Furthermore, there is some evidence that the presence of an anxiety disorder moderates the response to medications used to treat ADHD (Tannock et al., 1995), although other research suggests that this effect is not present when ADHD medication is properly titrated (Diamond et al., 1999; MTA Cooperative Group, 1999). Although high rates of comorbid depression are often reported in clinic-based samples of children with anxiety disorders (e.g., Bernstein, 1991; Manassis and Menna, 1999), the rate of comorbid depressive disorders in this sample was low (only 4.6%). This rate, however, is consistent with the rate of comorbid depressive disorders reported from a clinic sample in Australia (Barrett et al., 1996) with a similar age range (7–14 years). One reason for this lower rate of depressive disorders may be the restricted age range. Children with comorbid depression often are older than children with only anxiety disorders,
there is evidence suggesting a developmental progression (Kovacs et al., 1989), and the diagnosis of depression increases significantly after puberty, particularly in girls (e.g., Bird et al., 1993).
J . A M . A C A D . C H I L D A D O L E S C . P S YC H I AT RY, 4 0 : 7 , J U LY 2 0 0 1
793
Limitations
In general, comorbidity was defined in terms of broad categories (i.e., ANX/ANX, ANX/EXT), not specific disorders. Although this was necessary because of the numerous specific comorbid patterns, our results do not entirely address specific comorbid conditions. For example, as previously mentioned, the number of children with comorbid depression and anxiety was not large enough to permit meaningful examination of this subtype of comorbidity. This study relied on parent interviews for determinations of comorbidity. Changes in comorbid conditions may reflect actual child improvements or, because they are based on parent interviews, may also reflect parents’ lack of specificity in reporting changes. In addition, although diagnoses derived from parent and child reports were comparable, children in this study generally reported fewer disorders than did parents and tended to rate their symptoms as less severe on self-report measures than did their parents or teachers. It is unclear whether this phenomenon is a function of child underreporting or a reflection of limitations that children’s level of cognitive development places on their capacity to report on their own behavior and symptoms. Clinical Implications
The present results support the efficacy and robustness of cognitive-behavioral treatment for anxiety disorders in children with and without comorbidity. This finding is important because, frequently, children who present for treatment have comorbid disorders. This study also indicates that comorbid diagnoses were alleviated by anxietyfocused treatment, even though this treatment did not specifically target comorbid diagnoses. Although these findings are promising, further research is needed to examine the relationship of specific comorbid patterns on treatment outcome. REFERENCES Achenbach TM (1991), Integrative Guide for the 1991 CBCL/4–18, YSR, and TRF. Burlington: University of Vermont Department of Psychiatry Achenbach TM (1995), Diagnosis, assessment, and comorbidity in psychosocial treatment research. J Abnorm Child Psychol 23:45–65 Achenbach TM, Edelbrock C (1991), Manual for the CBCL and 1991 Profile. Burlington: University of Vermont Department of Psychiatry Albano AM, Chorpita BF (1995), Treatment of anxiety disorders of childhood. Psychiatr Clin North Am 18:767–784
KENDALL ET AL.
Barlow DH (1994), Comorbidity in social phobia: implications for cognitivebehavioral treatment. Bull Menninger Clin 58:43–57 Barlow DH, Craske MG, Cerny JA, Klosko JS (1989), Behavioral treatment of panic disorder. Behav Ther 20:261–282 Barlow DH, Rapee RM, Brown TA (1992), Behavioral treatment of generalized anxiety disorder. Behav Ther 23:551–570 Barrett PM (1998), Evaluation of cognitive-behavioral group treatments for childhood anxiety disorders. J Clin Child Psychol 27:459–468 Barrett PM, Dadds MR, Rapee RM (1996), Family treatment of childhood anxiety: a controlled trial. J Consult Clin Psychol 64:333–342 Basoglu M, Lax T, Kasvikis Y, Marks IM (1988), Predictors of improvement in obsessive-compulsive disorder. J Anxiety Disord 2:299–317 Bernstein GA (1991), Comorbidity and severity of anxiety and depressive disorders in a clinic sample. J Am Acad Child Adolesc Psychiatry 30:43–50 Bird HR, Gould MS, Staghezza BM (1993), Patterns of diagnostic comorbidity in a community sample of children aged 9 through 16 years. J Am Acad Child Adolesc Psychiatry 32:361–368 Borkovec TD, Abel JL, Newman H (1995), Effects of psychotherapy on comorbid conditions in generalized anxiety disorder. J Consult Clin Psychol 63:479–483 Brady EU, Kendall PC (1992), Comorbidity of anxiety and depression in children and adolescents. Psychol Bull 111:244–255 Brown TA, Antony MM, Barlow DH (1995), Diagnostic comorbidity in panic disorder: effect on treatment outcome and course of comorbid diagnoses following treatment. J Consult Clin Psychol 63:408–418 Brown TA, Barlow DH (1992), Comorbidity among anxiety disorders: implications for treatment and DSM-IV. J Consult Clin Psychol 60:835–840 Caron C, Rutter M (1991), Comorbidity in child psychopathology: concepts, issues, and research strategies. J Child Psychol Psychiatry 32:1063–1080 Chambless DL, Renneberg B, Goldstein A, Gracely EJ (1992), MCMI-diagnosed personality disorders among agoraphobic outpatients: prevalence and relationship to severity and treatment outcome. J Anxiety Disord 6:193–211 Deas-Nesmith D, Brady KT, Campbell S (1998), Comorbid substance use and anxiety disorders in adolescents. J Psychopathol Behav Assess 20:139–148 Diamond IR, Tannock R, Schachar RJ (1999), Response to methylphenidate in children with ADHD and comorbid anxiety. J Am Acad Child Adolesc Psychiatry 38:402–409 Francis G, Beidel D (1995), Cognitive-behavioral psychotherapy. In: Anxiety Disorders in Children and Adolescents, March JS, ed. New York: Guilford, pp 321–340 Heimberg RG, Salzman DG, Holt CS, Blendell KA (1993), Cognitivebehavioral group treatment for social phobia: effectiveness at five-year follow-up. Cognit Ther Res 17:325–339 Kearney CA, Silverman WK (1998), A critical review of pharmacotherapy for youth with anxiety disorders: things are not as they seem. J Anxiety Disord 12:83–102 Kendall PC (1994), Treating anxiety disorders in youth: results of a randomized clinical trial. J Consult Clin Psychol 62:100–110 Kendall PC, Cantwell DP, Kazdin AE (1989), Depression in children and adolescents: assessment issues and recommendations. Cognit Ther Res 13:109–146 Kendall PC, Clarkin JF (1992), Introduction to special section: comorbidity and treatment implications. J Consult Clin Psychol 60:833–835 Kendall PC, Flannery-Schroeder E, Panichelli-Mindel SM, Southam-Gerow M, Henin A, Warman M (1997), Therapy for youths with anxiety disorders: a second randomized clinical trial. J Consult Clin Psychol 65:366–380 Kendall PC, Kane M, Howard B, Siqueland L (1990), Cognitive-Behavioral Therapy for Anxious Children: Treatment Manual. Ardmore, PA: Workbook Publishing Kendall PC, Kortlander E, Chansky TE, Brady EU (1992), Comorbidity of anxiety and depression in youth: treatment implications. J Consult Clin Psychol 60:869–880
