Participants

The study included a total of 284 youth (see Table 1 for sample characteristics). Eighty-four were part of one sample comprising 55 participants with an anxiety disorder and 29 healthy volunteers (i.e., anxiety sample). The remaining 200 youth were part of another sample from the NIH Characterization and Treatment of Depression Study [23], comprising 127 participants with Major Depressive Disorder (MDD) and 73 healthy volunteers (i.e., depression sample).

The 84 participants in the anxiety sample were interviewed by trained clinicians masked to all other data using the Kiddie Schedule for Affective Disorders and Schizophrenia for School-Aged Children – Present and Lifetime Version (K-SADS-PL) [24]. Fifty-five of these participants met the criteria for one or more of the following anxiety disorders: Generalized Anxiety Disorder (GAD), Separation Anxiety Disorder (Sep AD), or Social Anxiety Disorder (Social AD) (see Table 1 for sample haracteristics). The 29 healthy volunteers within the anxiety sample were free from any psychiatric diagnoses. Exclusion criteria for the anxiety sample can be found at https://clinicaltrials.gov/study/NCT00018057 (diagnostic criteria listed in supplement).

In the depression sample, participants were interviewed by trained clinicians using the K-SADS-PL [24] who were not masked to scores on the DAWBA. Depressed participants met criteria for Major Depressive Disorder (MDD) while healthy volunteers were free from any psychiatric diagnoses (see Table 1 for sample characteristics). Exclusion criteria for the depression sample can be found at https://www.clinicaltrials.gov/study/NCT03388606 (diagnostic criteria listed in supplement).

MeasuresDevelopment and well-being assessment (DAWBA)

The Development and Well-Being Assessment (DAWBA) is a package of questionnaires, interviews, and rating techniques designed to generate Diagnostic and Statistical Manual of Mental Disorders (DSM-5) and International Classification of Diseases (ICD-10) psychiatric diagnoses for youth ages 11–17. The DAWBA was collected using an online platform developed by Youth in Mind [14], with all participants completing the self-report and their guardian completing the parent-report. For all participants, the online DAWBA assessment began with the Strengths and Difficulties Questionnaire (SDQ), a brief screening evaluation consisting of twenty-five questions that span five scales of behavioral difficulty: emotional symptoms (5 questions), conduct problems (5 questions), hyperactivity/inattention (5 questions), peer relationship problems (5 questions), and prosocial behavior (5 questions). Following the SDQ, participants were presented screening questions for each of 19 psychiatric disorder modules. For depression and anxiety modules, the DAWBA queries about symptoms ranging from the prior four weeks to six months. If the participant responds affirmatively to either the disorder-related SDQ question or one of the disorder-specific screening questions, meeting the set symptom threshold, the DAWBA will then prompt the individual to answer further explicating questions. However, if the participant does not meet the set symptom threshold, the DAWBA will bypass the remaining items for that disorder module.

DAWBA Band Scores. For all participants completing the DAWBA, three band scores are generated for each disorder module, representing the parent, child, and combined parent-child predictions of risk for a disorder. Band scores are integer values ranging from 0 to 5 indicating the following probabilities of a child meeting diagnostic criteria: “0” (p < 0.1%), “1” (p ~ 0.5%), “2” (p ~ 3%), “3” (p = 15%), “4” (p = 50%), and “5” (p ≥ 70%). For all analyses presented in the results section, we used the DAWBA combined (parent and child) band scores based on DSM-5 criteria, which is generated by the DAWBA’s computer algorithms [11]. Previous reports have also derived DAWBA symptom scales, which combine data from various disorders and offer a continuous measure of symptom experience [25]; see the supplement for analyses using symptom scales.

Screen for child anxiety related emotional disorders (SCARED)

The SCARED is a widely used child- and parent-reported instrument to assess childhood anxiety disorders including GAD, Sep AD, Panic Disorder, and Social AD [8]. Both the parent and child questionnaires comprise 41 items spanning five DSM-4 domains. The SCARED possesses moderate parent-child agreement and good internal consistency (intraclass correlation coefficients = 0.37–0.62; α = 0.7–0.9), discriminant validity, and test-retest reliability (ρ = 0.6–0.9) [26]. The SCARED generates a composite score, one from the parent assessment and one from the child assessment; in our analysis, we averaged these two scores to calculate a combined parent- and child-reported SCARED value (SCARED-CP) [27] to allow for comparison with the DAWBA’s combined parent-child band scores. Greater SCARED scores indicate more severe anxiety symptom presentation. Participants who completed the SCARED more than three months from the completion of the DAWBA were excluded from all analyses using SCARED data. The three-month cutoff allowed for a sufficient sample size across various measures while also excluding overt outliers. In an exploratory analysis, we also tracked anxious participants’ SCARED scores across treatment, using their baseline score collected at screening along with successive scores collected at predefined intervals during cognitive behavioral therapy (CBT): pre-exposure (week 3 of treatment), mid-exposure (week 8 of treatment), and post-exposure (week 12 of treatment).

