|Home | About | Journals | Submit | Contact Us | Français|
Reports the characteristics of a large, representative sample of treatment seeking anxious youth (N =488). Participants, aged 7–17 years (mean 10.7 yrs), had a principal DSM-IV diagnosis of separation anxiety disorder (SAD), generalized anxiety disorder (GAD), or social phobia (SP). Although youth with a co-primary diagnosis for which a different disorder-specific treatment would be indicated (e.g., major depressive disorder, substance abuse) were not included, there were few other exclusion criteria. Participants and their parent/guardian underwent an extensive baseline assessment using a broad array of measures capturing diagnostic status, anxiety symptoms and severity, and areas of functional impairment. Means and standard deviations of the measures of psychopathology and data on diagnostic status are provided. The sample had moderate to severe anxiety disorder and was highly comorbid, with 55.3% of participants meeting criteria for at least one non-targeted DSM-IV disorder. Anxiety disorders in youth often do not present as a single/focused disorder: such disorders in youth overlap in symptoms and are highly comorbid among themselves.
Anxiety disorders are among the most common conditions affecting youth (Costello, Egger, & Angold, 2005). Epidemiological reports based on large and/or nationally representative samples estimate the prevalence of anxiety disorders in youth to range between 10% and 20% (Achenbach, et al., 1995; Shaffer, Fisher, Dulcan, & Davies, 1996). In general, anxiety disorders are often comorbid among themselves and with other disorders. If left untreated, pediatric anxiety disorders predict adult anxiety disorders and depression; and other childhood sequelae, such as substance use problems, suicide attempts, and hospitalization (Ferdinand & Verhulst, 1995; Kendall, Safford, Flannery-Schroeder, & Webb, 2004; Pine, Cohen, Gurley, Brook, & Ma, 1998). Data from retrospective studies suggest that more than half of adults with anxiety or mood disorders had a history of a childhood anxiety disorder (Kessler et al., 1994).
Previous attempts to characterize anxious youth have included epidemiological studies, as well as data from psychotherapy and medication trials. Youth seeking treatment in a medication trial may differ from youth seeking treatment in a psychotherapy trial. For example, children seeking services in a medical setting might have been thought to be more severely disturbed, whereas youth seeking psychological services might have been thought to have less comorbidity. Also, different outcome measures and assessment approaches between psychotherapy and medication trials have made it difficult to compare samples and results across studies (Compton, Burns, Egger, & Robertson, 2002; Silverman, Pinna, & Viswesvaran, 2008). For example, psychotherapy outcomes have been assessed in terms of meeting or not meeting diagnostic criteria before and after treatment (e.g., Kendall et al. 1997; 2008), where medication trials (e.g., RUPP, 2002) have examined ratings of posttreatment improvement.
The Child/Adolescent Anxiety Multimodal Study (CAMS) was a randomized clinical trial investigating the relative efficacy of cognitive-behavior therapy (CBT), medication (sertraline; SRT), and their combination (COMB), as compared to pill placebo (PBO) for youth with a primary separation anxiety (SAD), social anxiety (SAD), or generalized (GAD) anxiety disorder (Walkup et al., 2008). CAMS included state-of-the-art measures of anxiety symptoms and severity, comorbid conditions and functional impairment that should allow generalization to treatment seeking youth in both psychotherapy and medication treatment studies. The present report describes the clinical characteristics of the 488 participants.
Enrollment sites represent a range of demographic and geographical locations and utilized a variety of recruitment strategies. Participants completed a three stage screening process that included an initial semi-scripted phone screen (Gate A), a diagnostic evaluation, physical exam, and pregnancy test (when appropriate) to ensure eligibility (Gate B), and if eligible, a baseline assessment to obtain pre-treatment parent and child ratings about the child’s anxiety and relevant comorbidity, and child and family functioning before beginning treatment (Gate C). Additional information about recruitment and enrollment is provided in Compton et al. (in press) and Walkup et al. (2008).
Participants who met DSM-IV diagnostic criteria for a principal diagnosis of SAD, GAD, or SP were eligible. With exception of major depressive disorder (MDD), bipolar disorder, pervasive developmental disorder, schizophrenia or schizoaffective disorder (all of which require a disorder-specific treatment not provided within CAMS), children who met diagnostic status for another DSM-IV disorder were also allowed to participate as long as the disorder was not principal or co-principal. Children who failed an adequate trial of CBT or an adequate trial of a psychotropic medication were also excluded. However, participants entering on a stable psychostimulant regimen for ADHD (operationalized as no significant change in medication dose for three months prior to study entry) were eligible.
