A new artificial intelligence tool called DiaCardia may soon make it possible to screen for prediabetes using only data from an electrocardiogram (ECG)—with no need for a blood test.
This breakthrough could allow people to check their risk of prediabetes from home using wearable devices, helping to catch the condition early and prevent full-blown diabetes.
Type 2 diabetes is a serious disease that occurs when the body can’t produce or use insulin properly. It leads to high blood sugar and raises the risk of many health problems, including heart disease and kidney failure. But before developing diabetes, people often go through a stage called prediabetes.
At this stage, blood sugar is higher than normal but not high enough to be called diabetes. It’s a key time to make lifestyle changes that can stop diabetes from developing.
Unfortunately, prediabetes can be hard to detect because it usually doesn’t cause symptoms. Many people don’t get regular checkups or avoid blood tests due to cost or inconvenience. That’s where this new AI tool could make a big difference.
The research team, led by Dr. Chikara Komiya, Dr. Ryo Kaneda, and Professor Tetsuya Yamada from Science Tokyo in Japan, created DiaCardia using a form of machine learning called LightGBM.
They trained it using health checkup data from more than 16,000 people, including their ECG signals, fasting blood sugar levels, hemoglobin A1c (a measure of long-term blood sugar), and whether they were being treated for diabetes.
The AI model looks at 269 small features hidden in ECG waveforms—tiny patterns in the heart’s electrical activity that can reflect changes in the body caused by prediabetes or diabetes.
It achieved strong results, correctly identifying people with prediabetes in tests with high accuracy. It worked well even when tested on data from another clinic that wasn’t used to train the model.
The team used a method called SHAP analysis to figure out how the AI was making its predictions. It turned out that certain heart signal features—like stronger R-waves and lower heart rate variability—were key clues.
These changes may be linked to insulin resistance and damage to the nervous system, both of which are known to occur in people with high blood sugar.
Most excitingly, DiaCardia still performed well even when using ECG data from just a single lead—the kind of data that can be collected by many smartwatches and fitness bands. This means people could eventually screen themselves for prediabetes using a device they wear every day.
“DiaCardia has the potential to make prediabetes screening scalable, accessible, and available anytime, anywhere, without a blood test,” said Dr. Komiya. If developed further, it could help millions of people catch blood sugar problems early and take action to prevent diabetes.
This technology opens the door for large-scale, affordable screening that could be done at home, at pharmacies, or even during routine heart health checks. It could make a big difference in how we fight one of the world’s fastest-growing health problems.
If you care about blood sugar, please read studies about why blood sugar is high in the morning, and how to cook sweet potatoes without increasing blood sugar.
For more information about brain health, please see recent studies about 9 unhealthy habits that damage your brain, and results showing this stuff in cannabis may protect aging brain, treat Alzheimer’s.
Copyright © 2025 Knowridge Science Report. All rights reserved.
