BACKGROUND

Unhealthy lifestyle behaviors have been identified as a major cause of numerous health issues, with a steady global increase in their prevalence. Addressing this challenge requires comprehensive behavioral changes to promote the adoption of a sustainable healthier lifestyle. However, despite the prevalent need, cost-effective and successful digital coaching for health-related behavior change remains scarce.

OBJECTIVE

This study aimed to present a holistic framework for designing, modeling, and executing behavior change strategies through a multiagent reasoning system that selected optimal digital coaching techniques based on individual assessments and integrated data-driven decision-making.

METHODS

Behavioral change theories have been explored to design a multiagent system aimed at achieving sustainable lifestyle changes. This system selected behavior change techniques based on individual user assessments, prioritizing those with the strongest impact on key behavioral components. The framework incorporated evidence-based practices stemming from behavioral change science and integrated them into Healthentia's behavioral change coaching scheme. Healthentia, a certified software as a medical device, implemented this framework in its non-medical modules that aim for lifestyle behavioral change and wellbeing specifically for chronic disease management, serving as an eHealth solution that advances decentralized care by enabling remote monitoring, data-driven content selection, and personalized digital coaching that adjusts to patient progress and engagement patterns.

RESULTS

This study explored the application of the Healthentia behavioral change coaching scheme in patients with type 2 diabetes. Behavioral attributes have been evaluated in 9 patients, yielding notable results in terms of fasting glucose dropping by an average of -17.3 mg/dL (Cohen d=1.5; P=.002), further underscored by a narrow 95% CI (-26.1 to -8.43), and in terms of weight and BMI, with mean reductions of -2.89 kg and -1.05 kg/m², respectively. These changes yielded large effect sizes (Cohen d approximately 1.05) and were statistically significant (P=.01). The positive outcomes were at least partly attributed to the personalized delivery of content, 71.66% (1125/1570) of which was well received by the patients.

CONCLUSIONS

Our study of this multiagent system, which was tested through simulated patient behavior and preliminary, limited behavior observations of patients with type 2 diabetes, promises improved health outcomes using personalized digital coaching strategies. Future directions include optimizing the multiagent selection process; further exploring the type 2 diabetes program; conducting an in-depth evaluation of its results, including glycated hemoglobin measurements; and expanding its applications to other chronic conditions.