Body mass index (BMI) is a rough proxy for body composition widely used to predict a person’s risk of poor health outcomes, including early death — but BMI is an unreliable metric that does not actually predict risk of death, new research suggests.
The study, published June 24 in the journal Annals of Family Medicine, found that BMI could not reliably predict mortality risk from any cause. However, body fat percentage, an alternative measure of body composition, was found to be far more accurate, showing a clear correlation to individuals’ risk of death.
“Ultimately, the current definition of obesity needs to change to one that is based on body fat percentage,” said senior study author Dr. Frank Orlando, the medical director of UF Health Family Medicine.
That could change how anti-obesity medications, such as Ozempic-style drugs, are prescribed, Orlando told Live Science in an email, as many people who might be at risk due to high body fat do not qualify for treatments based on BMI alone.
An imperfect measure
There are several ways to assess body composition, though none are as cheap and easy as BMI, which uses a person’s height and weight to calculate an indirect measure of body fat. It categorizes people as having underweight, normal weight, overweight or obesity.
“The assumption is that if you weigh over a certain amount for your height, you’re potentially carrying around excess body fat,” said Adam Collins, a professor of nutrition at the University of Surrey in the U.K. who was not involved in the new study. “By association, that excess body fat is increasing your risk of disease,” Collins told Live Science.
Related: People’s mental health often improves after weight-loss surgery. A study pinpoints the real reason why.
Studies have tied BMIs over 25 to an increased risk of various conditions, including type 2 diabetes and heart disease. However, past studies have found an inconsistent relationship between BMI and early death.
And BMI has recently been acknowledged as an “imperfect” measure of fat by medical organizations, including the American Medical Association. Experts argue that the metric can misclassify people who have a lot of muscle as having overweight or obesity, even if they have low or normal levels of body fat, for instance. BMI categories have also been primarily based on non-Hispanic white populations, which means they may not reflect the differing levels of risk that excess fat poses across all patients.
But despite these limitations, BMI continues to be used because there are no well-tested, inexpensive alternatives. In their new study, the researchers wanted to see how BMI stacked up against different body composition measures in predicting people’s risk of death over the next 15 years.
They used data from a representative sample of 4,252 U.S. adults who were between 20 and 49 years old at the start of monitoring. The team pitted BMI against two metrics: body fat percentage and waist circumference. Having a high waist circumference was associated with higher odds of dying from heart disease, but not from all causes, the study found.
To calculate body fat percentage, the researchers used a device that sends a small electrical current around the body and measures the amount of resistance it encounters in different tissues. As tissues like fat and muscle have different water contents, they conduct electricity in different ways; namely, the current moves more easily through muscle. This assessment — called bioelectrical impedance analysis (BIA) — estimates overall body composition, namely fat-free mass and body fat percentage.
Over the 15-year period, those with a high body fat percentage — over 27% in males and 44% in females — were 78% more likely to die than those with lower body fat levels.
What’s next?
BIA isn’t a perfect measure, either. “It’s not just about how much body fat you have in total, it’s about where that fat is,” Collins said. “If it’s just under the skin, subcutaneous fat, it is not so much of a problem. It’s that intra-abdominal fat, the fat around or inside your organs, where you start to get real issues of metabolic disease.”
Anything that replaces BMI will need to be as accessible as BMI, which can be calculated by patients themselves. BIA devices are becoming cheaper and more accessible to primary care practices in the U.S., he added. Bioelectrical impedance is also used in smart scales that can be bought for use at home.
Until now, we have based our definition of obesity on BMI because the measure is easy and cheap, Orlando said. “But now we have a study like ours showing that BIA is accurate and reliable when compared to BMI.”
Access to medicines that treat obesity — such as Wegovy, which has the same active ingredient as the diabetes treatment Ozempic — is available to people with a BMI in the obese category, as well as people with overweight who also have additional risk factors, like diabetes. Yet Orlando says he sees patients who have a BMI below the BMI cutoff that still have a high percentage body fat.
“Our study proves why any cut off for anti-obesity medicine needs to [change to] body fat percent because we are otherwise missing patients who have normal BMI but high body fat,” Orlando argued. “Guidelines need to be updated so that these patients get the medicines they need.”
This article is for informational purposes only and is not meant to offer medical advice.