Kendall PC, Southam-Gerow MA (1996), Long-term follow-up of a cognitivebehavioral therapy for anxiety-disordered youth. J Consult Clin Psychol 64:724–730 Kendall PC, Warman M (1996), Anxiety disorders in youth: diagnostic consistency across DSM-III-R and DSM-IV. J Anxiety Disord 10:453–463 Kessler RC, McGonagle KA, Zhao S et al. (1994), Lifetime and 12-month prevalence of DSM-III-R psychiatric disorders in the United States: results from the National Comorbidity Study. Arch Gen Psychiatry 51:8–19 Kovacs M (1992), Children’s Depression Inventory: Manual. Los Angeles: Western Psychological Services Kovacs M, Gatsonis C, Paulauskas SL, Richards C (1989), Depressive disorders in childhood, IV: a longitudinal study of comorbidity with and risk for anxiety disorders. Arch Gen Psychiatry 46:776–782 Manassis K, Menna R (1999), Depression in anxious children: possible factors in comorbidity. Depress Anxiety 10:18–24 March JS, Swanson JM, Arnold LE et al. (2000), Anxiety as a predictor and outcome variable in the Multimodal Treatment Study of Children with ADHD (MTA). J Abnorm Child Psychol 28:527–541 Masi G, Mucci M, Favilla L, Romano R, Poli P (1999), Symptomatology and comorbidity of generalized anxiety disorder in children and adolescents. Compr Psychiatry 40:210–215 McGee R, Feehan M, Williams S, Partridge F, Silva PA, Kelly J (1990), DSMIII disorders in a large sample of adolescents. J Am Acad Child Adolesc Psychiatry 29:611–619 Mendlowitz SL, Manassis K, Bradley S, Scapillato D, Miezitis S, Shaw BF (1999), Cognitive-behavioral group treatments in childhood anxiety disorders: the role of parental involvement. J Am Acad Child Adolesc Psychiatry 38:1223–1229 Merikangas KR, Stevens DE, Fenton B et al. (1998), Co-morbidity and familial aggregation of alcoholism and anxiety disorders. Psychol Med 28:733–788 MTA Cooperative Group (1999), Moderators and mediators of treatment response for children with attention deficit/hyperactivity disorder. Arch Gen Psychiatry 56:1088–1096 Nottelmann ED, Jensen PS (1995), Comorbidity of disorders in children and adolescents: developmental perspectives. Adv Clin Child Psychol 17:109–155 Noyes R, Reich J, Christiansen J, Suelzer M, Pfohl B, Coryell WA (1990), Outcome of panic disorder: relationship to diagnostic subtypes and comorbidity. Arch Gen Psychiatry 47:809–818 Reynolds CR, Richmond BO (1978), What I Think and Feel: a revised measure of children’s manifest anxiety. J Abnorm Child Psychol 6:271–280 Reynolds CR, Richmond BO (1985), Revised Children’s Manifest Anxiety Scale (RCMAS): Manual. Los Angeles: Western Psychological Services Reynolds WM, Coats KI (1986), A comparison of cognitive-behavioral therapy and relaxation training for the treatment of depression in adolescents. J Consult Clin Psychol 54:653–660 Silverman W (1991), Anxiety Disorders Interview Schedule for Children. Albany, NY: Graywind Silverman W, Albano AM (1997), The Anxiety Disorders Interview Schedule for Children (DSM-IV). San Antonio, TX: Psychological Corporation Silverman WK, Nelles WB (1988), The Anxiety Disorders Interview Schedule for Children. J Am Acad Child Adolesc Psychiatry 27:772–778 Spielberger C (1973), Manual for the State-Trait Anxiety Inventory for Children. Palo Alto, CA: Consulting Psychologists Press Stark KD, Reynolds WM, Kaslow KJ (1987), A comparison of the relative efficacy of self-control therapy and a behavioral problem-solving therapy for depression in children. J Abnorm Child Psychol 15:91–113 Tannock R, Ickowicz A, Schachar R (1995), Differential effects of methylphenidate on working memory in ADHD children with and without comorbidity. J Am Acad Child Adolesc Psychiatry 34:886–896 Turner RM (1987), The effects of personality disorder diagnosis on the outcome of social anxiety symptom reduction. J Personal Disord 1:136–143
794
J . A M . A C A D . C H I L D A D O L E S C . P S YC H I AT RY, 4 0 : 7 , J U LY 2 0 0 1
ABSTRACT Objective: Psychiatric comorbidity is common in anxious children. The purpose of this study was to investigate the impact of comorbidity on treatment outcome in anxious children. Method: Participants were 173 children between the ages of 8 and 13 years who met primary DSM-III-R/DSM-IV diagnoses of separation anxiety disorder, overanxious disorder/generalized anxiety disorder, or avoidant disorder/social phobia assessed by the Anxiety Disorders Interview Schedule for Children (ADIS-C). The majority (79%) had at least one comorbid diagnosis. Participants were randomly assigned to cognitive-behavioral therapy or waitlist. Group differences in ADIS-C diagnoses were compared after treatment. Multiple parent and child self-report measures were used to measure symptoms as well. Results: Pretreatment comorbidity was not associated with differences in treatment outcome: 68.4% of noncomorbid participants and 70.6% of comorbid participants were free of their primary diagnosis after treatment. Regarding parent and child self-report symptoms, multivariate analyses of variance revealed significant time (treatment) main effects, but no significant main effect for group (comorbid status) or time/group interaction. Conclusions: The cognitive-behavioral treatment program was similarly effective in anxious children with and without comorbid disorders; both groups showed clinically significant reductions in pretreatment diagnoses and symptoms. J. Am. Acad. Child Adolesc. Psychiatry, 2001, 40(7):787–794. Key Words: childhood, anxiety, comorbidity, treatment outcome.
