Off-campus University of Miami users: To download campus access dissertations, please use the following link to log into our proxy server with your University of Miami CaneID and Password.

Non-University of Miami users: Please talk to your librarian about requesting this dissertation through interlibrary loan.

Publication Date



UM campus only

Embargo Period


Degree Type


Degree Name

Doctor of Philosophy (PHD)


Psychology (Arts and Sciences)

Date of Defense


First Committee Member

Jill Ehrenreich-May

Second Committee Member

Brian D. Doss

Third Committee Member

Heather A. Henderson

Fourth Committee Member

Kiara R. Timpano

Fifth Committee Member

Neena M. Malik


Response to treatment in youth anxiety and depression treatment is known to vary (Curry et al., 2006; Ginsburg et al., 2011). The current study used latent growth curve modeling (LGCM) to examine the separate trajectories of change in adolescent anxiety and depressive symptoms (1) over the course of a transdiagnostic CBT intervention; and (2) up to six months following the end of treatment. In addition, we tested baseline variations in negative affect (NA), positive affect (PA), and anxiety sensitivity (AS) as predictors of anxiety and depression symptom change during treatment and follow-up. Participants were 59 adolescents (57.6% female) ages 12-17 years old (M = 15.42, SD = 1.71) who completed at least eight sessions of the Unified Protocol for the Treatment of Emotional Disorders in Adolescence (UP-A; Ehrenreich et al., 2008). Subjects received the UP-A as participants in either an open trial of the UP-A (n = 15) or a randomized controlled trial (n = 44). All subjects were diagnosed with a primary anxiety and/or depressive disorder using composite diagnoses derived from child and parent responses on the Anxiety Disorders Interview Schedule for the DSM-IV, Child and Parent versions (ADIS-IV-C/P; Silverman & Albano, 1996). Self-rated and parent-rated adolescent anxiety and depressive symptoms were measured with the Revised Children’s Anxiety and Depression Scale (RCADS; Chorpita, Yim, Moffitt, Umemoto, & Francis, 2000) and Revised Children’s Anxiety and Depression Scale-Parent version (RCADS-P; Ebesutani, Bernstein, Nakamura, Chorpita, & Weisz, 2010), respectively. Adolescent anxiety and depressive symptoms were assessed at five time points: baseline (Time 1), eight weeks following the start of treatment (Time 2), the end of treatment (Time 3), three months following the end of treatment (Time 4), and six months following the end of treatment (Time 5). Separate unconditional, piecewise LGCMs were conducted for anxiety and depressive symptom change, and for self-rated and parent-rated symptoms. Results from self-report models indicated that anxiety and depressive symptoms showed similar rates of decline during the UP-A. However, whereas self-rated anxiety symptoms continued to show significant reductions during follow-up, depressive symptoms remained at levels similar to the end of treatment. Both models showed significant slope variances, indicating significant inter-individual variation in the rate of change of anxiety and depressive symptoms across time. Parent-rated models displayed poor fit to the piecewise LGCMs, due to little change in symptoms following the end of treatment. Unconditional models with a single slope factor displayed better fit, with similar rates of change between parent-rated anxiety and depressive symptoms. We next examined conditional LGCMs by regressing the intercept and slope factors onto baseline levels of self-rated NA, PA, and AS. NA and PA were measured with the Negative Affect and Positive Affect subscales of the Positive and Negative Affect Schedule (PANAS; Watson, Clark, & Tellegen, 1988), respectively. AS was measured with the Child Anxiety Sensitivity Index (CASI; Silverman, Fleisig, Rabian, & Peterson, 1991). The PANAS and CASI were completed at the baseline assessment. Results for baseline associations were generally consistent with hypotheses. Higher levels of NA were associated with greater severity of adolescent self-rated anxiety and depressive symptoms. Lower levels of PA predicted greater depressive symptom severity at baseline, whereas PA was non-significantly associated with anxiety symptom severity. Higher levels of AS were associated with more severe anxiety symptoms, but were unrelated to depressive symptoms. Similar to adolescent self-report, NA was associated with more severe parent-rated anxiety and depressive symptom severity at baseline. However, in contrast to adolescent ratings, PA was associated with both anxiety and depressive symptom severity at baseline by parent-report. AS was unrelated to parent-rated baseline anxiety or depression symptom severity. Controlling for age, gender, and baseline symptom severity, there was little significant prediction of the slope factors. However, there were two exceptions. First, gender was a significant predictor of the treatment slope for adolescent self-rated anxiety symptoms, with girls showing a faster rate of change than boys. Second, and in the unexpected direction, higher levels of AS at baseline predicted greater declines in adolescent self-rated depressive symptoms during the course of the intervention. Results suggest that the UP-A may target both anxiety and depressive symptoms effectively, given similar rates of changes during the intervention. However, whereas anxiety symptoms displayed continued improvement following treatment, depressive symptoms flattened out after treatment. Future iterations of the UP-A may wish to include more relapse prevention techniques specific to depression (e.g., behavioral activation). In addition, results generally supported the tripartite model with regard to baseline associations, although some differences were noted between self and parent ratings. These discrepancies suggest important respondent considerations for conceptual understanding of similar and distinct affective characteristics of anxiety and depression. The relative lack of prediction regarding treatment response may speak to the robustness of the UP-A. Given the focus upon how adolescents respond to a broad array of emotional experiences, it may be that the intervention is effective for adolescents with a range of baseline NA and PA levels. One unexpected finding was that higher levels of AS were predictive of faster rate of change in depressive symptoms during treatment. It may be that adolescents with higher AS see a greater rationale for many treatment techniques (e.g., interoceptive exposure, behavioral activation) or responsiveness to such techniques, and thus show greater treatment engagement earlier in treatment. Despite study limitations, many of which were due to the small sample size, study findings offer important implications for conceptual understanding of adolescent anxiety and depression, as well as separate trajectories of adolescent anxiety and depression symptoms during a transdiagnostic treatment. Future work with larger samples should seek to replicate findings, as well as examine how changes in NA, PA, and AS may correspond to changes in anxiety and depressive symptoms during treatment.


anxiety; depression; adolescence; treatment; transdiagnostic; CBT