Artificial Intelligence for Type 2 Diabetes

Not yet recruiting at 1 trial location

Overseen By Avani A Narayan, MS

Age: < 65

Sex: Any

Trial Phase: Academic

Sponsor: University of California, San Francisco

Must not be taking: Glucocorticoids, Vitamin C

Disqualifiers: Pancreatic autoantibody positivity, Bariatric surgery, others

Compensation: Varies

Number of Visits: Unknown

What You Need to Know Before You Apply

What is the purpose of this trial?

This trial focuses on using Artificial Intelligence (AI) to better understand and predict the progression of Type 2 Diabetes (T2D), particularly in young people. The researchers aim to develop a digital twin—a virtual model of a patient—to demonstrate how different lifestyle and medication changes might affect the disease. Participants will be divided into two groups: one will receive personalized insights and recommendations based on their digital twin, while the other will receive standard care. The trial seeks young people aged 10 to 21 who have had T2D for at least three months and currently have an HbA1c level of 7% or higher, a common marker for blood sugar control. As an unphased trial, this study offers a unique opportunity to contribute to groundbreaking research that could transform diabetes care for young people.

Will I have to stop taking my current medications?

The trial requires that your medication regimen remains stable, meaning no changes to your medications or basal insulin dose by more than 20% in the two weeks before joining. You may need to stop taking more than 500mg/day of Vitamin C during the study.

What prior data suggests that these AI-based methods are safe for use in predicting Type 2 Diabetes progression?

Research has shown that using AI (artificial intelligence) in managing diabetes can improve patient health without major safety concerns. AI systems, for example, can analyze blood sugar data and adjust insulin levels in real-time, helping to maintain stable blood sugar levels.

A review of several studies found that AI can predict the risk of developing type 2 diabetes by examining large amounts of data. These AI tools enhance doctors' understanding of risk factors, potentially leading to early treatment.

Another report found that AI tools increased patient involvement and slightly lowered A1C levels, which measure blood sugar control, without any serious side effects.

Although AI in healthcare is still new, current evidence suggests that AI methods are generally safe and effective in diabetes care. ^{1 2 3 4 5}

Why are researchers excited about this trial?

Researchers are excited about AI-based methods for managing Type 2 Diabetes because they offer a personalized approach to treatment. Unlike standard care, which typically involves periodic medication adjustments and lifestyle advice based on routine blood tests, this innovative method uses a "digital twin" to simulate different treatment scenarios. This digital twin provides patients with projections of their disease progression in various alternate realities, allowing for tailored recommendations on medication and lifestyle changes. The interactive, game-like presentation on an iPad makes the information more engaging and easier to understand, potentially leading to better adherence and outcomes.

What evidence suggests that AI-based methods could be effective for Type 2 Diabetes?

Research has shown that AI-based methods can greatly improve blood sugar control in people with type 2 diabetes. One study found that AI interventions lowered HbA1c levels by up to 2.59%, a significant decrease. Another study discovered that over 70% of participants using an AI-powered lifestyle app managed to lower their HbA1c and even reduce or stop their medications. In this trial, participants in the digital twin arm will receive AI-based methods that simulate different treatment scenarios, illustrating how changes in medication, diet, or exercise can affect their health over time. This approach holds promise for better managing type 2 diabetes. ^{2 3 6 7 8}

Who Is on the Research Team?

Shylaja A Srinivasan, MD

Principal Investigator

University of California, San Francisco

Are You a Good Fit for This Trial?

This trial is for youth with Type 2 Diabetes. Participants should have a history of the disease and be willing to use an AI-powered phone application as part of their care. Specific eligibility criteria are not provided, but typically include factors like age range, disease severity, and willingness to adhere to study protocols.

Inclusion Criteria

HbA1C ≥ 7% which is the target HbA1C recommended by the American Diabetes Association

Willing to abide by recommendations and study procedures

I have had type 2 diabetes for at least 3 months.

Exclusion Criteria

Presence of a condition or abnormality that in the opinion of the Investigator would compromise the safety of the patient or the quality of the data

I expect to use steroids during the study.

I can't stop taking more than 500mg/day of Vitamin C during the study.

