Novel Artificial Intelligence Models Detect Type 1 Diabetes Risk Before Clinical Onset
New Studies Highlight Machine Learning as Potential Early Risk Detection Strategy
CHICAGO, June 20, 2025 /PRNewswire/ -- Developments from two studies highlighting the potential for machine learning leveraging artificial intelligence (AI) technology to improve early-stage identification of type 1 diabetes were presented as late-breaking poster sessions at the 85th Scientific Sessions of the American Diabetes Association® (ADA) in Chicago.
Each year, around 64,000 Americans are diagnosed with type 1 diabetes. As many as 40% are unaware they have the disease until they experience a life-threatening event that requires hospitalization. This is because the disease can progress silently until symptoms are present, such as excessive thirst, frequent urination, or diabetic ketoacidosis. By this point, significant and often irreversible damage to the cells that produce insulin has already occurred, highlighting the need for earlier detection and intervention strategies.
AI Model Reduces False Positives and Increases Accuracy in Type 1 Diabetes Risk Assessment Up to a Year Before Diagnosis
Presented as part of a late-breaking symposium, results from a new study demonstrate the potential for AI to more accurately identify individuals at risk for type 1 diabetes up to a year before diagnosis, with greater accuracy and fewer false positives than standard screening methods.
The retrospective cohort study developed two age-specific machine learning models—one for individuals aged 0–24 and another for those 25 and older—leveraging medical claims and lab test data from NorstellaLinQ. Researchers applied specific criteria to identify confirmed cases of stage 3 type 1 diabetes, including patients with at least two medical claims for type 1 diabetes, a higher frequency of type 1 versus type 2 diabetes claims, documented use of insulin or a continuous glucose monitor, and continuous medical and pharmacy claims activity in the two years leading up to diagnosis or treatment.
The models effectively identified risk of type 1 diabetes up to 12 months earlier than traditional screening methods. The models demonstrated high sensitivity in correctly identifying people with type 1 diabetes—approximately 80% in the younger group and 92% in adults. They also maintained improved precision compared to conventional screening methods, which typically yield a positive rate of just 0.3% in the general population.
"We're energized by the results of this study and what it could mean for early type 1 diabetes risk detection, potentially enabling more efficient and targeted screening for a disease that often goes undetected until a serious event prompts medical evaluation," said Laura Wilson, director health economics outcomes research, digital health at Sanofi. "By applying AI-driven predictive models to real-world data, we believe we can help identify individuals at high risk much earlier, giving them the opportunity to plan and prepare for the future."
The researchers plan to launch a multi-phase study to validate and refine a new clinical decision support tool for type 1 diabetes, working closely with leading hospital sites and experts. The research will integrate advanced AI models with hospital electronic health records, aiming to enable earlier, data-driven interventions for patients at risk.
AI Detects Type 1 Diabetes Risk More Than 18-fold Using U.S. Open Claims Data
Researchers used the Symphony Health Database, a large health care claims database covering 75 million patients, to train a machine learning model to identify people at risk for type 1 diabetes before they show symptoms. Records from nearly 90,000 individuals with type 1 diabetes to over 2.5 million people without type 1 diabetes were compared, using specific inclusion and exclusion criteria to define each group. Patterns in the records were analyzed to determine who was likely to develop type 1 diabetes. The model was tested on a large, real-world population and evaluated using a range of performance measures to determine how accurately it could predict risk.
Results showed machine learning models could successfully identify people at risk for type 1 diabetes before symptoms appeared, increasing detection efficiency more than 18-fold. Among those with type 1 diabetes, 29% had previously been misclassified as having type 2 diabetes or other forms, highlighting a critical gap in diagnostic accuracy that can delay appropriate treatment and increase the risk of complications.
Researchers found that the AI model that performed best was Bidirectional Encoder Representations from Transformers (BERT), a sophisticated tool originally designed for understanding language. BERT correctly identified 80% of true type 1 diabetes cases and was more accurate than other models, with a stronger odds ratio (97.27 vs. 38.01), meaning its predictions were far more likely to be accurate.
"By identifying individuals with presymptomatic type 1 diabetes, we have the opportunity to shift the entire timeline of care," said Jared Joselyn, senior vice president and global head of E.D.G.E at Sanofi. "These findings show how AI can uncover hidden patterns in routine health care data and help improve detection rates, with the goal of fostering more proactive, scalable care before disease progression."
Researchers note follow-up studies are needed to validate the approach using additional health care datasets from across the U.S. and internationally, as well as validating the predictions in a clinical setting. Future work will also explore enhancing model performance through multimodal AI techniques and by incorporating more longitudinal, genomic, and real-world data into broader clinical workflows to support earlier, data-driven intervention strategies.
Research presentation details:
Dr. Wilson will present the findings as a late-breaking poster session:
- Identification of Earlier Stage Autoimmune Type 1 Diabetes Using Machine Learning Algorithms
- Presented on Sunday, June 22 at 12:30 p.m. CT
Research presentation details:
Jared will present the findings as a late-breaking poster session:
- Predictive Modeling for Presymptomatic Type 1 Diabetes Detection Using Open Claims Data
- Presented on Sunday, June 22 at 12:30 p.m. CT
About the ADA's Scientific Sessions
The ADA's 85th Scientific Sessions, the world's largest scientific meeting focused on diabetes research, prevention, and care, will be held in Chicago, IL, on June 20–23. Thousands of leading physicians, scientists, and health care professionals from around the world are expected to convene both in person and virtually to unveil cutting-edge research, treatment recommendations, and advances toward a cure for diabetes. Attendees will receive exclusive access to thousands of original research presentations and take part in provocative and engaging exchanges with leading diabetes experts. Join the Scientific Sessions conversation on social media using #ADASciSessions.
About the American Diabetes Association
The American Diabetes Association (ADA) is the nation's leading voluntary health organization fighting to end diabetes and helping people thrive. This year, the ADA celebrates 85 years of driving discovery and research to prevent, manage, treat, and ultimately cure—and we're not stopping. There are 136 million Americans living with diabetes or prediabetes. Through advocacy, program development, and education, we're fighting for them all. To learn more or to get involved, visit us at diabetes.org or call 1-800-DIABETES (800-342-2383). Join us in the fight on Facebook (American Diabetes Association), Spanish Facebook (Asociación Americana de la Diabetes), LinkedIn (American Diabetes Association), and Instagram (@AmDiabetesAssn). To learn more about how we are advocating for everyone affected by diabetes, visit us on X (@AmDiabetesAssn).
Media Contact: Mimi Carmody, mcarmody@diabetes.org
View original content to download multimedia:https://www.prnewswire.com/news-releases/novel-artificial-intelligence-models-detect-type-1-diabetes-risk-before-clinical-onset-302486235.html
SOURCE American Diabetes Association