NCT05751993

Piloting a Reinforcement Learning Tool for Individually Tailoring Just-in-time Adaptive Interventions

• Status: Completed
• Phase: N/A
• Study type: Interventional
• Enrollment: 19 (actual)
• Sponsor: UNC Lineberger Comprehensive Cancer Center · Academic / Other
• Sex: All
• Age: 18 Years – 55 Years
• Healthy volunteers: Accepted

Summary

The purpose of this pilot study is to conduct a 12-week pilot feasibility study testing usability of a reinforcement learning model (AdaptRL) in a weight loss intervention (ADAPT study). Building upon a previous just-in-time adaptive intervention (JITAI), a reinforcement learning model will generate decision rules unique to each individual that are intended to improve the tailoring of brief intervention messages (e.g., what behavior to message about, what behavior change techniques to include), improve achievement of daily behavioral goals, and improve weight loss in a sample of 20 adults.

Detailed description

Reinforcement Learning (RL), a type of machine learning, holds promise for addressing the limitations of previous approaches to implementing JITAIs. Adaptive RL applications work by updating information about expected "rewards" (i.e., proximal outcomes) based on the results of sequentially randomized trials. To realize the potential of adaptive interventions to reduce health disparities in cancer prevention and control, mHealth interventionists first need to identify methods of using digital health participant data to continually adapt decision rules guiding highly tailored intervention delivery. This research team has developed a reinforcement learning model (AdaptRL) that reads in and analyzes user data (e.g., calories, weight, and activity data from Fitbit) in real-time, uses RL to efficiently determine which message a participant should receive up to 3 times per day, and creates a JITAI tailored to optimize daily behavioral goal achievement and weight loss for each participant. The objective of this study is to test the feasibility of using this reinforcement learning model in a pilot weight loss study.

Conditions

• Overweight and Obesity
• Overweight
• Obesity

Interventions

Type	Name	Description
BEHAVIORAL	ADAPT	The intervention is testing the feasibility of a reinforcement learning model to pull in participants' behavioral data (calories, activity, and weight) and use this data along with participants' past behavioral goal achievements to deliver the type of message that should be most effective for a given participant at a given time. At each decision point (morning, midday, and evening on a daily basis), the system evaluates which behaviors a participant is eligible to receive a message about (eating, activity, self-weighing), which intervention options a participant is eligible to receive, and then chooses what type of behavioral message a participant should receive. Over time, the model uses participant data and response to interventions to better tailor message choice.

Timeline

• Start date: 2025-04-12
• Primary completion: 2025-08-03
• Completion: 2025-08-18
• First posted: 2023-03-02
• Last updated: 2025-08-24

Locations

1 site across 1 country: United States