Nestoras Mathioudakis, MD MHS

Exciting news from our team! Our recent JAMA study tested an AI-powered Diabetes Prevention Program (DPP) and showed that a fully automated, digital approach can achieve results comparable to traditional human-coach programs.

Thank you to Dr. Leigh Perreault, a leading researcher in for her insightful JAMA editorial on our trial. She highlighted the key question we’re now tackling: how to bring these programs into real-world practice—expanding access while maintaining effectiveness over time.

As more AI-based prevention programs become available, we’ll continue learning what works best, for whom, and how to deliver diabetes prevention at scale.

💬 Editorial: An AI-led diabetes prevention program matched the effectiveness of human-led intervention, with greater adoption and maintenance, highlighting its potential for scalable impact on diabetes prevention.

https://ja.ma/4ou84rw

vs. Human coach for prediabetes: our randomized trial finds a tie.

In 368 adults with prediabetes, an AI-led Diabetes Prevention Program (Sweetch) matched human coach–led programs at 12 months on CDC benchmarks:
- 5% weight loss
- 4% weight loss plus 150 minutes/week of activity (tracked by actigraph wrist monitor)
- A1C drop of at least 0.2%

Why this matters:
- One of the first head-to-head randomized tests of AI vs humans in prediabetes (a condition affecting more than 1 in 3 U.S. adults)
- Shows AI can help deliver proven prevention at scale—alongside clinicians

Grateful to our study participants, my research team, collaborators, and JAMA for publishing this work. Johns Hopkins Medicine. This National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK) funded study provides new evidence that may reshape how we approach prediabetes prevention and behavior change at scale. Studies like this remind us why research saves lives – not just by discovering new treatments, but by revealing new ways to deliver proven interventions at scale.

Among adults with and overweight or , referral to a fully automated -led program was noninferior to referral to human-led coaching in achieving a composite outcome based on weight reduction, physical activity, and HbA1c.

https://ja.ma/48HwcCk

Proud to share results of our latest research study, 3 years in the making, involving an international team of experts trying to identifying novel markers to improve prediction of cardiovascular disease in . There are some signals of promising markers, but we need more rigorous studies to understand how well these markers help to enhance our ability to identify patients at risk of adverse cardiovascular outcomes. https://www.hopkinsmedicine.org/news/newsroom/news-releases/2024/01/international-consortium-identifies-biomarkers-that-improve-prediction-accuracy-of-cardiovascular-disease-risk-in-people-with-type-2-JHM Diabetes Prevention & Education Program

International Consortium Identifies Biomarkers That Improve Prediction Accuracy of Cardiovascular Disease Risk in People with Type 2 Diabetes Respiratory illnesses continue to increase across the nation, resulting in rising hospitalizations. Effective 1/12/2024, mandatory masking will resume for patients, visitors and employees in all Johns Hopkins Medicine locations in the state of Maryland. Masking is required regardless of vaccination....