Numerous studies document the high rate of comorbidity for both children and adults presenting with an anxiety disorder (Bernstein, 1991; Brown and Barlow, 1992; Deas-Nesmith et al., 1998; Kendall, 1994; Kessler et al., 1994; Masi et al., 1999; Merikangas et al., 1998). Studies such as these have led researchers to question the current diagnostic categories (e.g., Achenbach, 1995; Caron and Rutter, 1991) and to investigate the role of comorbidity in treatment outcomes (Kendall and Clarkin, 1992). Effective medications and psychosocial treatments have been identified for anxiety disorders in adults. Research on the psychosocial treatment of adult anxiety disorders has indicated effective cognitive-behavioral treatments for specific anxiety disorders (Barlow et al., 1989, 1992; Heimberg et al., 1993). Until recently, research on anxiety Accepted January 23, 2001. From the Child and Adolescent Anxiety Disorders Clinic, Department of Psychology, Temple University, Philadelphia. Funding was provided by two research grants (e.g., no. 44042) awarded to the first author. Reprint requests to Dr. Kendall, Department of Psychology, Temple University, 1701 North 13th Street, Weiss Hall, Room 478, Philadelphia, PA 19122. 0890-8567/01/4007–0787䉷2001 by the American Academy of Child and Adolescent Psychiatry.
disorders in children has lagged behind that of adultfocused research, has emphasized specific fears and phobias, and has consisted largely of case studies (e.g., Albano and Chorpita, 1995; Francis and Beidel, 1995; Kendall et al., 1992). Recent reports of randomized, controlled trials document the efficacy of individual and familybased cognitive-behavioral treatment approaches for childhood anxiety disorders (Barrett et al., 1996; Kendall, 1994; Kendall and Southam-Gerow, 1996). This report will focus on comorbidity and psychosocial treatments for childhood anxiety disorders. Investigating comorbidity is important for several reasons. First, comorbid disorders may moderate treatment outcome. Second, the impact of therapy on comorbid diagnoses may help to elucidate the relationship between the comorbid disorders and thus help to clarify issues related to diagnostic validity. Finally, the effect of treatment on comorbid conditions may indicate the generalizability of certain treatment strategies (Brown et al., 1995). Despite the widespread recognition of significant rates of comorbidity in subjects with anxiety disorders, few studies have addressed the impact of comorbidity on treatment efficacy. There are several possible reasons for this. Brown
J . A M . A C A D . C H I L D A D O L E S C . P S YC H I AT RY, 4 0 : 7 , J U LY 2 0 0 1
787
KENDALL ET AL.
and Barlow (1992) postulated that comorbidity has been understudied because of small sample sizes (inadequate to analyze the impact of specific patterns of comorbidity), the neglect of assessment for the presence of comorbidity, and the relative infancy of many treatments for certain disorders. Kearney and Silverman (1998) suggested that some studies do not address comorbidity because the presence of comorbid disorders was treated as an exclusionary criterion. The findings from the few studies of comorbidity and psychosocial treatments for adult anxiety disorders (e.g., Barlow, 1994; Noyes et al., 1990) have not been consistent. Certain comorbid conditions were associated with poorer outcome (Chambless et al., 1992; Turner, 1987), but other studies did not replicate these findings (Basoglu et al., 1988) and others reported that anxiety-focused treatments reduced comorbidity from pretreatment to posttreatment (Borkovec et al., 1995; Brown et al., 1995). With regard to treatments for anxious youths, few studies have examined the influence of comorbidity. Barrett and colleagues (1996) reported a significant reduction in self-reported depressive symptoms, and similar results were found by Mendlowitz et al. (1999) and Barrett (1998). Other studies have investigated the relationship of comorbid anxiety symptoms and treatment for other primary disorders. Studies by Reynolds and colleagues examined the effect of treatment for depression on anxiety symptoms in children (Reynolds and Coats, 1986; Stark et al., 1987) and found that several depressionfocused treatments reduced both depressive and anxious symptoms. The MTA Cooperative Group (1999) found that children with attention-deficit/hyperactivity disorder (ADHD) and comorbid anxiety who were receiving behavioral, psychopharmacological, or combined treatment showed reduced parent-reported ADHD and internalizing symptoms, although treatment did not specifically target anxiety. Behavior therapy was more effective than community care in anxious participants, but not in nonanxious participants. A later MTA report (March et al., 2000) suggested that parent-reported child anxiety in this sample reflected greater levels of negative affectivity and social problems and lower levels of anxious symptoms than those common to children with primary anxiety disorders. In addition, children with comorbid anxiety, conduct disorder, and ADHD were found to respond less favorably to psychosocial treatment than did anxious children with ADHD but not conduct disorder. The purpose of the present study is to examine, in youths with anxiety disorder, (1) the nature and occur788
rence of comorbidity in youths presenting for treatment of an anxiety disorder, (2) the effect of comorbidity on treatment outcome, and (3) the impact of treatment on comorbid diagnoses. Three main questions will be addressed: (1) Are patterns of comorbidity associated with differences in patient characteristics? (2) Does the presence of a comorbid disorder lessen treatment response? (3) Does treatment for the primary anxiety disorder lead to improvement in comorbid conditions? METHOD Participants Participants were 173 children between the ages of 8 and 13 years (mean = 11.2). Some of these participants were included in two previously reported studies (Kendall, 1994; Kendall et al., 1997), but the present sample also contains participants not included in the previous published reports of outcome. Although there was overlap in participants, this study was not a replication of prior studies; rather, the present study is a more comprehensive examination of the role of comorbidity. Children were referred from multiple community sources to the Child and Adolescent Anxiety Disorders Clinic and were included if they met diagnostic criteria for a primary DSM-III-R or DSM-IV diagnosis of separation anxiety disorder, overanxious disorder or generalized anxiety disorder (GAD), and avoidant disorder or social phobia. Kendall and Warman (1996) found that DSM-III-R and DSM-IV diagnostic categories resulted in similar cases; they found that children with a diagnosis of overanxious disorder were indistinguishable from children with a diagnosis of GAD, so these two diagnostic categories were combined. Likewise, children with avoidant disorder were similar to children with social phobia, and these categories were combined. The child and his or her parent(s) were interviewed separately. For this study, the parent interview was used to determine the presence of a primary anxiety disorder. Although the current Anxiety Disorders Interview Schedule for Children (ADIS-C) (Silverman and Albano, 1997) provides for a composite diagnosis, the ADIS-C used at the time of this study did not. Therefore, it was determined that the parent diagnosis was closest to this criterion. Exclusion criteria were few: the presence of a disabling physical condition, psychotic symptoms, or current use of antianxiety/antidepressant medication. Table 1 presents the distribution of primary diagnoses by parent report. Boys outnumbered girls (107:66), and the majority of participants were white (85%). Although our samples have a balance of males and females, the present sample probably resulted from the selection definition for the group with comorbid externalizing disorders. TABLE 1 Distribution of Primary Diagnoses (N = 173) Diagnosis GAD SAD SP
%
n
58.4 22.2 18.8
101 39 33
Note: GAD = generalized anxiety disorder; SAD = separation anxiety disorder; SP = social phobia.
