We’re finally abandoning BMI for better ways to assess body fat

I was so excited for my first job after university that even the physical assessment – required for my work as a field geologist – came with a certain thrill. That was until the doctor glanced at my chart and told me to lose 10 to 15 pounds. The proof, he said, was my body mass index (BMI), which placed me smack in the middle of the “overweight” category. Though this was supposedly a health concern, he then said I had passed every other test: my heart was strong, my blood markers were pristine, my risk of disease was comfortably below average. I stayed behind when he left the room, suspended between shame and disbelief. I had gone in feeling good about my body, but the doctor was telling me to think differently. I followed his advice through my 20s, chasing a “healthy” BMI through disordered eating. Only later did I find out the doctor was wrong. He had read my chart correctly, but BMI had failed to capture my health.

I’m not the only one who has been mislabelled as unhealthy. BMI, which is calculated by dividing a person’s weight by their height squared to estimate levels of body fat, is deeply embedded in healthcare. Yet this measure distorts health for many, with serious consequences.

BMI cut-offs determine access to knee surgeries, GLP-1 medications, infertility treatment, gender-affirming care, bariatric procedures and more. People outside the “acceptable” range may be denied care, while high-risk patients with “normal” BMIs may be overlooked. After 30 years as the status quo, a consensus is finally emerging that BMI is not appropriate. “There is no logic, no medical coherence to using BMI to define a disease. It’s just not suitable,” says Francesco Rubino at King’s College London.

Alongside that, the hunt is on for better measures. Even more importantly, this shift is forcing us to question the very interplay between health and body size, and completely rethink what a “healthy weight” really looks like.

The rise of BMI

BMI has persisted not because it is biologically precise, but because it is fast, cheap and simple. Developed by mathematician Adolphe Quetelet in the early 19th century as a statistical population tool, it was originally designed to document height and weight averages. As obesity increased in the 1970s, a study suggested it could be used to track this trend over large populations.

With rates of obesity rising further in the following decades, BMI became an increasingly popular research tool. And when the World Health Organization formally recognised obesity as a global epidemic in 1997, BMI was rolled out in healthcare as an assessment tool for individuals. Despite early critiques that this was inappropriate, its convenience meant it spread quickly, and BMI became medicine’s default screening gate.

Obesity is linked with several dangerous conditions, including cardiovascular problems, kidney disease, some cancers, high blood pressure and type 2 diabetes. Clinicians wanted a quick way to estimate when excess fat might become harmful, and BMI seemed to fill that role. The measurement assigns people a simple number, with those under 18.5 classed as underweight, over 25 as overweight and over 30 as obese. But Rubino says a risk-assessment tool for populations should never have been converted into a diagnostic tool for individuals.

A common critique is that BMI doesn’t reflect where fat is stored. It doesn’t acknowledge that men often carry visceral fat in the middle, in and around organs, which is more harmful than the subcutaneous fat on the arms, bottom and thighs that women are more likely to have. Not only that, it doesn’t distinguish between fat and muscle, meaning many athletes can end up being classed as unhealthy. This was the case for me when I underwent that shocking medical exam. I naturally carried more muscle and was an avid trail runner and swimmer. The very thing pushing me towards an “overweight” BMI was what made me fit and strong.

At the other end of the scale, someone can lack enough body fat for regular menstruation – which causes a host of other problems, including bone weakness and breakage, cardiovascular issues and pelvic pain – but be officially classed as “normal” according to their BMI. This isn’t uncommon in female athletes who carry enough muscle to keep them in the “normal” weight range, but don’t have enough fat. BMI is a proxy that can too easily mask true symptoms of illness.

“BMI is a good metric of volume – if a body is a cylinder. But I’m no cylinder. I have hips, I have curves on my body,” says Diana Thomas, a mathematician at the United States Military Academy at West Point in New York. “And so to just put a cylinder on me and say, ‘We’re going to use this body measurement from your outside to tell us what’s on the inside,’ seems pretty limited.”

