Abstract MP075: Profiles of Obese Depressed Adults in a Randomized Controlled Trial of Integrated Behavior Therapy
Objective: To identify distinct profiles of obese depressed patients in an ongoing trial that tests the effectiveness and implementation potential of an integrated care model in primary care combining standard behavioral weight loss treatment and problem solving therapy plus as-needed antidepressant(s).
Design and Methods: Analysis of variance for continuous and Chi-square tests for categorical variables compared 4 domains (sociodemographic, behavioral, clinical and psychosocial characteristics) among 4 severity categories of obese depressed participants (i.e., body mass index [BMI]<35 kg/m2 and Depression Symptom Checklist 20 [SCL20]<1.5; BMI<35 and SCL20≥1.5; BMI≥35 and SCL20<1.5; and BMI≥35 and SCL20≥1.5). Discriminant analysis identified combinations of characteristics that best differentiated these categories.
Results: Participants (n=322) had baseline mean (SD) age of 50.4 (12.2) years, BMI of 36.7 (6.4) kg/m2, and SCL20 of 1.5 (0.5). Pairwise comparisons showed participants in the 4 categories had similar sociodemographic characteristics, but they differed significantly in the other 3 domains (Table). Discriminant analysis showed the combination of higher abdominal obesity, problems with usual activities (eg, work and family or leisure activities), higher anxiety, poorer self-reported mental health, and more obesity-related problems differentiated the highest from lowest severity category. The combination of lower abdominal obesity, more obese-related problems, and higher anxiety differentiated the 2 intermediate (lower obesity/high depression vs. high obesity/lower depression) categories.
Conclusions: Discernable behavioral, clinical and psychosocial profiles reveal important variability according to comorbidity severity within a largely sociodemographically similar group of obese depressed adults. This insight may inform future analysis of heterogeneity of treatment effects and targeting of strategies to baseline participant profiles in a multicomponent intervention.
Author Disclosures: J. Ma: None. L. Xiao: None. N. Lv: None. L. Goldman Rosas: None. M. Snowden: None. M.A. Lewis: None. J.D. Goldhaber-Fiebert: None. E.M. Venditti: None. L. Lesser: None. E. Ward: None.
- © 2017 by American Heart Association, Inc.