J . A M . A C A D . C H I L D A D O L E S C . P S YC H I AT RY, 4 0 : 7 , J U LY 2 0 0 1
COMORBIDITY IN CHILDHOOD ANXIETY DISORDERS
Measures Anxiety Disorders Interview Schedule for Children. The revised ADISC/P (Silverman, 1991; Silverman and Albano, 1997) is a structured diagnostic interview that assesses anxiety disorders and other diagnoses. Interviewers were trained to reliability criterion (minimum κ = 0.80). Adequate interrater reliability has been reported (Barrett et al., 1996; Silverman and Nelles, 1988). Revised Children’s Manifest Anxiety Scale. The Revised Children’s Manifest Anxiety Scale is a 37-item scale that measures trait anxiety. It has a nine-item Lie scale and three anxiety factors: Physiological, Worry and Oversensitivity, and Concentration (Reynolds and Richmond, 1978, 1985). This measure has high internal consistency (Total Anxiety α coefficient = .83) and retest reliability (r = 0.68 after 9 months), and normative data are available (Reynolds and Richmond, 1985). Children’s Depression Inventory. The Children’s Depression Inventory (Kovacs, 1992) includes 27 items assessing the cognitive, affective, and behavioral signs of depression. The scale has high internal consistency (α coefficients = .71–.89), and item-total correlations were reported to be acceptable (Kovacs, 1992). Numerous studies have reported retest reliability; results indicate moderate stability. The scale correlates in the expected direction with measures of related constructs (Kendall et al., 1989; Kovacs, 1992), and normative data are available (Kovacs, 1992). State-Trait Anxiety Inventory for Children. The State-Trait Anxiety Inventory for Children has 20 items that assess how the child usually feels (A-Trait) and 20 assessing how the child currently feels (A-State). The A-Trait scale has adequate retest reliability (coefficients of 0.65 for males, 0.71 for females after 6 weeks). The stability of the correlation coefficients for the A-State scale was significantly lower (0.31 and 0.47), as expected given that it targets situational anxiety. The internal consistency of both scales is acceptable. Concurrent validity for the A-Trait scale is adequate, and normative data are available (Spielberger, 1973). Child Behavior Checklist. The 118-item Child Behavior Checklist (CBCL) assesses behavioral problems and social competencies. The mean retest reliability for the problems scale scores was 0.89 over 1 week and 0.75 over 1 year. The CBCL scale scores were favorably correlated with similar scales. Normative data are available (Achenbach and Edelbrock, 1991). Teacher’s Report Form. The 118-item Teacher’s Report Form (TRF), which is similar to the CBCL, provides information on the child’s classroom behavior. The mean retest reliability for the problems scale scores was 0.92, 0.75, and 0.66 after 2 weeks, 2 months, and 4 months, respectively. TRF scales are favorably correlated with similar scales. Normative data are available (Achenbach, 1991). Procedure Children and their parents were initially screened over the telephone. Eligible families were invited for the diagnostic intake. Parent self-report forms, the TRF, and a demographic information sheet were sent home to be completed and returned. The 3- to 4-hour intake consisted of separate diagnostic interviews with the child and the parents. Children completed self-report forms while parents were interviewed, and parents completed forms during the child interview. Children who met criteria were assigned randomly to the treatment or waitlist condition. Participants with ADHD taking medications at the start of their participation were not asked to discontinue medication. Treatment-condition children were randomly assigned to a therapist. Waitlist-condition children were assessed again after the 8-week waitlist and then were assigned randomly to a therapist. Treatment consisted of 16 to 20 weekly hour-long sessions. The first half of the cognitive-behavioral manual-based intervention (Kendall
J . A M . A C A D . C H I L D A D O L E S C . P S YC H I AT RY, 4 0 : 7 , J U LY 2 0 0 1
et al., 1990) focused on the recognition of anxious symptoms as cues and consisted of four steps: (1) recognizing anxious feelings and somatic reactions, (2) recognizing and modifying anxious cognition, (3) developing a plan to cope with the anxiety, and (4) evaluating performance and administering self-reward as appropriate. The second half of the program involved in vivo exposure to anxious situations. RESULTS
Although the results presented here are based on diagnoses derived from parent interviews, comparable results were found when the child interview data were analyzed. Nature and Characteristics of Pretreatment Comorbidity
The most frequent comorbid diagnoses for children with a primary anxiety disorder were simple phobia, social phobia, and GAD (Table 2). The overall rate and pattern of comorbidity was similar for children with each of the three primary anxiety disorders. To examine the role of comorbidity, children were divided into three groups according to parent-reported diagnoses: (1) those with only a primary anxiety disorder (ANX-only), (2) those with two or more anxiety disorders (ANX/ANX), and (3) those with comorbid externalizing disorders (ANX/EXT). There were only eight children with a comorbid diagnosis of major depression or dysthymia, a number too small to permit meaningful analysis of this group. These groups were compared with respect to age, gender, family income, TABLE 2 Distribution of Comorbid Diagnoses at Pretreatment Diagnosis GAD SAD SP SIM PTSD AG OCD PD SCH DEP ADHD ODD CD
%
n
29.0 17.0 33.5 46.2 0.0 0.0 1.2 1.7 20.8 4.6 15.0 9.2 1.2
50 29 58 80 0 0 2 3 36 8 26 16 2
Note: GAD = generalized anxiety disorder; SAD = separation anxiety disorder; SP = social phobia; SIM = simple phobia; PTSD = posttraumatic stress disorder; AG = agoraphobia; OCD = obsessivecompulsive disorder; PD = panic disorder; SCH = school phobia; DEP = major depression/dysthymia; ADHD = attention-deficit/ hyperactivity disorder; ODD = oppositional defiant disorder; CD = conduct disorder.
