Just-in-Time Feedback in Diet and Physical Activity Interventions: Systematic Review and Practical Design Framework

Abstract

Background: The integration of body-worn sensors with mobile devices presents a tremendous opportunity to improve just-in-time behavioral interventions by enhancing bidirectional communication between investigators and their participants. This approach can be used to deliver supportive feedback at critical moments to optimize the attainment of health behavior goals.

Objective: The goals of this systematic review were to summarize data on the content characteristics of feedback messaging used in diet and physical activity (PA) interventions and to develop a practical framework for designing just-in-time feedback for behavioral interventions.

Methods: Interventions that included just-in-time feedback on PA, sedentary behavior, or dietary intake were eligible for inclusion. Feedback content and efficacy data were synthesized descriptively.

Results: The review included 31 studies (15/31, 48%, targeting PA or sedentary behavior only; 13/31, 42%, targeting diet and PA; and 3/31, 10%, targeting diet only). All studies used just-in-time feedback, 30 (97%, 30/31) used personalized feedback, and 24 (78%, 24/31) used goal-oriented feedback, but only 5 (16%, 5/31) used actionable feedback. Of the 9 studies that tested the efficacy of providing feedback to promote behavior change, 4 reported significant improvements in health behavior. In 3 of these 4 studies, feedback was continuously available, goal-oriented, or actionable.

Conclusions: Feedback that was continuously available, personalized, and actionable relative to a known behavioral objective was prominent in intervention studies with significant behavior change outcomes. Future research should determine whether all or some of these characteristics are needed to optimize the effect of feedback in just-in-time interventions.

Keywords: Internet; accelerometer; activity monitor; diet; exercise; health behavior; mHealth; self-tracking; task performance and analysis; wearable sensors.