The ADIS (child and parent versions) assesses anxiety, mood, and externalizing disorders in youth and screens for additional disorders. Composite impairment ratings (Clinician Severity Rating; CSR) were generated for each diagnosis and used to determine diagnoses as principal (most disabling). A CSR rating of 4 or higher was required to meet criteria for a DSM-IV diagnosis. Versions of the ADIS have been reported to have excellent inter-rater reliability (r = .98 for the parent interview and .93 for the child interview; Silverman & Nelles, 1988) and strong retest reliability for the parent interview (k = .76; Silverman & Eisen, 1992) The ADIS has excellent interrater reliability, retest reliability, and concurrent validity (Lyneham, Abbott, & Rapee, 2007; Silverman, Saavedra, & Pina, 2001; Wood et al., 2002), and has been shown to be sensitive to treatment change (Hudson et al., 2009; Kendall, 1994; Kendall et al., 1997). Whenever possible, children and parents were interviewed separately by the independent evaluators (IEs).
The CGI-S provided a global rating of anxiety severity ranging from 1 (not at all ill) to 7 (extremely ill). The IE provided baseline CGI-S ratings for each participant following the Gate C assessment.
The PARS is a clinician-rated 50-item anxiety symptom checklist that is used to rate 6 anxiety severity rating subscales specifically addressing the combined severity of symptoms of SAD, GAD, and SP. The PARS has excellent inter-rater reliability (> 0.97), was used as the dimensional primary outcome measure in the CAMS trial, and provides for the current study an index of the general severity of anxiety.
The MASC is a 39-item, 4-point Likert, child self-report scale with four factors: physical symptoms, social anxiety, harm avoidance and separation/panic anxiety. MASC predictive and discriminative validity have been shown to be excellent (Baldwin, & Dadds, 2007; Rynn et al., 2006), with three-week retest reliability of .79 (March et al., 1997). A parent version of the MASC was created and used by parents to rate their impression of their child’s anxiety symptoms.
The SCARED is a 41-item child and parent self-report instrument assessing DSM-IV symptoms of panic disorder, SAD, GAD, SP, and school phobia. The SCARED has excellent psychometric properties in clinical (Birmaher et al., 1999), community (Hale, Raaijmakers, Muris, & Meeus, 2005), and primary care (Wren, Bridge, & Birmaher, 2004).
The PDS is a 5-item measure assessing stage of pubertal development. Validation data have been reported (Bond et al., 2006). A score of two differentiates prepubertal and pubertal youth.
The CGAS provides a measure of global impairment and functioning over the previous month. Scores range from 1 (lowest functioning) to 100 (highest functioning). The CGAS possesses acceptable psychometric properties (Green, Shirk, Hanze, & Wanstrath, 1994).
The CBCL is a 118-item parent report of child behavioral problems and social and academic competence. The CBCL has strong psychometric properties, and good interparent agreement and construct validity (Achenbach & Rescorla, 2001).
The NASSQ provides a child-report measure of specific self-statements associated with negative affect. Items are rated on a 5-point frequency scale. The NASSQ (Sood, & Kendall, 2007) consists of separate items for younger and older children and, for its initial and cross-validation samples, possesses strong internal consistency (.87; .94) and retest reliability (.96; .80; Ronan et al, 1994).
The CAIS-C and CAIS-P are parallel 27-item child- and parent-report ratings assessing anxiety-related difficulties in children’s school, social, and home/family functioning. The CAIS-P has been shown to possess good internal consistency, construct, and discriminant validity (Langley, Bergman, McCracken, & Piacentini, 2004).
The MFQ is a 33-item self-report measure of depressive symptoms. Derived from the DSM-III-R criteria for depression/dysthymia, the MFQ possesses strong psychometric properties including internal consistency (Angold, Costello, Messer, & Pickels, 1995; Kent, Vostanis, & Feehan, 1997). A parent version of the MFQ was created and used by parents to rate their impression of their child’s depressive symptoms.
In addition to parent, child, and clinician questionnaires, the primary outcome variables were assessed by blind independent evaluators (IEs). IEs were MA-level psychologists, social workers, a nurse practitioner, PhD psychologists and child psychiatrists, who were carefully selected based on experience and predetermined background criteria. IEs were trained to reliability both within and cross-site by a QA team through live workshops and telephone conferences, and were randomly checked (quality assurance) over the course of the study (Compton et al., in press; Walkup et al., 2008). Throughout the study, IEs participated in weekly on-site supervision, and bi-weekly cross-site supervisory conference calls.
Frequencies and percentages are provided for categorical variables and means and standard deviations (SDs) for continuous variables. Given the small amount of missing data (<0.25%) no imputation methods were used. Unless noted, all statistical summaries are based on the total randomized sample. Data analyses used SAS/STAT software, Version 9.1.3 for Windows.