789
KENDALL ET AL.
parental education level, and race via analyses of variance (ANOVAs) or χ2 tests. Table 3 displays the demographic information grouped by diagnosis and were found not to differ (all p values > .05). However, χ2 analysis revealed a significant difference with regard to gender (χ2 [2, N = 173] = 8.2, p < .05). The ANX/EXT group contained more males, consistent with the greater male-to-female ratio expected in these disorders. Multivariate analysis of variance (MANOVA) indicated that the groups differed significantly with respect to pretreatment internalizing symptom severity as assessed by child self-report questionnaires (F2,171 = 0.88, p < .05). For parent report of internalizing symptom severity, ANOVAs indicated that the groups differed significantly (F2,172 = 5.1, p < .05). Scheffé post hoc analyses indicated that the ANX/ANX group and the ANX/EXT group displayed more severe internalizing symptoms than the ANX-only group, but did not differ from each other. For parent report of externalizing symptom severity, ANOVAs indicated that the groups differed significantly (F2,172 = 25.2, p < .05). Scheffé post hoc analyses indicated that the ANX/ EXT group displayed more severe externalizing symptoms than the ANX-only and ANX/ANX groups, but the latter two groups did not differ significantly from each other. According to teacher reports of internalizing symptoms, the groups did not differ significantly (p = .99). However, teacher reports of externalizing symptoms indicated that the groups differed significantly (F2,167 = 3.9, p < .05). Scheffé post hoc analyses indicated that the ANX/EXT group displayed more severe externalizing symptoms than the ANX-only group and the ANX/ANX group, but the latter two groups did not differ significantly from each other. Occasional missing forms resulted in missing data, but analyses revealed no pattern to the missing scores. Table 4 presents the means for the dependent measures.
Impact of Comorbidity on Treatment Outcome
Note: ANX = primary anxiety disorder only; ANX/ANX = primary anxiety disorder and comorbid anxiety; ANX/EXT = primary anxiety disorder and a comorbid externalizing disorder.
To examine whether pretreatment comorbidity moderated treatment response, a χ2 analysis was conducted with a categorical measure of treatment outcome (presence or absence of pretreatment primary diagnosis at posttreatment). These analyses indicated that individuals with comorbid diagnoses at pretreatment were not significantly less likely to respond to treatment as indicated by remission of the primary diagnosis (all χ2 values < 1). The results of these analyses are depicted in Figure 1. For example, 68.4% of patients with no comorbid diagnoses and 70.6% of patients with one or more comorbid diagnoses were free of their primary diagnosis at the conclusion of treatment. To examine the impact of comorbidity on treatment outcome as indicated by changes on dependent measures (child self-report measures, parent report measures, and teacher report measure), 3 (groups, between subjects) ⫻ 2 (assessment periods, within subjects) MANOVAs or ANOVAs were conducted. The MANOVA revealed significant differences for the assessment period main effect (F1,148 = 3.90, p < .05), indicating that treatment produced gains from pretreatment to posttreatment. However, there was no main effect for group (p = .19) and no interaction between group and time (p = .21), indicating that the presence of comorbidity was not associated with significant differential treatment effects. Analyses yielded significant main effects for time (preto posttreatment) on the parent measures of internalizing and externalizing symptom severity and the teacher measure of internalizing symptom severity (all p values < .05). No significant association was found between treatment and change in severity of teacher-reported externalizing symptoms. Comorbidity was not significantly associated with pre- to posttreatment changes in severity of parentreported internalizing symptoms (p = .57) or externalizing symptoms (p = .94) on the CBCL. Similarly, nonsignificant interactions were found between comorbidity and pre- to posttreatment changes in teacher-reported internalizing symptoms (p = .19) and externalizing symptoms (p = .22). Table 4 presents means for dependent measures from pretreatment to posttreatment and follow-up. To examine the relationship of continued comorbidity (comorbid diagnoses at posttreatment) to outcome, with absence of primary anxiety disorder diagnosis as the outcome measure, a χ2 analysis was conducted. This analysis indicated that patients with continued presence of comorbidity were less likely to show remission of
790
J . A M . A C A D . C H I L D A D O L E S C . P S YC H I AT RY, 4 0 : 7 , J U LY 2 0 0 1
TABLE 3 Demographic Characteristics of Diagnostic Groups Variable Gender (n) Male Female Age (M) Race (n) White African American
ANX
ANX/ANX
ANX/EXT
20 15 11.2
48 38 11.3
35 9 10.9
31 4
74 12
35 9
COMORBIDITY IN CHILDHOOD ANXIETY DISORDERS
TABLE 4 Group Means and Standard Deviations: Pretreatment to Follow-up Pretreatment Scale RCMAS M SD A-Trait M SD A-State M SD CDI M SD CBCL-I M SD CBCL-E M SD TRF-I M SD TRF-E M SD
Posttreatment
Follow-up
ANX
ANX/ANX
ANX/EXT
ANX
ANX/ANX
ANX/EXT
ANX
ANX/ANX
ANX/EXT
51.2 12.3
53.1 10.3
54.3 11.5
40.7 12.4
42.7 10.5
42.2 10.6
41.9 9.1
41.4 12.0
41.9 8.4
47.9 13.9
51.3 12.7
47.5 11.9
35.9 13.0
51.9 12.4
48.1 12.3
37.3 10.5
38.5 13.2
37.6 12.5
52.8 9.6
52.9 11.0
50.1 11.5
42.5 8.5
45.2 10.2
45.9 12.9
45.4 10.2
47.7 11.2
48.2 8.8
7.4 6.4
11.2 8.5
11.2 9.3
4.3 4.8
5.3 4.9
6.1 7.4
4.2 3.6
5.4 7.5
5.8 7.5
66.6 8.0
71.1 7.8
70.9 6.8
57.5 9.1
60.0 11.8
64.2 9.8
56.3 8.0
58.6 12.4
60.2 9.8
47.6 8.7
50.8 8.7
59.6 7.2
44.6 9.4
45.7 9.0
56.1 8.2
46.0 8.5
46.5 10.9
56.5 9.9
61.7 12.2
61.7 11.7
61.5 11.0
58.5 10.2
56.0 11.0
57.6 11.5
61.4 11.3
56.8 9.7
51.4 10.1
49.6 7.6
49.5 7.0
53.5 10.2
47.9 6.5
46.4 7.0
52.6 12.2
48.1 8.6
47.7 7.8
48.6 10.1
Note: ANX = Primary anxiety disorder only; ANX/ANX = primary anxiety disorder and comorbid anxiety; ANX/EXT = primary anxiety disorder and a comorbid externalizing disorder. RCMAS = Revised Children’s Manifest Anxiety Scale; A-Trait = State-Trait Anxiety Inventory for Children, Trait subscale; A-State = State-Trait Anxiety Inventory for Children, State subscale; CDI = Children’s Depression Inventory; CBCLI = Child Behavior Checklist, Internalizing scale; CBCL-E = Child Behavior Checklist, Externalizing scale; TRF-I = Teacher’s Report Form, Internalizing scale; TRF-E = Teacher’s Report Form, Externalizing scale.
