JMIR Publications

ABSTRACT

Background:

Personalized behavioral recommendations through mobile applications have proven effective in preventing serious chronic diseases such as diabetes. Recent studies have primarily focused on the personally optimizing recommendations using reinforcement learning. However, the main problem with these approaches is that they focus on behavior changes and overlook clinical outcomes.

Objective:

In this study, we proposed a method for online planning of dietary and exercise recommendations to directly optimize postprandial glucose levels through behavioral changes.

Methods:

The proposed method is a multi-arm bandit based on a two-stage reward prediction model, wherein an action is a combination of the total carbohydrate intake and postprandial walking duration, and the reward is how low the postprandial glucose levels are. We realized predicting the reward for each action based on the predicted behavioral responses to an action, and subsequently, the postprandial glycemic response, using the predicted responses.

Results:

In a simulation experiment, we demonstrated that the proposed online algorithm can significantly improve postprandial glucose levels with personalized recommendations, compared to the randomized policy. Furthermore, we conducted a small real-world experiment with a simplified proposed method involving a single update of the recommendation policy into a personalized one. For six human subjects, compared to the randomized policy, we confirmed 23% improvement (on average) in actual glucose responses along with the behavioral adherence to the recommendations concerning carbohydrate intake and postprandial walking.

Conclusions:

The effectiveness of the proposed method was demonstrated from both the simulation experiment and the small real-world experiment. However, further longitudinal real-world experiment in patients with diabetes is needed to confirm the finding.