New waist-based criteria suggest many Americans classified as overweight may actually have clinical obesity

A study says that 26% of people who fall into the normal BMI range meet a newer definition of “clinical obesity” when waist size and existing health damage are taken into account.

According to research published in Annals of Internal Medicine, about 26% of U.S. adults with a normal body mass index meet the criteria for clinical obesity under the revised approach, which looks beyond weight alone. Among adults in the overweight BMI range, the figure rises to 50%.

Researchers at Keck Medicine of USC analyzed data from 5,642 nonpregnant adults examined through the National Health and Nutrition Examination Survey between 2021 and 2023. NHANES uses mobile clinics to measure, weigh and draw blood from a nationally representative sample of Americans. The average age of participants in the study was 49.

The report is a short research paper of roughly four pages, and the authors say it is the first national estimate broken down by BMI category.

Two ways to define obesity

BMI is calculated by dividing weight in kilograms by height in meters squared. A BMI below 18.5 is considered underweight, 18.5 to 24.9 is normal, 25 to 29.9 is overweight and 30 or above is obese. The formula dates back to the 1830s and has been the standard screening tool in U.S. clinics since the 1980s, largely because it can be calculated with any scale and height chart.

But BMI does not separate fat from muscle or bone. A very muscular person may have a high BMI without carrying excess fat, while someone with little muscle can have a normal BMI yet still have substantial fat around the organs.

A commission from The Lancet Diabetes & Endocrinology proposed an alternative in early 2025. The group included 58 members, among them clinicians, researchers and people living with obesity.

Its approach starts with the waist. Doctors measure waist circumference, waist-to-hip ratio or waist-to-height ratio, and high values indicate excess body fat. The clinician then checks whether that fat is already causing harm. If a patient has conditions such as high blood pressure, sleep apnea, fatty liver disease or joint pain that limits movement, the diagnosis is “clinical obesity.” If not, the patient is labeled as having “preclinical obesity.” In all, the commission listed 18 qualifying conditions for adults.

Other groups have already tested the framework. In a later analysis, a Mass General Brigham team applied it to more than 300,000 people in the All of Us research program and found that seven in 10 adults qualified. That group, however, skews older and sicker than the general population. The USC researchers chose NHANES because it is designed to represent the U.S. population, then separated the findings by BMI class, something the earlier study did not do.

Clinical obesity increases as BMI rises

After adjusting for age, sex and ethnicity, clinical obesity was found in 26.1% of adults with a normal BMI and 50.3% of those in the overweight range. The share climbed to 65.6% in class I obesity (BMI 30 to 34.9), 77.8% in class II (35 to 39.9) and 85.3% in class III (40 and above).

Across the full sample, 53.8% of adults had clinical obesity and 24.2% had preclinical obesity.

The authors also ran two versions of the numbers for excess body fat alone. Requiring two or three abnormal waist-based measurements, with no BMI condition, identified 78% of adults. Requiring an abnormal BMI plus at least one abnormal waist measurement identified 40.9%.

That 78% figure is close to a number that has circulated since a 2024 forecast from The Lancet, which estimated that about 75% of American adults were overweight or obese by BMI. The USC authors did not compare the two directly. Their own data do show movement between categories: some high-BMI adults had normal waist measurements, and about one in four adults with a normal BMI had abnormal waist readings.

Why the difference matters for care

Rules for prescribing GLP-1 drugs such as semaglutide, along with surgical eligibility and most insurance coverage decisions, still begin with BMI. A patient with a BMI of 24.9, a 40-inch (102-centimeter) waist and fatty liver disease would usually not qualify for medication or surgery. The authors note that such a patient might not even be screened for obesity-related conditions.

Dr. Brian P. Lee, a hepatologist and liver transplant specialist at Keck Medicine of USC and the paper’s corresponding author, said in a statement that BMI reflects total body weight, including muscle and bone, rather than fat. He added that millions of Americans may be missing needed health interventions.

“The good news is that obesity can be treated,” Lee said, pointing to lifestyle changes and medication as ways to reduce excess fat and lower the risk of later complications.

The authors wrote that if the commission’s definition is adopted, screening guidelines and treatment thresholds would need to be revised. They also called for studies on whether treating clinical obesity in people with a normal BMI is safe and effective. Current drug trials, they noted, enrolled patients according to BMI.

Limits of the study

NHANES does not record every condition included in the commission’s list. It does not capture raised pressure inside the skull, atrial fibrillation or chronic joint stiffness, so the authors say their 53.8% estimate for clinical obesity is likely an undercount.

The commission’s framework is also based on expert judgment and has not yet been validated as a diagnostic test. No one has shown that a normal-BMI patient labeled with clinical obesity does better with treatment than without it.

Waist measurement is also more difficult to standardize than it may seem. Clinics use different protocols, two clinicians can place the tape at different points on the same patient and a standard medical tape measure is too short for some people.

The most recent BMI-only estimate comes from a later paper in Circulation, which put obesity at 41% of U.S. adults in 2023. A waist-based national count covering more than one year has not yet been published.

The study appears in Annals of Internal Medicine.

Ways to get involved

Join the Obesity Action Coalition if you want a patient perspective on how obesity is defined and covered. The coalition advocates with insurers and Congress for better access to treatment and offers free education for members.

Read the annual State of Obesity report from Trust for America’s Health if you want to follow the data, since it compiles state-by-state figures and grades policy responses each year.

Support the American Heart Association if you are concerned about related health damage. Its research, funding and guidelines influence how doctors screen for weight-related heart and metabolic disease.

Take part in the National Health and Nutrition Examination Survey if you are selected. Studies like this depend on ordinary Americans agreeing to be weighed, measured and tested.

Use World Obesity Day, observed on March 4 and coordinated by the World Obesity Federation, to share accurate information about body fat, BMI and newer definitions.