Timeline for a Trial Participant

Screening

Participants are screened for eligibility to participate in the trial

2-4 weeks

Digital Twin Model Development

Development and validation of an AI-powered prediction model for T2D disease progression using existing data

Ongoing throughout the trial

Randomized Controlled Trial

Evaluation of the digital twin model by comparing generated data to observed values over one year

12 months

Quarterly visits (in-person or virtual)

Follow-up

Participants are monitored for safety and effectiveness after the main trial

4 weeks

What Are the Treatments Tested in This Trial?

Interventions

• AI-based Methods

Trial Overview

The trial is testing an AI prediction model against standard diabetes care. It aims to predict long-term disease progression using clinical data and motivate behavioral change in young patients through personalized healthcare recommendations based on their 'digital twin'—a virtual simulation of their health.

How Is the Trial Designed?

Group I: Digital twin arm Experimental Treatment 1 Intervention

Participants in the digital twin arm will receive information on their disease progression which will be based on projected change in HbA1C in alternative realities and specific recommendations on medication dosing and lifestyle changes based on this data. The digital twin information will be presented on an iPad in a game- like manner. The alternate realities will include scenarios of change in medication adherence, physical activity metrics, dietary changes etc.

Group II: Control arm Placebo Group 1 Intervention

Participants in the control arm will receive standard of care which is medication change recommendations based on HbA1C and blood glucose values every 3 months and standard lifestyle education.

Who Is Running the Clinical Trial?

University of California, San Francisco

Lead Sponsor

Trials

2,636

Recruited

19,080,000+

American Diabetes Association

Collaborator

Trials

148

Recruited

102,000+

Stanford University

Collaborator

Trials

2,527

Recruited

17,430,000+

Citations

1.

pmc.ncbi.nlm.nih.gov

pmc.ncbi.nlm.nih.gov/articles/PMC12456749/

A Scoping Review of Artificial Intelligence-Based Health ...

(2) Effectiveness: AI interventions enhanced glycemic control, yielding reductions in glycosylated hemoglobin (HbA1c) of up to 2.59%, improved ...

2.

consultqd.clevelandclinic.org

consultqd.clevelandclinic.org/randomized-clinical-trial-demonstrates-ai-supported-coaching-program-improved-outcomes-in-patients-with-t2d

AI-Supported Coaching Program Improved T2D Outcomes

More than 70% of people with type 2 diabetes using a new AI-powered lifestyle app reduced their HbA1C and reduced or eliminated medications.

3.

sciencedirect.com

sciencedirect.com/science/article/pii/S2666990024000089

Artificial intelligence for diabetes: Enhancing prevention, ...

This systematic review explores the role of Artificial Intelligence (AI) in enhancing the prevention, diagnosis, and management of diabetes.

4.

nature.com

nature.com/articles/s41746-024-01034-7

AI-based diabetes care: risk prediction models and ...

Notably, usage has expanded to predict an individual's risk for developing type 2 diabetes. A scoping review of 40 studies by Mohsen et al.

5.

diabetesjournals.org

diabetesjournals.org/care/article/46/5/959/148541/An-Integrated-Digital-Health-Care-Platform-for

An Integrated Digital Health Care Platform for Diabetes ...

Among adults with T2D, use of an integrated digital health care platform with AI-driven dietary management resulted in better glycemia and more ...

6.

pmc.ncbi.nlm.nih.gov

pmc.ncbi.nlm.nih.gov/articles/PMC10591058/

Artificial intelligence in diabetes management - PubMed Central

AI could be employed in understanding risk factors for diabetes onset owing to human limitations and biases when dealing with large spaces of risk factors.

7.

dmsjournal.biomedcentral.com

dmsjournal.biomedcentral.com/articles/10.1186/s13098-025-01819-0

Effectiveness and safety of AI-driven closed-loop systems in ...

AI-based closed-loop systems can analyze glucose data in real-time and automatically adjust insulin delivery, resulting in reduced time outside target glucose ...

8.

aha.org

aha.org/aha-center-health-innovation-market-scan/2024-07-02-type-2-diabetes-patients-can-benefit-ai-powered-nudges-report

Type 2 Diabetes Patients Can Benefit from AI-Powered ...

The report states that patient engagement tripled and A1C numbers decreased by 0.4 points, according to new data that the telehealth provider ...