J . A M . A C A D . C H I L D A D O L E S C . P S YC H I AT RY, 4 0 : 7 , J U LY 2 0 0 1
791
% of Patients Improved Fig. 1 Percentage of patients who improved. ANX = primary anxiety disorder only; ANX/ANX = primary anxiety disorder and comorbid anxiety; ANX/ EXT = primary anxiety disorder and a comorbid externalizing disorder.
their primary pretreatment diagnosis (χ2 [2, N = 129] = 7.2, p < .05). The impact of comorbidity on maintenance of treatment gains was examined with separate 3 (groups; between subjects) ⫻ 2 (assessment periods; posttreatment to follow-up, within subjects) MANOVAs. No significant change was found in level of symptoms (p = .07), although the continued decrease in self-reported symptoms approached significance. There was no effect of comorbidity on maintenance of treatment gains (p = .45), and no interaction effect between comorbidity and gains (p = .90). Because attrition between posttreatment and follow-up may have influenced these results, participants who completed the follow-up assessment were compared with participants who declined to complete the follow-up. These groups were compared with t tests or χ2 tests and found not to differ significantly with regard to demographic variables, pre/posttreatment diagnostic status, or pre/posttreatment dependent variables (e.g., self- and parent reports).
KENDALL ET AL.
The present results were consistent with published reports documenting a high rate of comorbidity in children with anxiety disorders (Brady and Kendall, 1992). The majority of children presenting for treatment with a primary anxiety disorder had at least one comorbid diagnosis (79%). The most common additional diagnoses were secondary anxiety disorders including GAD, simple phobia, social phobia, and separation anxiety disorder. Also
consistent with the literature, comorbidity was associated with greater severity of internalizing symptoms (Brown et al., 1995; Manassis and Menna, 1999; Nottelmann and Jensen, 1995). Pretreatment comorbidity was not associated with treatment outcome or maintenance of treatment gains at follow-up, defined broadly. However, those individuals who continued to have comorbid diagnoses at the end of the treatment were significantly less likely to show remission of their primary anxiety diagnosis (see also Borkovec et al., 1995; Brown et al., 1995). The present treatment, designed to address the child’s primary anxiety disorder, was nevertheless associated with a reduction in comorbid diagnoses. The literature suggests that comorbidity is associated with greater severity of symptoms, more persistent difficulties, and a greatly increased likelihood of seeking treatment (Manassis and Menna, 1999; McGee et al., 1990; Nottelmann and Jensen, 1995). The finding of increased severity of symptoms in comorbid groups in this sample is consistent with published reports, yet it raises issues regarding nosology. Is comorbidity within the anxiety disorders simply a proxy for general severity rather than an indication of an overlap of different types of psychopathology? Although the present study does not directly test this notion, the fact that children with multiple disorders exhibited greater levels of internalizing symptoms is consistent with the notion of proxy. The findings that comorbidity did not moderate response to treatment and that the treatment yielded a reduction in multiple anxiety diagnoses also seem to support the notion of a general anxiety syndrome. Brown and Barlow (1992) and Borkovec et al. (1995) articulated three reasons why treatment of a primary anxiety disorder might contribute to the reduction of comorbid diagnoses. First, the patient may generalize treatment skills and apply them to other areas not specifically targeted. Second, many disorders, particularly anxiety disorders, share overlapping features and symptoms, and improvement in some of these overlapping symptoms may account for reduced comorbidity after treatment. Third, there may be basic underlying psychological processes common to anxiety disorders even if secondary differentiating processes and symptoms are present. In this study, presence of disorders other than comorbid anxiety disorders did not moderate response to treatment, although this may have been due to the large treatment effect. Rather, there was a reduction in the rate of ADHD
792
J . A M . A C A D . C H I L D A D O L E S C . P S YC H I AT RY, 4 0 : 7 , J U LY 2 0 0 1
Impact of Treatment on Comorbid Conditions
To evaluate whether treatment for a primary anxiety disorder resulted in changes in comorbid diagnoses, a McNemar test was conducted to determine whether the overall rate of comorbidity changed significantly from pretreatment to posttreatment and follow-up. At pretreatment, 79% of the children had at least one additional diagnosis; at posttreatment, this rate had decreased to 47%. At follow-up, only 20% of the children continued to evidence comorbid diagnoses. These reductions in the rate of comorbidity were significant (McNemar test, p values < .05). Additional McNemar tests were conducted to determine whether anxiety treatment was associated with the reduction in specific comorbid diagnoses apart from anxiety diagnoses. At pretreatment, 15% of the children had comorbid ADHD. At posttreatment, this rate was 4.7%, and at follow-up it was 3.6%. These results were significant (McNemar test, p < .05). Similarly, there was a significant reduction in the rate of oppositional defiant disorder (ODD) from pretreatment (9.2%) to follow-up (1.8%), and posttreatment to follow-up, but not from pretreatment to posttreatment (6.8%) (McNemar test, p < .05). Because at least one study in the adult literature (Borkovec et al., 1995) indicated that the effectiveness of treatment for the primary condition affects reduction in comorbidity, differential decrease in comorbidity in treatment responders was evaluated. Participants were categorized into either a treatment success group (no longer met criteria for the pretreatment primary anxiety disorder diagnosis) or a treatment nonsuccess group. Chi-square analyses examined the rate of comorbidity in treatment responders versus treatment nonresponders at posttreatment and at follow-up. Treatment responders were significantly more likely to show a reduction in comorbidity at posttreatment (χ2 [1, N = 129] = 7.2, p < .05) and at follow-up (χ2 [1, N = 126] = 4.0, p < .05). DISCUSSION
COMORBIDITY IN CHILDHOOD ANXIETY DISORDERS
