Abstract
INTRODUCTION: Depression and obesity are two of the most common chronic conditions among US adults. The Research Aimed at Improving Both Mood and Weight (RAINBOW) study was the first randomized clinical trial to demonstrate the effectiveness of an integrated behavioral treatment for depression and obesity. We aimed to characterize the trajectories of both depression symptoms and weight loss during the RAINBOW intervention. METHODS: Participants (N = 201) randomly assigned to the RAINBOW treatment arm who provided serial data on depression symptoms (via the 9-item Patient Health Questionnaire [PHQ9]) and self-measured weight were included. Group-Based Multi-Trajectory Modeling is a statistical technique that identifies participants who follow similar trajectories of multiple outcomes over time. This approach was used to characterize latent groups of both PHQ9 and weight loss percentage during treatment. Demographic differences between groups were investigated using Chi-Square, Fisher's exact tests, or ANOVA tests. Group membership was investigated for relationships with 12-month change in independently measured depression (via the 20-item Symptom Checklist, or SCL20) and weight (measured by trained research staff) using linear regression models. RESULTS: Participants were classified into 3 groups based on their depression symptom and weight trajectories: Moderate Depression Improvements/Minimal Weight Loss (N = 108, 53.7%), Substantial Depression Improvements/Moderate Weight Loss (N = 58, 28.9%), and Substantial Depression Improvements/Substantial Weight Loss (N = 35; 17.4%). Older participants were more likely to be in the Substantial/Substantial group. Participants in the Substantial/Moderate and Substantial/Substantial groups had larger 12-month depression symptom score changes and 12-month weight loss compared to the Moderate/Minimal group. CONCLUSION: We identified three joint trajectories of depression symptom scores and obesity. Members in the three trajectories saw varying degrees of symptom reduction from baseline to 12 months. Results suggest adaptive treatment trials may be needed for participants in groups with lower anticipated depression symptom scores and/or weight reductions.