When is excess weight harmful to health?

When is excess weight harmful to health?

Not every extra pound automatically makes you sick. A new model now reveals where the true risk lies. © smartboy10/ iStock

BMI alone is not enough: How susceptible an overweight person is to diabetes, cardiovascular diseases, etc. does not just depend on body weight. A new model now reveals how high the individual risk actually is using 20 simple questions and a blood test. It shows that the highest risk for many diseases is not always those with the highest body mass index, but often people who are “just” overweight instead of obese, as the researchers report in “Nature Medicine”.

More and more people around the world are overweight or obese – often with consequences for their health. Because obesity promotes, among other things, diabetes, high blood pressure, chronic inflammation and cardiovascular diseases. The risk of some types of cancer such as colon cancer or breast cancer as well as dementia is also increased. At the same time, obesity causes changes in the brain, in the intestinal flora and in the fat cells themselves, which make it more difficult to lose weight – this is often a vicious circle for those affected.

Obesity and the individual health risk

But not every person who is overweight or obese automatically becomes ill. It has long been known that the body mass index (BMI) alone is not sufficient to determine the individual risk of disease. “But the parameters that are necessary to identify clinical, disease-causing obesity and to predict possible complications have so far remained unclear,” explain Kamil Demircan from Queen Mary University of London and his colleagues.

Risk profiles
Risk profiles of three people with the same age, gender and BMI. The differences are clearly visible. © Demircan et al. / Nature Medicine, CC by 4.0

A new model now offers a solution. To do this, Demircan and his team analyzed the health data of almost 200,000 people from the UK Biobank, a British long-term study. Using artificial intelligence, they evaluated more than 2,000 different parameters, including blood values, lifestyle factors and body measurements. “Our approach represents one of the largest studies of people with overweight or obesity,” the team writes.

20 factors reveal risk for 18 diseases

The result is the OBSCORE model, which can predict a person’s individual risk of 18 different diseases based on 20 factors, including type 2 diabetes, high blood pressure, apnea, gout, kidney damage and various cardiovascular diseases. “By systematically, data-driven analysis of a variety of health factors, we were able to identify a small group of factors that, taken as a whole, can help identify people at highest risk earlier,” says Demircan.

What’s interesting is that people with the highest BMI didn’t always have the highest risk of many of these diseases. Instead, a certain combination of lifestyle and metabolic factors can also put people with a BMI under 30 in the highest risk category, the researchers found. Conversely, the comparison of three people with the same age, gender and BMI showed a very different risk profile.

Help with prevention and more targeted treatment

These results confirm that body mass index alone is not sufficient to predict a person’s individual risk of disease. In the future, the OBSCURE model could help to distinguish “healthy” overweight people from people at risk of illness. The latter could then receive more targeted help than before and, for example, receive a GLP-1 active ingredient such as semaglutide or tirzepatide, even if they are not yet obese.

“As more and more people worldwide are affected by obesity, preventing the associated long-term health complications has become one of the greatest challenges for global health systems,” says senior author Claudia Langenberg from Queen Mary University and the Berlin Institute of Health at the Charité. “Here we have successfully developed a data-supported framework that identifies people at increased risk of complications.”

Source: Kamil Demircan (Queen Mary University of London) et al., Nature Medicine, 2026; doi: 10.1038/s41591-026-04353-2

Recent Articles

Related Stories