What’s the deal with BMI, aka body mass index?

It takes mere minutes to calculate your body mass index, more commonly known as BMI, by using your height and weight. With countless online charts, you can find out not only what your BMI is, but which of its categorical classifications you fall into: underweight, normal, overweight or obese.

For decades, BMI has been used to stratify people for research and clinical care. A particular BMI can also be used as a cutoff for certain treatments: People with too high a BMI may not qualify for an organ transplant while people whose BMI is not high enough will not receive weight loss drugs or surgery. In other cases, it is used as rationale for recommending weight loss, with clinicians encouraging people to aim for a BMI in the normal range.

Recently, however, some researchers and clinicians have been questioning BMI's value. The number does not capture a person's muscle mass; where on their body fat is stored; or how their race, ethnicity and gender affect the complex relationship between their body composition and health risks. Other measures of body composition have entered the discussion as BMI alternatives.

In the face of this criticism, we asked Stanford Medicine scientists and clinicians what the public should know about BMI and how they think the number should be used -- or not used. They said that its simplicity and years of use mean it is likely not going away anytime soon, but that any conclusions drawn from BMI should always be put in the context of someone's broader health.

The background

Nearly 200 years ago, a Belgian scientist developed a simple way to capture what he dubbed the "socially ideal male body." His formula -- a person's body mass divided by the square of their height -- later became known as BMI.

Throughout the 20th century, scientists and insurance companies created BMI tables to capture the range of normal BMI values and categorize people's risks of developing cardiovascular disease and diabetes based on these values. In general, the higher a person's BMI, the higher their risk of these diseases. However, most of the early studies of BMI that were used to decide the cutoff numbers between risk categories included only non-Hispanic white men. 

"One of the problems with BMI is that it does not translate well across different races, ethnicities and genders," said Stanford Medicine's Dan Azagury, MD, an associate professor of surgery who serves as medical director for the Stanford Lifestyle and Weight Management Center.

The crux of the problem, Azagury said, is that poor health outcomes are more closely associated with levels of deep abdominal fat -- fat that surrounds internal organs like the liver and stomach -- than with a person's body size. A powerlifter with large muscles may have a BMI putting them in the obese category despite low levels of abdominal fat. Women, compared with men, tend to store more fat in their hips and thighs rather than their abdominal region -- yet BMI is not adjusted for gender.

Moreover, ethnicity plays a role in body composition and disease risk. For instance, people of Polynesian descent with high BMI levels tend to have less fat on their bodies than Europeans with the same BMI. People of Japanese ancestry, on the other hand, have higher rates of metabolic disease at lower BMIs than other ethnic groups.

Despite these problems with BMI, the number is used to determine eligibility and insurance coverage for everything from weight loss medications to fertility treatments and joint replacement surgeries.

The debate

Because BMI is not a perfect measure of someone's health, some scientists and organizations are moving away from using BMI in certain clinical settings. In 2023, the American Medical Association released a statement emphasizing that BMI is "an imperfect way to measure body fat in multiple groups given that it does not account for differences across race/ethnic groups, sexes, genders, and age-span." The organization recommended that BMI be used only in conjunction with other measures of disease risk such as direct measurements of visceral fat or body composition.

BMI is an imperfect way to measure body fat in multiple groups given that it does not account for differences across race/ethnic groups, sexes, genders, and age-span.

American Medical Association

However, many clinicians argue that BMI is so engrained in the system -- and so easy to calculate -- that moving completely away from the number will be difficult.

"Absolutely everything relies on BMI right now," Azagury said. "Health screening, weight guidelines, clinical trial endpoints, insurance reimbursements -- everything."

Moreover, many suggested alternatives to BMI -- such as body roundness index (BRI), which involves measuring waist circumference, or body composition scans -- are harder for people to measure on their own at home.

"We already have trouble getting people in the door for weight loss interventions, and we have this relatively easy screening method with BMI," Azagury said. "I worry that if we make the guidelines more complicated, we'll have even fewer people who are aware they should seek treatment."

BRI, like BMI, also has its own shortcomings when it comes to capturing the diversity of body compositions and associated health risks for the entire human population.

