Researchers find that BMI is a flawed measure of metabolic health

By Lisa Marie Basile | Fact-checked by Davi Sherman
Published June 22, 2023

Key Takeaways

  • The body mass index (BMI) is a widely used tool for measuring obesity. However, new research sees its ability to accurately detect metabolic health as limited—especially in Asian American and Hispanic individuals.

  • Researchers out of Rutgers Robert Wood Johnson Medical School in New Brunswick, NJ, and Rutgers School of Public Health in Piscataway, NJ, found that many Americans with a “normal” BMI were still technically obese, according to dual-energy x-ray absorptiometry (DEXA) scans.

  • Some physicians believe BMI is a starting point, but MDs should also embrace other tools for understanding patients’ metabolic health.

Researchers out of Rutgers University’s Robert Wood Johnson Medical School in New Brunswick, NJ, and the Rutgers School of Public Health in Piscataway, NJ, shared their findings regarding the limitations of the body mass index (BMI) at ENDO 2023, the Endocrine Society’s annual meeting held from June 15 to 18.[] 

Their key point? “BMI alone may not be sufficient to detect metabolically unhealthy adiposity, especially in Asian Americans and Hispanics,” according to the study’s abstract. 

A brief look at BMI’s background

The BMI scale (originally called Quetelet’s Index) was devised in the 1830s by Lambert Adolphe Jacques Quetelet, a Belgian polymath interested in finding “l’homme moyen,” or “the average man.” Essentially, his index required dividing a person’s body weight (in kilograms or pounds) by their height (in meters or feet) squared—and it was used to inform epidemiological studies.[][]   

It wasn’t until 1972 that Quetelet’s Index was rebranded as “BMI'' by Ancel Keys, an American physiologist. And although BMI has since become an industry standard, it has failed to tell a full picture of health, according to research published in the Journal of the History of the Behavioral Sciences: “The epidemiological community accepted BMI despite Keys's failure to demonstrate that either it or adiposity (body fat content), were superior as predictors of heart disease.”[]

Today, BMI is widely used—but often under scrutiny. To determine the “extent of discordance between BMI, true adiposity, and body fat distribution,” the researchers set out to describe the racial differences in BMI and dual-energy x-ray absorptiometry (DEXA) scan–based adiposity measures by looking at a wide range of adults. 

The researchers looked at 9,784 participants (the average age was 39, 41% were female, and 61% were non-Hispanic white, or NHW). To find participants, the team pulled from the National Health and Nutrition Examination Survey (from people who had whole-body DEXA scans available). BMI was categorized into ethnicity-specific categories, with different BMI ranges for Asian and non-Asian individuals. This is because some experts believe that international criteria for obesity cannot be applied to Asian populations.[] For non-Asian individuals, the researchers used the following measurements: 

  • Underweight = less than 18.5

  • Normal = 18.5-24.9

  • Overweight = 25-29.9

  • Obese = more than or equal to 30. 

For Asian individuals, the researchers used the following measurements:

  • Underweight = less than 18.5

  • Normal = 18.5-22.9

  • Overweight = 23-27.4

  • Obese = 27.5

Key research findings

The researchers found that a high percentage of Americans with “normal” BMI were still technically obese, according to DEXA scans—underscoring the fact that BMI is not a full or accurate picture of metabolic health. 

Specifically, the researchers found that about 36% of participants had a BMI over 30—the definition of obesity. 74% had obesity per body fat percentage. Specifically, among adults with “normal” BMI, 44% of NHW, 27% of non-Hispanic Black, 49% of Hispanic, and 49% of Asian individuals had obesity per body fat percentage. 

In summary, “Nearly 3 in 4 young-to-middle-aged U.S. adults have obesity via total body fat percentage estimated from DEXA scans,” the authors write. “Despite ethnicity-specific BMI cutoffs, normal BMI Asian Americans, as well as Hispanics, were more likely to have obesity and more likely to have a greater proportion of abdominal fat than NHW. On the other hand, NHB had significantly lower likelihood of obesity at normal/overweight BMI ranges and lower proportion of abdominal fat. These suggest that BMI alone may not be sufficient to detect metabolically unhealthy adiposity, especially in Asian Americans and Hispanics,” the researchers said.