Five years ago, Rubino had also grown frustrated with BMI. So he created a commission to re-evaluate the definition and diagnosis of obesity, the results of which were finally published last year as part of the Lancet global commission on obesity. The reliance on BMI created a “catch-22”, says Rubino, because it was used to diagnose obesity, which is purportedly a disease, but failed to meaningfully reflect health.

The commission’s report called for a major overhaul. Just as there is a distinction between pre-diabetes and diabetes, the report said obesity should be divided into pre-clinical obesity, when fat levels are a risk factor, and clinical obesity, when it should be considered a stand-alone illness. Crucially, the authors said BMI alone shouldn’t be used to determine obesity status.

Some people can be overweight but have no markers of ill health, which the Lancet commission flagged as another serious shortcoming of BMI. Over the past 20 years, there has been a growing understanding that obesity doesn’t always lead to disease. “Some people who carry some extra fat may actually fare better or may not be affected by it, like Queen Victoria, who lived into her 80s and died from unrelated causes,” says Rubino.

The Lancet’s recommendations were immediately taken up by more than 75 international medical organisations working on obesity, reflecting a seismic shift in the field: BMI is on its way out.

When does weight affect your health?

The connection between weight and health is much more complex than a number on a scale or chart. And with BMI not fit for purpose, researchers are looking for better tools.

We know that the location of fat, rather than the total amount, is a much better predictor of health. People with high visceral fat have more than double the risk of heart disease and greater risks of high blood pressure and type 2 diabetes. They are also three times more likely to develop dementia by their mid-70s to early 80s. That’s why the Lancet commission recommends that healthcare providers determine excess body fat by directly measuring it if possible, with technology like bioimpedance tools that use low-voltage electricity to asses body composition.

Alternatively, they should use a criteria that estimates visceral fat – such as waist circumference, waist-to-hip ratio or waist-to-height ratio – in addition to BMI. Regardless of the tool used, age, gender and ethnicity should all influence the interpretation of results, based on two decades of evidence showing that these factors affect where fat is stored and the associated risks.

The measures can be fairly simple: Sonia Anand at McMaster University in Ontario, Canada, points to the findings of the the landmark INTERHEART study, which first showed that waist-to-hip ratio outperforms BMI as a predictor of heart attacks more than 20 years ago. Research since then has shown it is also a better predictor of mortality.

Another strong alternative to BMI is the weight-adjusted waist index (WWI), which divides waist circumference by the square root of body weight. Rather than estimating mass alone, it captures central fat storage – a pattern closely tied to inflammation, hypertension and cardiometabolic disease – plus it can be quickly calculated, using only a tape measure and scale.

In a study last year of 239 white women, those in the highest WWI quartile had significantly higher levels of visceral fat, higher blood pressure and higher levels of inflammatory markers than those in the lowest quartile, even when their BMIs were similar.

This is particularly meaningful for women, says Naveed Sattar, a professor of cardiometabolic health at the University of Glasgow, UK. Women typically have greater subcutaneous fat storage capacity, and waist circumference in women is a stronger prediction of diabetes than BMI. “Where you put your fat matters,” he says.

Height is also important. Taller people have bigger bones and more muscle, none of which is accounted for by BMI. A little over a decade ago, Thomas decided to create her own metric. As an athlete, she knew of the critiques of BMI, while her background in maths naturally pushed her to question the accuracy of nutrition and health advice. She came to believe health research needed “more math in it”.

Thomas used datasets from more than 7000 people to build a model linking body shape to fat patterns associated with health risk. From that, she developed the body roundness index (BRI) formula, which uses height, waist circumference and weight to model body geometry. Studies have since shown that BRI predicts total and visceral fat more accurately than BMI, waist or hip measures alone, and allows shape to be analysed as a continuum rather than a binary label.