"I would never advise a patient or physician to focus on one variable, whether that is BMI or BRI," said Yulia Zak, MD, a clinical assistant professor of surgery and colleague of Azagury in the Bariatric and Minimally Invasive Surgery program.

What the science says

Stanford Medicine statistician Maya Mathur, PhD, had long heard that being overweight decreases a person's lifespan. But when she started looking into the numbers, she realized that there wasn't as strong a link as she had assumed.

"With something like BMI, it is incredibly hard to tease apart causation and correlation," Mathur said. "You can assign people to weight loss interventions and test what happens, but you can never assign someone to a particular BMI."

Mathur was especially surprised to discover that many previous studies on BMI and mortality did not control for factors like diet and exercise. This means that a poor diet or lack of exercise could be associated with both a high BMI and an increased mortality risk -- but without the high BMI directly causing the increased mortality.

In a 2022 study, Mathur reported that so much previous research had this oversight that, when you looked at it all together, there was actually no robust evidence for an association between having an overweight BMI (a BMI between 25 and 30) and increased mortality.

"We really don't have statistically sound evidence that having an overweight BMI is bad for mortality," Mathur said. "It might actually be good for mortality; it's quite hard to tease apart."

We really don't have statistically sound evidence that having an overweight BMI is bad for mortality. It might actually be good for mortality; it's quite hard to tease apart.

Maya Mathur

Having a BMI in the obese category, on the other hand, is more strongly associated with increased mortality, she added. A BMI of more than 30 has been associated with cardiovascular disease, diabetes, osteoarthritis, cancer and other chronic conditions.

Azagury -- who has found associations between the BMI levels of patients who have obesity and their health outcomes after weight loss surgery -- pointed out that BMI is a risk factor for health, not a direct measure of health at any given time.

"Because the prevalence of obesity is so high right now, you can certainly find people who have obesity and no other disease," he said. "But the vast majority of people who have a BMI in the obese range will likely develop some of these conditions if they don't already have them."

---

---

Data is still needed to show that alternative measures of body composition, like BRI, are better at predicting health risks. One study this year found that especially high and low BRI numbers are associated with increased mortality but did not compare BMI and BRI head-to-head. Zak admits that BRI -- because it does a better job of gauging whether someone's fat is stored around their abdomen -- could identify some people prone to poor health outcomes that BMI would miss. But she cautions against abandoning BMI.

"We have so many years of data backing up associations between obesity and health," she said. "Sure, there are some whose risk of disease is overestimated or underestimated based on BMI but at a population level it is still catching most people."

Where does this leave us?

Most clinicians are settling on the idea that BMI is one imperfect measure of disease risk that should be used in the context of someone's broader health.

"I don't think it's necessary that BMI be completely thrown out," Mathur said. "But we need to be more careful about how it's being used in clinical practice."

Azagury, for instance, says that the Stanford Lifestyle and Weight Management Center has changed its BMI cutoffs for Asian people seeking treatment, and that they often try to pair BMI with other measures of health.

"We use BMI to triage patients for different weight loss interventions and to comply with insurance requirements," he said. "But when we're following a patient to see how they're doing, we look more closely at how other things -- like blood pressure and cholesterol levels -- are changing."

Mathur said public health messaging on BMI, emphasizing its weaknesses and its role as one risk factor among many that shape health, needs to change. She encourages patients to question their clinician's use of BMI -- particularly if they are in the overweight, rather than obese, category.

"If your doctor raises BMI as a health risk, ask them what the evidence is for any recommendation they are making," she said.

If your doctor raises BMI as a health risk, ask them what the evidence is for any recommendation they are making.

Maya Mathur

But Azagury, in his role managing the weight management center, has a different message -- that people be aware of their BMI and body composition, and make lifestyle changes or seek out medical help if they have obesity.

"Even if you don't love BMI as a single measure of health, treating obesity can have an incredibly beneficial impact on many aspects of your health and on your life expectancy," he said. "Ultimately, I think there will be blood tests that can let us screen people for disease risk more accurately. Twenty years from now we might look back and be shocked that we based so much risk assessment on this very crude number."

Photo illustration by Emily Moskal

---

---