MDs weigh in on the limitations and benefits of BMI

Scott Cunneen, MD, FACS, FASMBS, Director of Bariatric Surgery at Cedars-SinaiMedical Center in Los Angeles, also says that BMI should be seen as a starting point, not the be-all and end-all.

BMI’s limitations are many. For one, it cannot differentiate between muscle and fat, Cunneen says. If a patient has “a lot of muscle mass and very little body fat, BMI is still going to say that [they’re] overweight when, in fact, [they] may not actually belong in that category.” Additionally, BMI doesn’t indicate where an individual’s body fat is stored. “Visceral fat or central fat, that is, fat that is stored in the organs, is more dangerous and is usually more predictive of health-related problems,” he adds.

According to David Cutler, MD, Family Medicine Physician at Providence Saint John’s Health Center in Santa Monica, CA, BMI also “cannot consider other important health factors, such as blood pressure, cholesterol levels, or blood sugar levels—which are essential for assessing overall health and disease risk.”. “Individuals with similar BMIs can have different body compositions, fitness levels, and overall health statuses,” he adds.

MDs should focus not only on BMI but also family history, body shape, muscle mass percentage, and fitness level, Cunneen suggests. 

Aayush Visaria, MD, MPH, an internal medicine resident at  Rutgers Robert Wood Johnson Medical School in New Brunswick, NJ, and co-author of the research, says, “Waist circumference is a fairly well-established measure of adiposity and is also a proxy for visceral adiposity.”

There are, however, still some instances in which BMI can be helpful—to a point. Visaria says its value lies in its accessibility: “Despite its limitations, [BMI] does have strong predictive value at the population level, especially at very high or very low BMI ranges. It also….[only requires] a simple calculation of height and weight. This makes it much more reproducible, scalable, and trackable,” he says.

Cutler also adds that while BMI shouldn't be used as a stand-alone measure of obesity, it’ll likely be hard to replace, especially as it’s so widespread. “More expensive and cumbersome tools like the DEXA scan are unlikely to gain widespread acceptance,” he says. However, he notes, “[BMI] should be interpreted alongside other measurements and clinical evaluations to obtain a more comprehensive understanding of an individual's health status and obesity risk.”

Cunneen also says, “The gold standard might be to weigh [the patient] and then measure [their] displacement of water so that you know their volume, their density.” “But obviously having to jump into a swimming pool and judg[e] how much water is displaced isn’t going to happen. It's something that you can use for scientific research, but it will never be used on a daily basis.”

Not just MDs who continue to turn to BMI, Visaria says, but also “public health and policy folks, researchers, and fitness people.” However, he thinks this reliance on BMI will eventually change, and the shift is already taking place, with more health-measuring options coming down the pipeline. 

“There are many new devices, wearables, and scales that use bioimpedance technology to estimate body fat percentage,” Visaria says. “Although more research needs to be done to validate the various scales and wearables in the market, I suspect in the next 10 years that these types of devices will be commonplace in doctor's offices,” he adds.

The importance of taking culture and ethnicity into consideration around patient health Visaria’s team’s research underscores that MDs should be aware that racial and ethnic differences in obesity exist. “Race/ethnicity can be a surrogate for a multifactorial set of values, lifestyle habits, and genetics,” Visaria says. “I think it's important for providers to be aware that there exist racial/ethnic differences in obesity, and thus to use clinical judgment when applying many of our clinical guidelines—which unfortunately may not be validated in all minority populations.”

Visaria adds that MDs should consider cultural context when counseling patients: “There are many ongoing studies trying to decipher what it is about race and ethnicity that drives the differences. Is it more so cultural and lifestyle habits, or is it primarily genetics? Either way, we know that some groups are at higher risk than others and thus may benefit from earlier screening in order to reduce disparities.”

Share with emailShare to FacebookShare to LinkedInShare to Twitter