Diabetes is often classified into two main types: Type 1, which usually starts in childhood, and Type 2, which is more common in adults and linked to obesity. But researchers now know that not all Type 2 diabetes cases are the same.
People with Type 2 diabetes can vary widely in body weight, age of onset, and other factors. Recognizing these differences is important for understanding the disease and finding the best treatments.
A team from Stanford Medicine has created a new tool to help with this. They developed an artificial intelligence (AI) algorithm that uses data from continuous glucose monitors to identify three of the four most common subtypes of Type 2 diabetes.
Continuous glucose monitors are small devices worn on the arm that track blood sugar levels throughout the day. This technology could help people understand their risk of diabetes earlier and take steps to manage it, such as adjusting their diet or exercise routine.
Around 13% of people in the United States, or about 40 million individuals, have diabetes, and another 98 million have prediabetes. Most of these cases are Type 2 diabetes.
While it’s traditionally been treated as one condition, doctors now recognize that Type 2 diabetes has different causes, and these differences affect how well certain treatments work.
For example, some medications might work better for one subtype than another. The new AI tool aims to provide an easier way for people to identify their specific subtype and improve their health.
One of the study’s leaders, Dr. Michael Snyder, shared a personal experience that highlights the need for better tools. Years ago, he found out he was prediabetic. To prevent diabetes, he tried increasing his muscle mass, a common strategy to improve insulin sensitivity.
However, it didn’t help because his diabetes risk wasn’t caused by insulin resistance. Instead, it stemmed from a problem with his beta cells, which produce insulin. This example shows how knowing the specific cause of high blood sugar can guide more effective actions.
The research was published in Nature Biomedical Engineering. The team studied 54 participants, including 21 with prediabetes and 33 healthy individuals. They used the continuous glucose monitors to track changes in blood sugar levels after participants drank a sugary drink.
The AI algorithm analyzed patterns in these changes and linked them to specific subtypes of Type 2 diabetes. For instance, some patterns indicated insulin resistance, while others pointed to problems with beta cells or other hormonal imbalances.
Traditionally, diagnosing diabetes involves checking blood sugar levels through a simple blood test. However, these tests don’t reveal why blood sugar is high. More detailed tests can uncover the underlying cause, but they are expensive and not practical for regular use.
Continuous glucose monitors offer a simpler and more accessible way to gather detailed information about blood sugar patterns and identify these causes.
The AI tool proved highly accurate, correctly identifying the subtypes of Type 2 diabetes about 90% of the time. This was better than standard metabolic tests and could be done at home using data from the monitors.
Knowing their specific subtype could help people choose the right treatment and even prevent complications, such as heart disease or liver problems.
The researchers hope this technology will make diabetes care more accessible, especially for people who live far from medical facilities or face financial challenges. By identifying early signs of insulin resistance or beta cell dysfunction, people can take steps to prevent diabetes or manage it better.
The team plans to continue testing the algorithm with people who already have Type 2 diabetes to refine the tool further.
In summary, this study highlights how AI and wearable technology can improve diabetes care. By using data from continuous glucose monitors, the tool identifies different subtypes of Type 2 diabetes, offering a more personalized approach to treatment.
This technology has the potential to make early diagnosis and management more accessible, improving health outcomes for millions of people.
If you care about diabetes, please read studies about Vitamin D and type 2 diabetes, and to people with diabetes, some fruits are better than others.
For more health information, please see recent studies that low calorie diets may help reverse diabetes, and 5 vitamins that may prevent complication in diabetes.
The research findings can be found in Nature Biomedical Engineering.
Copyright © 2025 Knowridge Science Report. All rights reserved.
