Remote Approaches for Optimizing Weight Loss (Elevate Weight Loss Trial)

Participants will be randomized at baseline to one of four weight loss intervention conditions: 1) automated program + coaching, 2) automated program + no coaching, 3) group videoconference program + coaching, and 4) group videoconference program + no coaching. All participants will be provided with weekly goals, instructed to self-monitor diet, exercise, and weight daily, and provided with feedback based upon these data. Those receiving the online program will also be instructed to view video lessons (24 in total, 10-15 min each) which focus on behavioral strategies for changing diet and exercise. Individuals randomized to the videoconference program will participate in 24 group sessions (1 hour each) designed to mimic in-person treatment and allow for participant interaction via large and small group discussions. The content across conditions will be similar and the lessons will be provided weekly during months 1-3, twice per month during months 4-6, and monthly during months 7-12. Those receiving coaching will have individual, monthly meetings with a coach via videoconference (\~15 min each). Coaching sessions will focus on individual barriers, problem solving, goal setting, and fostering support and accountability. The primary aim of this study is to examine the effects of delivery format (automated online vs. videoconference) and coaching (individualized vs. none) on weight loss at 12 months. Secondary aims will examine the effects of delivery format and coaching on intervention engagement (e.g., frequency of self-monitoring), psychosocial outcomes (e.g., perceived support, self-efficacy, and motivation), 18-month weight loss, and the cost per kilogram of weight loss (to examine whether the addition of human support is cost-effective). Two algorithms will also be developed to predict which treatment type should be recommended to whom, using only baseline characteristics: 1) a 'widely-applicable' algorithm which will use metrics common to electronic medical records (sex, BMI, age, race, ethnicity), and 2) a 'more comprehensive' algorithm which will further include additional baseline characteristics (e.g., education, household income, health literacy, group preference, etc).