and ODD as well as in the overall rate of comorbidity. These results suggest that the treatment program may have robust effects, even on disorders that are not directly targeted for intervention and that are conceptualized as distinct from anxiety disorders. Reduction in comorbid anxiety diagnoses can be explained in terms of the three hypotheses listed above, as well as the capacity of the therapists to flexibly apply the manual-based treatment to address secondary anxiety-based concerns. However, it is more difficult to account for reduction in disorders that are not theoretically related to anxiety or the targets of an anxiety treatment program. For example, although ADHD and anxiety disorders do share some overlapping features, it is difficult to account for the decrease in diagnoses of ADHD on this basis alone. In some cases, a change in overlapping targeted symptoms may have resulted in a child’s not meeting criteria for ADHD (e.g., relaxation training helping a child to reduce excess fidgeting, a symptom of both anxiety and ADHD). An alternative explanation for the reduction in the rate of ADHD and ODD is that these disorders were causally related to the anxiety disorder or were exacerbated by anxiety. The present results indicate that children who evidenced a clinically significant change in anxiety problems were significantly more likely to show remission of comorbid diagnoses at posttreatment and follow-up. This finding could be interpreted as support for the argument that anxiety was exacerbating symptoms of ADHD or ODD. For example, anxiety might interfere with task-focused attention, and an anxious child’s behavior may appear similar to that of a child with ADHD under conditions of anxious arousal. Furthermore, there is some evidence that the presence of an anxiety disorder moderates the response to medications used to treat ADHD (Tannock et al., 1995), although other research suggests that this effect is not present when ADHD medication is properly titrated (Diamond et al., 1999; MTA Cooperative Group, 1999). Although high rates of comorbid depression are often reported in clinic-based samples of children with anxiety disorders (e.g., Bernstein, 1991; Manassis and Menna, 1999), the rate of comorbid depressive disorders in this sample was low (only 4.6%). This rate, however, is consistent with the rate of comorbid depressive disorders reported from a clinic sample in Australia (Barrett et al., 1996) with a similar age range (7–14 years). One reason for this lower rate of depressive disorders may be the restricted age range. Children with comorbid depression often are older than children with only anxiety disorders,
there is evidence suggesting a developmental progression (Kovacs et al., 1989), and the diagnosis of depression increases significantly after puberty, particularly in girls (e.g., Bird et al., 1993).
J . A M . A C A D . C H I L D A D O L E S C . P S YC H I AT RY, 4 0 : 7 , J U LY 2 0 0 1
793
Limitations
In general, comorbidity was defined in terms of broad categories (i.e., ANX/ANX, ANX/EXT), not specific disorders. Although this was necessary because of the numerous specific comorbid patterns, our results do not entirely address specific comorbid conditions. For example, as previously mentioned, the number of children with comorbid depression and anxiety was not large enough to permit meaningful examination of this subtype of comorbidity. This study relied on parent interviews for determinations of comorbidity. Changes in comorbid conditions may reflect actual child improvements or, because they are based on parent interviews, may also reflect parents’ lack of specificity in reporting changes. In addition, although diagnoses derived from parent and child reports were comparable, children in this study generally reported fewer disorders than did parents and tended to rate their symptoms as less severe on self-report measures than did their parents or teachers. It is unclear whether this phenomenon is a function of child underreporting or a reflection of limitations that children’s level of cognitive development places on their capacity to report on their own behavior and symptoms. Clinical Implications
The present results support the efficacy and robustness of cognitive-behavioral treatment for anxiety disorders in children with and without comorbidity. This finding is important because, frequently, children who present for treatment have comorbid disorders. This study also indicates that comorbid diagnoses were alleviated by anxietyfocused treatment, even though this treatment did not specifically target comorbid diagnoses. Although these findings are promising, further research is needed to examine the relationship of specific comorbid patterns on treatment outcome. REFERENCES Achenbach TM (1991), Integrative Guide for the 1991 CBCL/4–18, YSR, and TRF. Burlington: University of Vermont Department of Psychiatry Achenbach TM (1995), Diagnosis, assessment, and comorbidity in psychosocial treatment research. J Abnorm Child Psychol 23:45–65 Achenbach TM, Edelbrock C (1991), Manual for the CBCL and 1991 Profile. Burlington: University of Vermont Department of Psychiatry Albano AM, Chorpita BF (1995), Treatment of anxiety disorders of childhood. Psychiatr Clin North Am 18:767–784
KENDALL ET AL.
Barlow DH (1994), Comorbidity in social phobia: implications for cognitivebehavioral treatment. Bull Menninger Clin 58:43–57 Barlow DH, Craske MG, Cerny JA, Klosko JS (1989), Behavioral treatment of panic disorder. Behav Ther 20:261–282 Barlow DH, Rapee RM, Brown TA (1992), Behavioral treatment of generalized anxiety disorder. Behav Ther 23:551–570 Barrett PM (1998), Evaluation of cognitive-behavioral group treatments for childhood anxiety disorders. J Clin Child Psychol 27:459–468 Barrett PM, Dadds MR, Rapee RM (1996), Family treatment of childhood anxiety: a controlled trial. J Consult Clin Psychol 64:333–342 Basoglu M, Lax T, Kasvikis Y, Marks IM (1988), Predictors of improvement in obsessive-compulsive disorder. J Anxiety Disord 2:299–317 Bernstein GA (1991), Comorbidity and severity of anxiety and depressive disorders in a clinic sample. J Am Acad Child Adolesc Psychiatry 30:43–50 Bird HR, Gould MS, Staghezza BM (1993), Patterns of diagnostic comorbidity in a community sample of children aged 9 through 16 years. J Am Acad Child Adolesc Psychiatry 32:361–368 Borkovec TD, Abel JL, Newman H (1995), Effects of psychotherapy on comorbid conditions in generalized anxiety disorder. J Consult