Millions of people around the world are living with type 2 diabetes without knowing it. Because the disease often develops slowly, many people do not visit a doctor until complications appear.
Researchers have now created an online screening tool that could help identify people at risk from the comfort of their own home. Their findings were published in the Journal of Clinical Epidemiology.
The new system, called MEDWACS, was created by researchers at the Technical University of Denmark together with collaborators in Italy. It uses artificial intelligence to estimate a person’s likelihood of having prediabetes or undiagnosed type 2 diabetes.
Unlike a medical diagnosis, the tool is meant as an early warning system. If it suggests someone is at high risk, they can then arrange a blood test with their doctor to confirm whether diabetes is present.
The AI was trained using more than three decades of information from a large American health survey that included adults of different ages and backgrounds. Researchers then tested the program on separate groups of people from both the United States and South Korea.
The results showed that MEDWACS accurately identified many people with prediabetes or undiagnosed diabetes. Its performance matched or exceeded several existing screening approaches.
To use the tool, people answer a few simple questions and take basic measurements using items commonly found at home, including a bathroom scale, tape measure and blood pressure monitor. The most unusual measurement is the length of the thigh.
Researchers explain that thigh length may provide clues about nutrition during early childhood. The thighs also contain the body’s largest muscles, which help remove sugar from the bloodstream. People with smaller muscle mass may have a reduced ability to process glucose efficiently.
Experts say one of the biggest strengths of the tool is that it removes barriers to screening. Many people avoid medical tests because they feel healthy or do not realise they are at risk. A simple online assessment could encourage earlier medical evaluation.
The researchers believe the tool is likely to perform well in Denmark because it has already worked across very different populations. However, they emphasise that studies using Danish health data are still needed before widespread use.
The research demonstrates how artificial intelligence may improve early disease detection without replacing doctors. MEDWACS cannot diagnose diabetes, but it may help identify people who would benefit from further testing.
If future studies confirm its accuracy in more countries, it could become a valuable public health tool for finding diabetes earlier and preventing serious complications.
If you care about diabetes, please read studies about bananas and diabetes, and honey could help control blood sugar.
For more health information, please see recent studies about Vitamin D that may reduce dangerous complications in diabetes and plant-based protein foods may help reverse type 2 diabetes.
Source: Technical University of Denmark.
