SP 15: Microrandomized Trial to Optimize Use of Burden-reducing Self-monitoring Approaches in Behavioral Obesity Treatment

Self-monitoring (SM) is vital for effective behavioral obesity treatment (BOT), but adherence to dietary SM is often poor due to its high burden, leading to weight regain. Alternative, less burdensome SM approaches show promise in improving adherence and weight loss, yet optimal timing and application remain unclear. To address this, we will conduct a 24-week micro-randomized trial (MRT) with overweight/obese adults to compare gold-standard SM and four alternatives – reduced-frequency SM, lapse-only SM, smartwatch-based intake monitoring, and weight-only SM. Using MRT’s repeated randomizations, we will identify which SM strategies work best for whom and when, considering individual differences and social factors. Reinforcement learning will develop an adaptive algorithm to personalize SM recommendations, maximizing adherence and weight loss. This data-driven approach aims to enhance scientific understanding and optimize SM for improved obesity treatment outcomes.

This SP is interested in the personalization algorithms proposed by TR&D2, particularly how causal domain expertise (Aim 2) can be utilized to improve personalization.  Further this SP is interested in potentially using the (under development) JusTIn Toolkit for Just-in-Time Adaptive mHealth Interventions and the pJITAI toolbox.  If the TR&D algorithms under Aim 2 are demonstrated to be robust, this SP is potentially interested in assessing feasibility of these algorithms for use in informing the SP’s future research. TR&D2 pushes its personalization algorithms, specifically those that incorporate causal domain expertise to improve SM strategy personalization. A more sophisticated, personalized approach to dietary SM will enhance BOT outcomes by improving SM adherence and weight loss.

TBA