Landmark Nature study: most BMI-defined obesity isn't clinical

In the new study, BMI both overestimated and underestimated obesity in certain groups when checked against the new framework. The error runs in both directions, which makes it harder to patch.

What happened to the 844,000 adults?

The researchers tracked the two categories over roughly a decade and found sharply different trajectories:

• Most people with preclinical obesity did not go on to develop clinical obesity over the decade.
• Those with clinical obesity faced a substantially higher risk of death.
• Only an estimated 22.7% to 31.3% of people who met the BMI criteria for obesity had clinical obesity; the remainder were classified as preclinical.

So the group BMI treats as one homogeneous population actually splits into people whose excess fat is already costing them health, and people whose risk has not yet converted into disease.

What did the researchers actually say?

"This work tells us that the BMI system is not accurate in identifying the health status of an individual," said Francesco Rubino, chair of metabolic and bariatric surgery at King's College London and a senior author of the study, at a press briefing.

Rubino argued that medicine usually defines disease by what clinicians see at the bedside. "Diseases are normally defined in medicine by the way they look at the bedside," he explained. Obesity has historically been defined using measurements that reflect population-level health risks, not the health of an individual patient — a category error he suggests the new framework corrects.

The verdict from the authors is not that BMI should be abandoned, but that it is an inadequate stand-in for diagnosis on its own.

How soon could this change practice?

Not soon, on the evidence so far. An Pan, a professor of epidemiology at Huazhong University of Science and Technology in China and co-lead author, said at the same press event that more work is needed before the new framework could be implemented in clinical practice. Specifically, the team wants to identify the factors that predict which individuals with preclinical obesity will progress to clinical obesity — the group most likely to be misclassified as low-risk today.

If the guidelines were adopted more broadly, Rubino said, care could be tailored to whether excess fat is already impairing health or represents a risk of future disease, which would bring obesity diagnosis in line with how other complex conditions are handled. He acknowledged the cost: "The fact that [physicians] will have to work harder to make a diagnosis is certainly true compared to what we're used to."

That is the trade on offer: a more accurate, more laborious diagnosis in exchange for catching the people BMI currently treats as a single undifferentiated risk category.

The takeaway

The strongest reading of this study is that obesity needs a bedside definition, not a scale reading — and that the new framework, promising as it looks, is still a proposal under test rather than a clinical rule.

The take-home in plain terms: BMI alone overstates how many people are sick from obesity, and the fix will require doctors to measure function, not just size.