This nuance means that waist-to-hip ratio, WWI and BRI are all more effective at highlighting the metabolic risk factors for conditions like diabetes, cardiovascular disease, high blood pressure and cancers – all of which are linked to stomach fat, not just fat in general. And rather than solely relying on body measurements, physicians can add blood tests to evaluate risk. Testing for liver function and levels of triglycerides and HDL cholesterol is cheap and easy, says Sattar.

These measures have changed my understanding of my own body. My WWI says I’m in the first quartile, the “most healthy”; my waist-to-hip measurements put me in the “low risk level” category; and my BRI says I’m in the healthy zone and my visceral fat levels are “excellent”. Quite a different message than the BMI score.

The fight to do away with BMI

As well as misclassifying individuals like me, evidence accumulated over two decades shows that BMI gives a distorted health reading for whole ethnic groups. BMI was originally devised using metrics based on white people and, though studies exposing these blind spots began years ago, the implications have only started to shape policy recently.

South Asian, Chinese and Black people all get diabetes at higher rates and lower BMIs than white people. The reasons for this disparity aren’t fully understood for Black or Chinese people, but the effect in South Asian populations is linked to their genetic make-up. South Asian people naturally tend to have lower muscle mass and more fat mass than white people at the same BMI, says Sattar, and more rapidly add fat around the centre of the body, especially the liver.

These insights led South Asian countries to be among the first to challenge BMI. By 2022, 10 Asia-Pacific nations including India, Singapore, Sri Lanka and the Philippines created their own BMI thresholds, with lower cut-off points for obesity, and began using additional measures like waist circumference.

Some of the research questioning BMI started with Anand. “Being a South Asian woman, my family history of early-onset heart disease and diabetes being very prevalent, I was very passionate to try and understand our biological differences,” she says. Her first few grant applications to study BMI variations across ethnic groups were deemed unimportant, she says, and rejected. But she persevered. Her work ultimately showed that South Asian and Chinese people experience cardiometabolic risk at significantly lower BMI thresholds than white Europeans, while the measure also fails to predict risk among Black and Inuit populations.

Almost two decades on from her initial idea, Anand’s work helped spur a 2020 clinical practice guideline update in Canada, ensuring both lower BMI cut-offs for South Asian people and the recommendation that non-BMI measures be considered in evaluations of obesity for all groups.

Other countries have also started to adapt. In 2022, the National Institute for Health and Care Excellence, an independent regulatory advice body in England and Wales, issued new guidance encouraging the use of waist-to-height ratio in routine care. Critics of BMI – including Rubino – say it should never be used in isolation, and hope that support for the Lancet commission from so many groups, including the World Obesity Federation and World Health Organization, will put an end to this practice in the UK. In the US, the American Medical Association declared BMI “imperfect” in 2023 and recommended it be used only alongside other health measures.

In the era of weight-loss drugs like Mounjaro and Wegovy, this matters more than ever. Medication for obesity is currently distributed based on BMI cut-offs. In England, GLP-1 treatments are restricted to those with a BMI of 40 and above for Mounjaro or 35 and above for Wegovy (both adjusted for ethnicity) plus multiple obesity-linked conditions. In Japan, the drugs are reserved for those with a BMI above 35, while in the US, the Food and Drug Administration recommends the medications for anyone with a BMI of 30 or higher, although many insurance companies force higher limits.

With 1 billion people globally classed as obese according to BMI, there isn’t capacity to treat everyone, says Rubino. “And by the way, not all of the 1 billion need it.” Relying on BMI to determine both who is considered to have obesity and who gets treatment creates problems at every stage of the treatment process.

Medicine’s understanding of obesity is unlikely to improve unless the field is willing to accept – and even embrace – the complexity of human bodies. “One of the things I always ask [doctors] is, why does everything have to be easy? My body’s complicated. Embrace that complexity!” says Thomas.

For myself and many others whose lives have been affected by BMI’s authority, the shift towards shape and distribution-based measures isn’t about dismissing obesity-related disease. It is about recognising individual biology, ethnicity and lived experience. The emerging science points to a simple truth: health isn’t a single number, and bodies aren’t interchangeable cylinders.