Search
Health

BMI misjudges obesity in a notable share of adults, study of 844,000 people finds

A Nature study tested a 2025 definition that separates obesity already causing harm from obesity that hasn't yet, and the gap in death risk was wide.

A clinic room with a tape measure, scale and stethoscope, and two anonymous figures with similar outlines but different inner glows.
Illustration: ViewTime

Key Takeaways

  • BMI alone missed obesity in up to 8.6% of participants and wrongly flagged up to 3.1% of them
  • Up to a third of obesity cases met the bar for clinical obesity, meaning organ damage or physical impairment
  • Clinical obesity was tied to a 69% higher death risk over more than a decade, versus 19% for preclinical obesity

Body-mass index, the height-and-weight formula doctors have used for decades to diagnose obesity, gets the call wrong for a meaningful share of adults, according to a large international study published in Nature. When researchers compared BMI with a newer, more detailed definition of obesity, BMI missed the condition in up to 8.6 percent of participants and labeled up to 3.1 percent as having obesity when they did not meet the newer criteria.

The analysis covered 844,108 adults drawn from four groups in China, the U.K., the U.S. and Spain. The "up to" figures reflect the high end across those groups, so the size of the error varied by population. Even so, at the scale of the world's adult population, error rates in that range would translate into millions of people with the wrong diagnosis.

The more important finding may be what the new definition reveals about risk. It splits obesity into two categories, and the people in them fared very differently over time.

Two kinds of obesity, two levels of risk

The framework being tested was put forward in 2025 by a group of experts. It moves beyond BMI by adding measures such as waist circumference, waist-to-height ratio and body fat scans. It also asks whether excess fat is already causing harm, such as damage to organs or physical limits that interfere with daily activities.

People whose excess fat is doing that kind of damage have what the framework calls clinical obesity, an active disease. People who carry excess fat without measurable harm to bodily function have preclinical obesity, a state of elevated risk rather than illness.

In the new study, clinical obesity was linked to a 69 percent higher risk of dying over follow-up periods longer than a decade. For preclinical obesity, the figure was 19 percent. Up to a third of all obesity cases qualified as clinical, said Francesco Rubino, a bariatric surgery consultant at King's College London, who called it "an active disease that should be treated as such."

That gap is the strongest argument in the paper. Two people with identical BMIs could land in different categories, and the study suggests their outlooks differ sharply. BMI by design cannot see that difference, because it knows nothing about where fat sits or what it is doing to the body.

Preclinical obesity is not a clean bill of health, though. "Most people with preclinical obesity, but not all, remained free of clinical obesity for a decade," Rubino said. He was careful about the middle ground, too: "Preclinical obesity should therefore be neither treated automatically as active disease nor dismissed as harmless." Some people in that group may benefit from preventive treatment, he said, with intensity matched to their individual risk.

An Pan, an epidemiologist at Huazhong University of Science and Technology in China, said the differences in outlook between the two categories held up consistently across countries and ethnic groups. In his view, that consistency supports adopting the framework worldwide.

What the findings mean for treatment

The practical stakes center on GLP-1 drugs, the class that includes semaglutide and tirzepatide. Proponents of the 2025 framework hope that if doctors adopt it, those medicines can be steered toward patients who need them most, instead of being offered to everyone who crosses a BMI line.

That is a sensible goal. A definition that separates active disease from risk gives doctors and insurers a clearer basis for prioritizing care than a single number does. It also cuts the other way: the 8.6 percent of people BMI missed are, by this measure, people with obesity who might not be getting attention at all.

"People with obesity, like everyone else, deserve care matched to their needs: no less, and no more," Rubino said.

The researchers also see uses beyond the clinic. A sharper definition could help studies untangle whether a condition like heart disease stems from obesity or from something else.

There are real limits, however. The mortality figures describe associations observed in large populations over time; they show that people with clinical obesity died at higher rates, not that the label itself predicts any one person's outcome. The study also compared definitions rather than testing treatments, so it does not show whether acting on the new categories improves health.

Then there is the question of whether busy clinics can use the framework at all. A tape measure is cheap and fast. Body fat scans and assessments of organ function and daily impairment take time, equipment and money, and BMI has lasted as long as it has largely because it takes seconds to calculate. A more accurate definition helps only if it fits into a routine appointment.

The team plans further research on how to apply the criteria effectively and on refining the diagnostic methods. Future studies could also test how treatments affect people in each category, which is the evidence doctors will need before the new definition changes who gets prescribed what.

Sources

Iris Walker

Iris Walker writes about health research, with a focus on the background needed to interpret new papers and public announcements. Her coverage connects findings with earlier evidence and explains the questions researchers are investigating. She emphasizes clear terminology, appropriate context, and the uncertainties that remain as scientific understanding develops.

View all stories