A novel, semi-automated self-monitoring feedback system for health promotion in community settings

The National Institute of Diabetes and Digestive and Kidney Diseases awarded Aurora Health Care, Inc. $1,471,108 on September 15, 2025, under the Diabetes, Digestive, and Kidney Diseases Extramural Research program (CFDA 93.847) to develop a semi-automated self-monitoring feedback system for weight loss interventions in community settings.

The project will develop and optimize a feedback system that delivers weekly automated messages to participants based on their self-monitoring of dietary intake, physical activity, and weight. Study 1 will employ a micro-randomized factorial trial design with 300 adults with overweight or obesity over 16 weeks. Participants will use Fitbit activity monitors, Fitbit e-scales, and the Fitbit smartphone app to self-monitor daily and will receive basic weekly feedback on self-monitoring frequency and weight trajectory, plus randomly assigned feedback components addressing calorie goal attainment, diet quality, physical activity goal attainment, and prompted goal setting. The study aims to optimize the efficacy of feedback message construction by evaluating how different types and amounts of feedback affect weight change. The work will generate a feedback library and optimized algorithm for feedback provision to address the gap between community intervention weight losses (approximately 4% of baseline weight) and the 8–10% observed in effectiveness trials.

Performance occurs in Milwaukee, Wisconsin. The period of performance extends from the award date through June 30, 2030.

Abstract

Weight management interventions delivered in community settings demonstrate weight losses of ~4% of baseline weight, smaller than the 8-10% observed in effectiveness trials. The provision of weekly feedback related to self-monitoring of dietary intake, physical activity, and weight improves weight loss outcomes in adults with obesity; however, this feedback is often not provided in interventions delivered in community settings due to lack of training and time. To support effective feedback provision in community interventions, we propose to develop a semi-automated self-monitoring feedback system that can be integrated into a range of community settings. First, to address the lack of empirical evidence regarding optimal feedback construction, aim 1 will focus on optimizing the efficacy of feedback messages (including development of a feedback library and an optimized algorithm for feedback provision). In study 1, we propose to use a highly efficient micro- randomized factorial trial design to evaluate the impact of different types (in relation to behavioral target and interactivity) and amount of feedback on weight change in 300 adults with overweight or obesity. Participants in this study will receive a core weight loss 101 session and will be asked to self-monitor dietary intake, physical activity, and weight daily using study-provided tools (a fitbit activity monitor, fitbit E-scale, and the fitbit smartphone APP). Each week for 16 weeks participants will receive basic feedback on the frequency of their self-monitoring of dietary intake, physical activity, and weight and their weight trajectory, and will be randomly assigned to receive (or not receive) each of four additional feedback components (calorie goal attainment, diet quality, physical activity goal attainment, and prompted goal setting). We will use mixed-effects models to investigate the impact of each feedback component on weight change (primary outcome) and on proposed mechanisms (adherence to self-monitoring and to caloric intake and physical activity goals) during the week after IT is received. We will also investigate the impact of the number of feedback components received, potential moderators of feedback efficacy (e.g., age, SEX/gender, race/ethnicity, socioeconomic status, social support, food security) and whether different combinations of feedback are more effective when individuals are doing well in the program (i.e., meeting program goals) versus when they are not doing well, allowing US to develop personalized algorithms for feedback provision. For aim 2, we will employ user-centered design methods to develop, refine (via iterative development/testing cycles in study 2), and conduct "real world" usability testing with community intervention facilitators (study 3) of a semi-automated self-monitoring feedback system that aims to combine the efficiency of computer automation with the expertise of target users. Community intervention facilitators will be able to use this system to quickly develop effective self-monitoring feedback for delivery of this essential treatment component across a range of community intervention settings, broadly improving access to effective treatment for adult overweight and obesity.