Clin Psychol 63:479–483 Brady EU, Kendall PC (1992), Comorbidity of anxiety and depression in children and adolescents. Psychol Bull 111:244–255 Brown TA, Antony MM, Barlow DH (1995), Diagnostic comorbidity in panic disorder: effect on treatment outcome and course of comorbid diagnoses following treatment. J Consult Clin Psychol 63:408–418 Brown TA, Barlow DH (1992), Comorbidity among anxiety disorders: implications for treatment and DSM-IV. J Consult Clin Psychol 60:835–840 Caron C, Rutter M (1991), Comorbidity in child psychopathology: concepts, issues, and research strategies. J Child Psychol Psychiatry 32:1063–1080 Chambless DL, Renneberg B, Goldstein A, Gracely EJ (1992), MCMI-diagnosed personality disorders among agoraphobic outpatients: prevalence and relationship to severity and treatment outcome. J Anxiety Disord 6:193–211 Deas-Nesmith D, Brady KT, Campbell S (1998), Comorbid substance use and anxiety disorders in adolescents. J Psychopathol Behav Assess 20:139–148 Diamond IR, Tannock R, Schachar RJ (1999), Response to methylphenidate in children with ADHD and comorbid anxiety. J Am Acad Child Adolesc Psychiatry 38:402–409 Francis G, Beidel D (1995), Cognitive-behavioral psychotherapy. In: Anxiety Disorders in Children and Adolescents, March JS, ed. New York: Guilford, pp 321–340 Heimberg RG, Salzman DG, Holt CS, Blendell KA (1993), Cognitivebehavioral group treatment for social phobia: effectiveness at five-year follow-up. Cognit Ther Res 17:325–339 Kearney CA, Silverman WK (1998), A critical review of pharmacotherapy for youth with anxiety disorders: things are not as they seem. J Anxiety Disord 12:83–102 Kendall PC (1994), Treating anxiety disorders in youth: results of a randomized clinical trial. J Consult Clin Psychol 62:100–110 Kendall PC, Cantwell DP, Kazdin AE (1989), Depression in children and adolescents: assessment issues and recommendations. Cognit Ther Res 13:109–146 Kendall PC, Clarkin JF (1992), Introduction to special section: comorbidity and treatment implications. J Consult Clin Psychol 60:833–835 Kendall PC, Flannery-Schroeder E, Panichelli-Mindel SM, Southam-Gerow M, Henin A, Warman M (1997), Therapy for youths with anxiety disorders: a second randomized clinical trial. J Consult Clin Psychol 65:366–380 Kendall PC, Kane M, Howard B, Siqueland L (1990), Cognitive-Behavioral Therapy for Anxious Children: Treatment Manual. Ardmore, PA: Workbook Publishing Kendall PC, Kortlander E, Chansky TE, Brady EU (1992), Comorbidity of anxiety and depression in youth: treatment implications. J Consult Clin Psychol 60:869–880
Kendall PC, Southam-Gerow MA (1996), Long-term follow-up of a cognitivebehavioral therapy for anxiety-disordered youth. J Consult Clin Psychol 64:724–730 Kendall PC, Warman M (1996), Anxiety disorders in youth: diagnostic consistency across DSM-III-R and DSM-IV. J Anxiety Disord 10:453–463 Kessler RC, McGonagle KA, Zhao S et al. (1994), Lifetime and 12-month prevalence of DSM-III-R psychiatric disorders in the United States: results from the National Comorbidity Study. Arch Gen Psychiatry 51:8–19 Kovacs M (1992), Children’s Depression Inventory: Manual. Los Angeles: Western Psychological Services Kovacs M, Gatsonis C, Paulauskas SL, Richards C (1989), Depressive disorders in childhood, IV: a longitudinal study of comorbidity with and risk for anxiety disorders. Arch Gen Psychiatry 46:776–782 Manassis K, Menna R (1999), Depression in anxious children: possible factors in comorbidity. Depress Anxiety 10:18–24 March JS, Swanson JM, Arnold LE et al. (2000), Anxiety as a predictor and outcome variable in the Multimodal Treatment Study of Children with ADHD (MTA). J Abnorm Child Psychol 28:527–541 Masi G, Mucci M, Favilla L, Romano R, Poli P (1999), Symptomatology and comorbidity of generalized anxiety disorder in children and adolescents. Compr Psychiatry 40:210–215 McGee R, Feehan M, Williams S, Partridge F, Silva PA, Kelly J (1990), DSMIII disorders in a large sample of adolescents. J Am Acad Child Adolesc Psychiatry 29:611–619 Mendlowitz SL, Manassis K, Bradley S, Scapillato D, Miezitis S, Shaw BF (1999), Cognitive-behavioral group treatments in childhood anxiety disorders: the role of parental involvement. J Am Acad Child Adolesc Psychiatry 38:1223–1229 Merikangas KR, Stevens DE, Fenton B et al. (1998), Co-morbidity and familial aggregation of alcoholism and anxiety disorders. Psychol Med 28:733–788 MTA Cooperative Group (1999), Moderators and mediators of treatment response for children with attention deficit/hyperactivity disorder. Arch Gen Psychiatry 56:1088–1096 Nottelmann ED, Jensen PS (1995), Comorbidity of disorders in children and adolescents: developmental perspectives. Adv Clin Child Psychol 17:109–155 Noyes R, Reich J, Christiansen J, Suelzer M, Pfohl B, Coryell WA (1990), Outcome of panic disorder: relationship to diagnostic subtypes and comorbidity. Arch Gen Psychiatry 47:809–818 Reynolds CR, Richmond BO (1978), What I Think and Feel: a revised measure of children’s manifest anxiety. J Abnorm Child Psychol 6:271–280 Reynolds CR, Richmond BO (1985), Revised Children’s Manifest Anxiety Scale (RCMAS): Manual. Los Angeles: Western Psychological Services Reynolds WM, Coats KI (1986), A comparison of cognitive-behavioral therapy and relaxation training for the treatment of depression in adolescents. J Consult Clin Psychol 54:653–660 Silverman W (1991), Anxiety Disorders Interview Schedule for Children. Albany, NY: Graywind Silverman W, Albano AM (1997), The Anxiety Disorders Interview Schedule for Children (DSM-IV). San Antonio, TX: Psychological Corporation Silverman WK, Nelles WB (1988), The Anxiety Disorders Interview Schedule for Children. J Am Acad Child Adolesc Psychiatry 27:772–778 Spielberger C (1973), Manual for the State-Trait Anxiety Inventory for Children. Palo Alto, CA: Consulting Psychologists Press Stark KD, Reynolds WM, Kaslow KJ (1987), A comparison of the relative efficacy of self-control therapy and a behavioral problem-solving therapy for depression in children. J Abnorm Child Psychol 15:91–113 Tannock R, Ickowicz A, Schachar R (1995), Differential effects of methylphenidate on working memory in ADHD children with and without comorbidity. J Am Acad Child Adolesc Psychiatry 34:886–896 Turner RM (1987), The effects of personality disorder diagnosis on the outcome of social anxiety symptom reduction. J Personal Disord 1:136–143
794
J . A M . A C A D . C H I L D A D O L E S C . P S YC H I AT RY, 4 0 : 7 , J U LY 2 0 0 1