A recent study published in JAMA has found that an AI-powered diabetes prevention program (DPP) app can reduce the risk of diabetes in adults with prediabetes similarly to traditional human-led programs. The study, funded by the National Institutes of Health, compared the effectiveness of an AI-powered DPP app to human-led programs in over 368 participants.
Key Findings
- Similar outcomes: Both the AI-powered DPP app and human-led programs achieved similar rates of diabetes risk reduction, with 31.7% and 31.9% of participants meeting the CDC-defined composite benchmark, respectively.
- Higher initiation and completion rates: The AI-powered DPP app had higher rates of program initiation (93.4% vs 82.7%) and completion (63.9% vs 50.3%) compared to traditional human-led programs.
Implications
The study suggests that AI-powered DPPs could be an effective alternative to existing human-coached programs, especially for patients with logistical constraints. The AI-powered app’s ability to provide personalized interventions and always be available could extend its reach and make it a valuable tool in diabetes prevention.
Future Directions
The study team plans to explore how the AI app outcomes translate to broader, underserved patient populations and investigate patient preference, engagement, and costs associated with AI-led DPPs. This research has the potential to inform the development of more effective and accessible diabetes prevention programs.