The mood and feelings questionnaire (MFQ)

The MFQ provides an assessment of depression symptom severity and generates a composite score from both child and parent reports [9]. As with the SCARED, these composite scores were averaged together in our analyses to provide a combined parent and child MFQ score (MFQ-CP) [28] in order to assess the relationship with the DAWBA’s combined parent-child band score. Within the MDD sample, the MFQ was collected from both depressed and healthy volunteers at baseline when participants were enrolled in the study; this baseline score was used in all analyses. Each of the 13 items was scored and summed to generate a composite score ranging from 0 to 26, with greater scores indicating more severe depression or mood disorder presentation. For participants who completed the MFQ more than three months before or after the completion of the DAWBA, their data were omitted from the respective analyses.

Pediatric anxiety rating scale (PARS)

The PARS is a clinician-rated assessment of anxiety symptoms and is widely used in treatment studies [10, 17]. When administered, both child and parent responses are considered in the clinician’s assessment, resulting in a combined overall score. A higher score indicates more severe anxiety symptom presentation. PARS was collected at the screening visit and at three time points during treatment: CBT week 3, CBT week 8, and CBT week 12. Youth who completed the DAWBA and PARS within three months of each other were included in analyses of PARS data.

Procedure

Written informed consent and assent were obtained from the guardian(s) and the child, respectively. All procedures were approved by the NIMH Institutional Review Board, and all participants were compensated for participation. All data presented in this study were collected from children and their guardian(s) as part of larger treatment studies of children with mood or anxiety disorders. DAWBA was administered at screening in both groups. However, as noted above, in the anxiety sample, clinicians were masked to the results of the DAWBA when assigning a diagnosis via the K-SADS-PL; in contrast, clinicians performing the K-SADS-PL in the depression sample were not masked to the DAWBA.

The relationship between the DAWBA and established self-report instruments was evaluated within the two samples. In the anxiety cohort, both SCARED-CP and PARS were administered at the screening visit and across treatment at the intervals defined above. In the depression sample, the MFQ-CP was administered at screening. All questionnaire data collected more than three months from the DAWBA were excluded in both the anxiety and depression samples. In both samples, if participants completed repeated symptom measures within three months of the DAWBA, we used data temporally closest to the collection of the DAWBA.

Statistical analysis

All statistical analyses were performed in RStudio (version 2022.07.1).

1a and 1b – self-report measures: SCARED-CP and MFQ-CP

A linear regression analysis was performed to assess whether the three DAWBA band scores for anxiety disorders (GAD, Sep AD, and Social AD) predicted the SCARED-CP. Similarly, a linear regression was performed in the depression sample between the DAWBA MDD band score and MFQ-CP.

1c – SCARED-CP across treatment

To assess whether the DAWBA predicted treatment outcomes, we calculated the difference score for the SCARED-CP by subtracting the pre-treatment (baseline) score from the post-treatment (CBT week 8 or CBT week 12) score. A linear regression analysis was performed using the DAWBA band scores to predict the SCARED-CP difference score. In calculating the difference score, we used the latest SCARED-CP score a subject completed, either from week 8 or 12 of CBT.

2a – clinician-report measure: PARS

A comparable approach as outlined in Statistical Analyses 1a was used for PARS in which linear regression was used to predict the PARS score from the DAWBA anxiety band scores (GAD, Sep AD, and Social AD) in the anxiety sample.

2b – PARS across treatment

A comparable approach as outlined in Statistical Analysis 1c was used for PARS, in which we performed a linear regression analysis to predict the PARS difference score (post-treatment - pre-treatment) from the DAWBA band scores (GAD, Sep AD, and Social AD). In calculating the difference score, we used the latest PARS a subject completed, either from week 8 or 12 of CBT.

Table 1 Demographic characteristics of sample