The mean age of the sample was 10.70 years (SD=2.80; range=84–214 months), with the majority (74.2%) aged 7 to 12 years. The gender distribution of the sample was evenly distributed, with 246 boys (50.4%) and 242 girls (49.6%). The ethnic/racial distribution of the sample was 78.9% white (n = 385), 9.0% black (n = 44), 2.5% Asian (n = 12), 1.2% American Indian (n = 6), 0.4% Native Hawaiian/Other Pacific Islander (n = 2), and 8.0% other (n = 39). The sample was also predominately non-Hispanic (87.9%).
Of the 488 youth, 480 (98.36%) met all eligibility criteria, and 8 (1.64%) were granted an eligibility waiver (e.g., permitting entry when the child’s age was very close to but not within the target ages) by the CAMS Steering Committee. Although the mean age of boys and girls did not differ significantly, girls were further along in pubertal development (Mean (SD) PDS: boys=1.69 (0.81), girls=2.00 (1.00), t=3.82, p<0.0001).
Clinical characteristics of the sample are summarized in Table 1. The CAMS sample had similar PARS scores to those in prior trials (e.g., Pine and colleagues (Pine et al., 2001); M=19.2 vs. 18.7, respectively) and higher scores than found with nonanxious youth (Ginsburg, Keeton, & Riddle, 2007). Overall, 54.5% of parents reported that their child had one or more previous mental health contacts. Two hundred and thirty-four (47.9%) youngsters had received some form of psychotherapy in the past and 104 (29.7%) participants had taken a psychotropic medication prior to study entry. Psychiatric hospitalization was reported for 1.8% of participants.
Parents reported higher scores than their children on the anxiety rating scales (MASC: 64.5 vs. 45.9, t=16.81, p<0.0001; SCARED: 32.1 vs. 23.4, t=9.68, p<0.0001. In contrast, youth reported significantly higher MFQ scores than parents (17.8 vs. 13.8, t=5.69, p<0.0001). Though CAMS excluded youngsters meeting criteria for current MDD, the sample did evidence elevated depressive symptoms; mean youth and parent MFQ scores corresponded to the 75–80%ile based on prorated norms derived from Angold and colleagues (Angold, Erkanli, Silberg, Eaves, & Costello, 2002).
Data from extant RCTs (Table 2) provide a backdrop. Most medication trials targeted a single diagnosis, whereas psychotherapy trials typically targeted multiple anxiety disorders. The present sample appears similar to the published studies in racial/ethnic makeup and gender. Relative to existing trials, CAMS included a higher proportion of youth with SAD and SP, and was comparable with regard to GAD.
Table 3 provides the prevalence of the targeted anxiety disorders, alone and in combination with each other. 270 participants (55.33%) met criteria for one or more DSM-IV disorders in addition to one of the three principal anxiety disorders. Of these, 170 (34.8%) had one additional disorder, 79 (16.2%) had two, 14 (2.9%) had three, and 6 (1.2%) met criteria for four DSM-IV disorders. The most common comorbidities were ADHD, ODD, and OCD (Table 4).
The sample was sorted into the following diagnostic groups based on highest CSR score(s): (1) SAD only, (2) SP only, (3) GAD only, (4) SAD and SP, (5) SAD and GAD, (6) SP and GAD, and (7) SAD, SP, and GAD (Table 3).
Among participants who met diagnostic criteria for only one of the three targeted anxiety disorders (n=105), adolescents were significantly more likely to receive a principal diagnosis of SP than children (20.6% vs. 8.3%; chi-square=13.94; p<0.0002). No significant differences by age were found for SAD only or GAD only.
Among participants who met diagnostic criteria for two of the three anxiety disorders (n=207), children were more likely than adolescents to meet criteria for SAD plus either SP (8.7% vs. 0.8%, p<0.002) or GAD (10.3% vs. 1.6%, p<0.002). In contrast, significantly more adolescents (47.6%) than children (20.8%) met criteria for SP+GAD (p<0.0001). Children were significantly more likely than adolescents (41.7% vs. 20.6%) to meet diagnostic criteria for all three targeted disorders SAD+SP+GAD (p<0.0001).
Consistent with the finding that more adolescents than children met criteria for SP, ADIS CSRs for SP were significantly higher for adolescents (M=5.8, SD=1.1) than for children (M=5.2, SD=.98; F=21.79, p<0.0001). The CSR ratings for SAD and GAD did not differ significantly between children and adolescents. Total PARS score was significantly higher for adolescents (M=20.1, SD=4.1) than children (M=18.8, SD=4.2; F=8.28, p<0.004).
There were no statistically significant relationships between anxiety diagnostic status or severity and gender.
A higher percentage of white versus nonwhite participants received a principal diagnosis of GAD-only (8.3% vs. 1.0%, chi-square=6.96, p<0.008). Nonwhites were significantly more likely than whites (11.7% vs. 5.5%) to meet criteria for SAD plus SP (chi-square=4.91, p<0.03). No other statistically significant relationships were found between race and diagnostic status or anxiety severity as measured by CSR or PARS ratings.
