AI Predicts and Identifies Subtypes of Type 2 Diabetes from Continuous Blood Glucose Monitor
By HospiMedica International staff writers Posted on 08 Jan 2025 |

Diabetes has traditionally been classified into two types — Type 1, which typically develops in childhood, and Type 2, which is often linked to obesity and tends to occur later in life. However, scientists have discovered that not all Type 2 diabetes patients are alike, with variations in factors such as body weight, age of onset, and other characteristics. There has been an increasing push to subclassify Type 2 diabetes, which accounts for 95% of all diabetes cases, in order to better understand the risk of related conditions, including cardiovascular, kidney, liver, and eye complications, and identify the underlying physiology of each individual's diabetes. At present, diabetes diagnosis relies solely on blood glucose levels, which can be measured through a simple blood test. However, these tests do not provide insight into the biological mechanisms behind elevated blood sugar. Understanding the underlying physiology requires metabolic tests, which are typically conducted in a research setting, but these tests are costly, cumbersome, and not feasible for routine clinical use.
On the other hand, continuous glucose monitors, which are available over the counter, can track high blood sugar and provide more detailed information about a person’s metabolic biology. Insulin, a hormone produced by the pancreas, regulates blood glucose by helping cells absorb it and use it for energy. When the pancreas produces insufficient insulin — a condition known as insulin deficiency — blood glucose levels rise. Insulin resistance, a common marker of diabetes, occurs when cells fail to respond to insulin, causing glucose to accumulate in the blood. Type 2 diabetes can also result from a defect in the production of incretin, a hormone released by the gut after eating that prompts the pancreas to secrete insulin, or from insulin resistance in the liver. Each of these four physiological subtypes of diabetes may require different therapeutic approaches.
Researchers from Stanford Medicine (Stanford, CA, USA) have now developed an AI-based algorithm that uses data from continuous glucose monitors to differentiate between three of the four most common Type 2 diabetes subtypes. The researchers sought to determine whether a widely used device, such as a continuous glucose monitor, could provide data with hidden signals that correlate to different subtypes of diabetes. The monitor, which is worn on the upper arm, measures real-time fluctuations in blood sugar levels. When people consume a glucose drink, their blood sugar levels typically spike, but the intensity and pattern of these spikes can differ between individuals.
In a study involving 54 participants — 21 with prediabetes and 33 healthy individuals — the researchers applied an AI-powered algorithm to identify patterns in the peaks and troughs of blood sugar levels that correspond to various Type 2 diabetes subtypes. Participants who used the continuous glucose monitors also underwent the oral glucose test at a doctor’s office. When the data from the continuous glucose monitors were compared with clinical data and other metabolic biomarkers, the algorithm was able to predict metabolic subtypes, such as insulin resistance and beta-cell deficiency, with greater accuracy than traditional metabolic tests.
The tool, detailed in a paper published in Nature Biomedical Engineering, was able to accurately detect and identify subtypes approximately 90% of the time. In addition to providing higher-resolution data for individuals with diabetes or prediabetes, the continuous glucose monitor offers additional benefits. Even if someone with insulin resistance doesn’t develop diabetes, it remains important to identify the condition, as insulin resistance is a risk factor for several other health issues, such as heart disease and fatty liver disease. The researchers plan to continue testing the algorithm on individuals diagnosed with Type 2 diabetes and hope that the broad availability of the technology will improve access to care, especially for patients who are unable to attend doctor’s appointments. They also view this technology as an invaluable health tool for individuals who face economic or geographical barriers to accessing health care.
“People have looked at that for decades and have found certain parameters that indicate insulin resistance or beta cell dysfunction, which are the main drivers of diabetes,” said Tracey McLaughlin, MD, a professor of endocrinology. “But now we have the monitors, and you can get a much more nuanced picture of the glucose pattern which predicts these subtypes with greater accuracy and can be done at home.”
Channels
Artificial Intelligence
view channel
Innovative Risk Score Predicts Heart Attack or Stroke in Kidney Transplant Candidates
Heart researchers have utilized an innovative risk assessment score to accurately predict whether patients being evaluated for kidney transplants are at risk for future major cardiac events, such as a... Read more
AI Algorithm Detects Early-Stage Metabolic-Associated Steatotic Liver Disease Using EHRs
Liver disease, which is treatable when detected early, often goes unnoticed until it reaches advanced stages. Metabolic-associated steatotic liver disease (MASLD), the most prevalent form of liver disease,... Read moreSurgical Techniques
view channel
Easy-To-Apply Gel Could Prevent Formation of Post-Surgical Abdominal Adhesions
Surgical adhesions are a frequent and often life-threatening complication following open or laparoscopic abdominal surgery. These adhesions develop in the weeks following surgery as the body heals.... Read more
Groundbreaking Leadless Pacemaker to Prevent Invasive Surgeries for Children
Leadless pacemakers marked a significant advancement in cardiac care, primarily because traditional pacemakers are dependent on leads, which are prone to breakage over time. Currently, two FDA-approved... Read morePatient Care
view channel
Portable Biosensor Platform to Reduce Hospital-Acquired Infections
Approximately 4 million patients in the European Union acquire healthcare-associated infections (HAIs) or nosocomial infections each year, with around 37,000 deaths directly resulting from these infections,... Read more
First-Of-Its-Kind Portable Germicidal Light Technology Disinfects High-Touch Clinical Surfaces in Seconds
Reducing healthcare-acquired infections (HAIs) remains a pressing issue within global healthcare systems. In the United States alone, 1.7 million patients contract HAIs annually, leading to approximately... Read more
Surgical Capacity Optimization Solution Helps Hospitals Boost OR Utilization
An innovative solution has the capability to transform surgical capacity utilization by targeting the root cause of surgical block time inefficiencies. Fujitsu Limited’s (Tokyo, Japan) Surgical Capacity... Read more
Game-Changing Innovation in Surgical Instrument Sterilization Significantly Improves OR Throughput
A groundbreaking innovation enables hospitals to significantly improve instrument processing time and throughput in operating rooms (ORs) and sterile processing departments. Turbett Surgical, Inc.... Read moreHealth IT
view channel
Printable Molecule-Selective Nanoparticles Enable Mass Production of Wearable Biosensors
The future of medicine is likely to focus on the personalization of healthcare—understanding exactly what an individual requires and delivering the appropriate combination of nutrients, metabolites, and... Read more
Smartwatches Could Detect Congestive Heart Failure
Diagnosing congestive heart failure (CHF) typically requires expensive and time-consuming imaging techniques like echocardiography, also known as cardiac ultrasound. Previously, detecting CHF by analyzing... Read morePoint of Care
view channel
Handheld, Sound-Based Diagnostic System Delivers Bedside Blood Test Results in An Hour
Patients who go to a doctor for a blood test often have to contend with a needle and syringe, followed by a long wait—sometimes hours or even days—for lab results. Scientists have been working hard to... Read more
Smartphone-Enabled, Paper-Based Quantitative Diagnostic Platform Transforms POC Testing
Point-of-care diagnostics are crucial for public health, offering rapid, on-site testing that enables prompt diagnosis and treatment. This is especially valuable in remote or underserved regions where... Read moreBusiness
view channel
Becton Dickinson to Spin Out Biosciences and Diagnostic Solutions Business
Becton, Dickinson and Company (BD, Franklin Lakes, NJ, USA), has announced that its board of directors has unanimously authorized BD management to pursue a plan to separate BD's Biosciences and Diagnostic... Read more