Carrying too much fat in the body, known as adiposity, is a well-known cause of serious health problems like heart disease, stroke, type 2 diabetes, and kidney disease.
But measuring a person’s true health risk isn’t always easy. Traditional tools like body mass index (BMI) don’t tell the full story.
BMI can’t tell the difference between fat and muscle and doesn’t show where fat is located in the body.
A new study from Mass General Brigham and their partners has found a better way. Researchers used artificial intelligence (AI) to quickly and accurately measure fat in different parts of the body using full-body scans.
Their results, published in the journal Annals of Internal Medicine, showed that not all fat is equally dangerous. The location of fat in the body plays a big role in disease risk.
The research team hopes this new approach can be used to screen people for hidden health risks using scans they’ve already had for other reasons.
“We hope this could lead to a new tool that repurposes existing MRI and CT scans from hospitals to find people with dangerous fat patterns who may need help preventing diabetes and heart disease,” said Dr. Vineet K. Raghu, co-leader of the study.
To test this idea, the team used data from the UK Biobank, a large research project that includes health information from hundreds of thousands of people in the United Kingdom. They looked at full-body MRI scans from over 33,000 adults who had never been diagnosed with diabetes or heart disease. These people were followed for about four years.
Using AI, the researchers studied different types of fat and muscle in the scans. They found that fat stored deep in the belly, around organs (called visceral fat), and fat inside muscles were both strongly linked to a higher risk of developing diabetes and heart disease. This was true even after accounting for BMI and waist size.
In men, another risk factor was low muscle mass. Men with less skeletal muscle were more likely to develop these health problems, but this link was not seen in women.
The results show that using AI to analyze body scans can give doctors a much clearer picture of someone’s health risk than traditional methods. Fat stored around organs and inside muscles appears to be more harmful than fat stored elsewhere in the body.
The researchers say more studies are needed to confirm these findings and to make sure AI tools can work well in real-world hospital settings. But if validated, this technology could help doctors spot people at high risk much earlier and offer them personalized prevention plans.
If you care about diabetes, please read studies about This drug combo can treat type 2 diabetes in the long run effectively and findings of Eating fewer than 3 meals a day may help reduce risk of type 2 diabetes and obesity.
If you care about diabetes, please read studies about How to choosing the right fruits for type 2 diabetes and findings of New higher dose diabetes drug promises better blood sugar control and weight loss.
The study is published in Annals of Internal Medicine.
Copyright © 2025 Knowridge Science Report. All rights reserved.
