Category: Child / Adolescent - Anxiety

PS11- #B44 - What Can Anxiety Tell Us? Classifying Depressed and Nondepressed Youth Using a Measure of Anxiety

Saturday, Nov 18
12:15 PM – 1:15 PM
Location: Indigo Ballroom CDGH, Level 2, Indigo Level

Keywords: Child Anxiety | Child Depression


It is well known that, by early adolescence, depressive disorders are highly comorbid with anxiety disorders. Furthermore, several studies have documented the utility of anxiety- and depression- specific measures for uniquely assessing these domains. When considering youth self-report measures, there is a high degree of overlap in individual anxiety and depressive symptoms and items used to assess such symptoms. Furthermore, in younger children, depression may not be adequately captured by depression-specific self-assessment tools because of lower symptom thresholds overall and unique behavioral indicators of emotional distress in younger versus older children, among other factors. Therefore, the purpose of the current study was to examine whether a measure of youth anxiety alone, the Screen for Child Anxiety and Related Emotional Disorders (SCARED), could classify potentially depressed and non-depressed youth.

Discriminant function analysis (DFA) was used with a clinical sample of youth (n = 289, 51.9% male, age M=11.03 years) presenting for an initial evaluation at a university-based specialty clinic for child and adolescent emotional disorders. The best-fitting model included parent- and child-rated subscales of the SCARED, as well as age, and clinician-rated Global Severity. This model correctly classified 67.7% of the original cases into three categories (depression, subclinical depression, and no depression diagnosed based on a structured clinical interview (Wilk’s λ=.63, χ2(24)=73.75, p < .01). Several child-rated subscales were shown to contribute more to the model than their corresponding parent-rated subscales, suggesting that child-report of anxiety symptoms may be particularly informative for predicting clinical impressions of depressive disorders than parent report.  However, age contributed the most to the model, indicating that older age may be a stronger predictor of depression status than the SCARED. These findings suggest that parent- and child-rated anxiety symptoms may be informative as a risk screening for the presence of co-occurring depression, particularly in clinical samples and/or when depression-specific measures are unavailable.

Rebecca Grossman

Graduate Student
University of Miami
Miami, Florida

Jill Ehrenreich-May

Associate Professor
University of Miami
Coral Gables, Florida