This treatment-seeking sample had moderate to severe anxiety disorder(s). Given the relatively few exclusion criteria, wide age range, and multiple sites, the sample represents a real-world presentation of distressing anxiety among youth, particularly prepubertal youth. It is noteworthy that less than 25% of the sample met criteria for only one of the three targeted anxiety disorders, whereas 36% met criteria for all three disorders. These findings are consistent with other reports (e.g., Essau, Conradt, & Petermann, 1999; Essau, Petermann, 1999; Wittchen, Stein, & Kessler, 1999). In light of the high degree of anxiety comorbidity (Foa et al., 2005), it is of interest to consider the degree to which anxiety presents as separate disorders in youth. Given that youth are more likely than adults to lack a fully developed understanding of emotions, social expectations, and the parameters of risk, and given the rapid changes in biological, social, and familial systems occurring during development, it may not be surprising that anxious symptom presentation by youth does not fall lock-step into specified categories. Development, almost by definition (Lerner, 1986), suggests successive change, such that continuity of a specific anxiety disorder may not be expected. There may be a general problem with anxiety, yet differing presentations across development (see also Last, Perrin, Hersen, & Kazdin, 1992; Ollendick & King, 1994). With ongoing development in youth, an anxiety disorder may present differentially at different points in development (e.g., as separation when starting preschool and as social anxiety when entering adolescence). In contrast, although anxiety disorders in adults also evidence comorbidty (Craske, Rauch, Ursano, Prenoveau, Pine, & Zinbarg, 2009), their presentation may be less variable. Perhaps as evidenced by the finding that adolescents were more likely to meet criteria for social phobia than younger children, increased age and development is required for specific anxieties to take form and be expressed. Nevertheless, the anxiety disorders in youth overlap in symptoms and are highly comorbid among themselves.
Although youth with a co-primary diagnosis for which a different disorder-specific treatment would be indicated were not included, the present sample was highly comorbid: 55.3% met criteria for at least one additional disorder. These findings are consistent with other studies (Angold, Costello, & Erkanli, 1999; Lewinsohn, Zinbarg, Seeley, Lewisohn, & Sack, 1997; Verduin & Kendall, 2003) and further evidence that anxiety disorders in youth often do not present as single/focused disorder. These data also provide evidence that anxiety disorders often co-occur with externalizing problems, a finding that has treatment implications (March et al., 2000).
The severity of the sample was evident across multiple measures. Mean scores on the ADIS CSR, CGI-S and CGAS were all indicative of marked illness and functional impairment. On the PARS, the sample was comparable to youth in other studies (e.g., The RUPP Fluvoxamine Study (RUPP Anxiety Study Group, 2002)). The mean total score for the present sample on the CAIS-P, reflecting parental report anxiety-related functional interference, was significantly higher than that for the original CAIS-P validation sample (22.2 vs. 18.8, Hedges’ g=0.41, p<0.001). In addition, CBCL Anxiety/Depression, Internalizing, and total problem mean scores were all >1.5 SDs above the mean. Children’s mean self-talk (NASSQ) was in the negative and maladaptive range.
Consistent with prior literature (Comer & Kendall, 2004), parent-reports of their children’s symptoms on the MASC and SCARED were higher than children’s self-reports. Children may present themselves as less impaired for a number of reasons including social desirability, avoidance of their anxiety, because they are uncomfortable with the mental health system, or lack developmental awareness of their anxiety as pathological. Likewise, parents may over-report anxiety in their children for a variety of reasons, including their own personal distress and motivation to receive treatment (Connell & Goodman, 2002).
There were no meaningful differences in anxiety diagnosis and severity for gender or race. This finding is consistent with existing data (Costello et al., 1996; Ginsburg & Silverman, 1996; Treadwell, Flannery-Schroeder, & Kendall, 1995). Although the sample was primarily Caucasian, race was not related to diagnostic status except for a higher percentage of Caucasians with GAD only. Further study is needed, but the present findings suggest that, among treatment-seeking anxious youth, the anxiety disorder presentation is consistent across the races and gender. With regard to age, it was not surprising that, among participants who met diagnostic criteria for only one of the three anxiety disorders, a significantly higher percentage of adolescents were diagnosed with SP. This likely reflects the increasing concerns with social relationships that emerge during adolescence (Crawley, Beidas, Benjamin, Martin, & Kendall, 2008).
Despite the large sample, multisite recruiting, comprehensive assessment, and rigorous quality assurance, a few limitations warrant consideration. First, although the sample included the full range of SES, participants were predominately middle class or higher (75%) and European-American in background (78%). Second, study eligibility criteria precluded inclusion of youngsters with MDD, thus limiting our ability to examine the relationship between anxiety and depression. Finally, the sample does not include participants diagnosed with principal panic disorder, PTSD, or OCD of any severity.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.