Nature study finds millions of obesity diagnoses would flip if a health-based definition replaced BMI

A study published in Nature finds that replacing body mass index with an alternate, health-based definition of obesity would reclassify millions of people, some losing an obesity diagnosis and others gaining one. The analysis applies a framework advanced last year by a Lancet commission, which argued that obesity should be identified when excess body fat is already harming health or is likely to drive serious future disease, not by a height-and-weight ratio alone. The unlocked news account does not name the cohort or the exact number behind the word millions.

The available account describes a reclassification analysis, not a treatment trial. Investigators appear to have applied the proposed definition in place of standard BMI cutoffs and counted who would gain or lose an obesity label. BMI still governs regulatory decisions, insurance coverage, and access to treatments, including GLP-1 weight-loss drugs. The authors argue that tying the diagnosis to present harm and future risk could improve care. Absolute counts, confidence intervals, event rates, and funding are not in the unlocked summary, so no relative figure can be checked against a base rate here.

BMI was built as a population screening ratio, not a diagnosis of disease. It tracks adiposity and risk only roughly: it can miss harmful visceral fat at a normal BMI and can label muscular people as ill. The Lancet framework separates clinical obesity, excess fat plus impaired organ or tissue function, from preclinical obesity, excess fat without current dysfunction. That is a conceptual correction with large downstream effects on drug and surgery coverage, who is offered intensive care, and who is told they are sick. Relabeling millions is not the same as showing fewer heart attacks, less disability, or longer life.

A reclassification study can show how many diagnoses would move. It cannot show that the new label improves care, that organ-function checks are reliable in ordinary clinics, or that drug benefit differs by group. The next study should compare heart attacks, diabetes, and death among people whose BMI and clinical labels agree and those whose labels conflict, and test treatment effects in both groups. Until then, BMI is likely to remain the cutoff payers use, even if some clinics add a second exam. Switching coverage rules would be a policy decision, not a result this design can deliver.