Type 2 diabetes often develops quietly over many years. Many people have no obvious symptoms until serious damage has already occurred.
A new study published in the Journal of Clinical Epidemiology describes an artificial intelligence-based online screening tool called MEDWACS that may help people identify whether they are at risk of prediabetes or undiagnosed type 2 diabetes without leaving home.
Prediabetes means blood sugar is higher than normal but not yet high enough to be diagnosed as diabetes. Without lifestyle changes or treatment, many people with prediabetes eventually develop type 2 diabetes. Detecting these conditions early can help prevent complications such as heart disease, kidney disease, nerve damage and vision loss.
Researchers from the Technical University of Denmark developed MEDWACS using artificial intelligence trained on more than 30 years of health information from the U.S. National Health and Nutrition Examination Survey. The goal was to create a simple tool that anyone could use with basic household equipment.
Instead of relying on blood tests, the system asks questions and uses seven measurements that can be collected at home. These include age, sex, body mass index and blood pressure. One surprising measurement is thigh length.
The researchers selected these seven factors from nearly 3,700 possible variables after the AI identified those that best predicted diabetes risk. They then tested the system using separate groups of people from the United States and South Korea whose diabetes status was already known.
The validation showed that MEDWACS performed as well as, or better than, several recognised screening methods. This suggests it could become another useful way to encourage people at high risk to seek medical testing.
The inclusion of thigh length surprised the researchers. Doctors believe it reflects early childhood nutrition because poor nutrition can slightly reduce bone growth. Shorter thighs may also indicate lower muscle mass, and muscles are one of the body’s main tissues for removing glucose from the bloodstream.
The researchers believe the tool could also work well in Denmark, although it still needs to be validated using Danish health data. Experts from the Danish Diabetes Association welcomed the research while noting that established tools such as the Leicester Risk Assessment already perform well.
The authors stress that MEDWACS is not designed to diagnose diabetes. Instead, it is intended to identify people who should visit a doctor for proper blood testing.
This study highlights the growing role of artificial intelligence in preventive healthcare. Although the tool still requires validation in additional populations, it offers a convenient, inexpensive way to identify people who may unknowingly have diabetes or prediabetes. Earlier detection could allow treatment to begin sooner and reduce the long-term burden of diabetes.
If you care about diabetes, please read studies about diabetes and vitamin B12, and the right diet for people with type 2 diabetes.
For more health information, please see recent studies about how to eat smart with diabetes, and turmeric and vitamin D: a duo for blood pressure control in diabetic patients.
Source: Technical University of Denmark.
