BMI Is Losing Its Monopoly: Huge Nature Study Backs a New Way to Diagnose Obesity

Your body mass index is older than the telephone. The Belgian statistician Adolphe Quetelet came up with the formula in the 1830s to describe the “average man,” not to diagnose anybody. The American physiologist Ancel Keys gave it its modern name in 1972. Almost two centuries later, a number that fits on a sticky note still decides who gets labelled obese in doctors’ offices, insurance forms and drug trials. A large study published on November 11 in Nature suggests that era is ending.

The paper tested a new definition of obesity on more than 844,000 adults across four countries. Its verdict on BMI is polite but unmistakable: useful as a first screen, misleading as a final answer.

What the Nature study actually tested

In January 2025, a Lancet commission of 56 experts chaired by Francesco Rubino, a professor of bariatric and metabolic surgery at King’s College London, proposed rethinking obesity from the ground up. Instead of one BMI cut-off, the commission suggested confirming excess body fat with additional measurements such as waist circumference or waist-to-height ratio. It then split obesity into two categories. Clinical obesity means excess fat plus at least one of 18 signs of organ dysfunction or physical impairment, such as heart problems, joint pain or sleep apnea. Preclinical obesity means excess fat without that kind of damage, at least so far.

The new study, led by researchers at Huazhong University of Science and Technology and North China University of Science and Technology with partners including Pennington Biomedical Research Center and King’s College London, put that framework through a real-world stress test. Rubino is one of the co-authors, which is worth knowing when you read the conclusions. The team analysed four large cohorts: the Kailuan Study in China, the UK Biobank, the All of Us Research Program in the United States and the SIDIAP primary care database in Spain. Participants were 18 to 74 years old and had no history of stroke, heart attack or cancer at the start.

BMI gets it wrong in both directions

The headline finding is that BMI misclassifies people both ways. According to the study summary published by News-Medical, BMI overestimated obesity in 1.2 to 3.1 percent of participants and underestimated it in 5.2 to 8.6 percent. Adding waist measurements raised obesity prevalence by 3.3 to 7.4 percentage points depending on the country. In the US cohort, the share rose to 48.1 percent.

That sounds like bad news. In raw numbers, it is. More interesting is what happened to risk. People who counted as overweight or obese by BMI alone, but did not meet the combined body measurement criteria, showed no increase in all-cause mortality. Their risk was even slightly lower than that of people without overweight or obesity. Meanwhile, clinical obesity was linked to a 69 percent higher risk of death from any cause compared with people who had neither excess fat by the combined measures nor organ dysfunction. Preclinical obesity carried a 19 percent higher risk.

In plain terms: the tape measure around your waist and the state of your organs told the researchers more than the bathroom scale did. Fashion people have known for a long time that two bodies with the same weight can wear the same dress very differently. Medicine is now catching up with the fitting room, with far better data.

Preclinical does not mean “on the way to sick”

One of the more reassuring results concerns progression. Most people with preclinical obesity did not develop clinical obesity over the follow-up period. Live Science reports that 77 to 91 percent did not progress within ten years, depending on the cohort. The ten-year incidence of clinical obesity was 21.2 percent in the Chinese cohort, 8.7 percent in the UK Biobank and 22.6 percent in the Spanish data.

The authors also tested a simple universal waist-to-height threshold of 0.50, a rule of thumb often summed up as “keep your waist under half your height.” Applied to everyone, it would sharply increase the number of people classified as obese. Those flagged only by that threshold showed no extra mortality risk, while higher, ethnicity-specific cut-offs separated risk more meaningfully. One size, it turns out, does not fit all here either.

What experts are cautious about

The study is observational and retrospective. It relies on health records. The cohorts also differed in which organ dysfunction markers were available and how often they were checked. The researchers used a statistical method to estimate how much of that dysfunction was attributable to obesity at the population level, but they stress that this cannot establish the cause in an individual patient. They call for prospective studies with clinically validated diagnoses in more diverse populations.

Outside experts welcomed the direction while flagging open questions. Jennifer Hwang of the University of Chicago told Live Science that “we are definitely missing some people who actually need obesity treatment but just did not meet the BMI criteria.” Aayush Visaria of Rutgers Robert Wood Johnson Medical School argued that “there needs to be more consensus on what’s actually defining clinical obesity.” Rubino’s own summary was blunter: BMI, he said, “misses a lot of information.”

There is also a policy dimension. With weight loss drugs in high demand and expensive, the definition of obesity shapes who qualifies for treatment and who pays for it. The authors argue that a preclinical label should not automatically rule out active treatment when it is clinically appropriate. How health systems and insurers translate that into rules is still open. No official guideline has switched over on the basis of this paper.

What this means for you

If your BMI says one thing and your doctor’s broader assessment says another, this study suggests the broader assessment deserves your attention. Waist measurements, blood pressure, blood sugar, joint health and sleep tell a fuller story than height and weight squeezed into one number. That is not permission to ignore BMI or to self-diagnose with a sewing tape. It is a reason to ask for the whole picture.

And there is a cultural shift hiding in the statistics. For decades, BMI has been used as a moral scorecard in diet culture, on magazine covers and in casual comments at family dinners. A definition built on how a body actually functions, rather than how it ranks on a chart from the 1830s, is better medicine. It might also turn out to be a kinder way of talking about bodies.

This article is for general information only and does not replace professional medical advice. If you have questions about your weight or health, please talk to a doctor.