Study shows BMI’s weakness as a predictor of future health

A new University of Florida Health study shows that body mass index, or BMI—commonly used to measure obesity and health risk—fails in predicting the risk of future death, suggesting the calculation is deeply flawed.

The study, published June 24 in the The Annals of Family Medicine, concludes that a direct measure of body fat using a relatively inexpensive device that notes the resistance of body tissue to a small electrical current is far more accurate.

This method, bioelectrical impedance analysis, is increasingly found in medical practices.

Individuals who have high body fat as measured directly by bioelectric impedance are 78% more likely to die of any cause than those with healthy body fat levels during the 15 years they were tracked, the study found. They are also about three-and-a-half times more likely to die of heart disease.

By contrast, BMI proved entirely unreliable. The study showed no statistically significant association with 15-year mortality risk from any cause, including heart disease, when BMI is used.

BMI and bioelectrical impedance data were analyzed in a way that removes the effects of age, race and income.

Researchers in the UF College of Medicine’s Department of Community Health and Family Medicine who conducted the study urged physicians to move away from BMI as the standard body composition measure.

The study’s authors likened their research to a test between two competing means of measuring body fat, much like a product comparison.

“This study is a game-changer,” said Arch Mainous, Ph.D., lead author of the study and a professor and vice chair of research in the department. “This is the ultimate Coke versus Pepsi test. And BMI failed.”

The study pulled data from the National Health and Nutrition Examination Survey, which is linked to the national death index. The analysis involved 4,252 people, a representative sample of the U.S. population.

The study’s senior author, Frank Orlando, M.D., medical director of UF Health Family Medicine—Springhill, said BMI’s use as a “vital sign” of health isn’t supported by good evidence.

“I’m a family physician, and on a regular basis, we’re faced with patients who have diabetes, heart disease, obesity and other conditions that are related to obesity,” said Orlando, an associate professor in the department.

“One of the routine measures we take alongside traditional vital signs is BMI. We use BMI to sort of screen for a person having an issue with their body composition, but it’s not accurate for everyone like vital signs are,” he added.

BMI is calculated using an individual’s height and weight—the formula is weight divided by height, squared. But the measure cannot distinguish between muscle and fat mass, only providing an indirect indication.

“For example, people who are bodybuilders can really elevate their body mass index,” Orlando said. “But they’re healthy even with a BMI indicating they’re obese.”

Mainous said BMI gained popularity because of its simplicity. It requires no equipment, just a patient’s weight and height.

“It’s very easy to get,” he said. “The problem was, we didn’t have a practical alternative.”

Devices that measure bioelectrical impedance, however, are increasingly affordable, with many models under $300. They are now found in some primary care practices. In fact, Orlando noted, one is available at the UF Health Springhill family practice.

Mainous and Orlando said some researchers have criticized the devices for not being as accurate as a DEXA scan, or a dual-energy X-ray absorptiometry. That scan uses low-dose X-rays to measure bone, fat and lean muscle mass. DEXA, of course, requires expensive equipment, although it is considered the gold standard to measure body fat.

“If you talk to obesity researchers, they’re going to say you have to use the DEXA scan because it’s the most accurate,” Mainous said. “And that’s probably true. But it’s never going to be viable in a doctor’s office or family practice.”

Bioelectrical impedance, Mainous said, doesn’t need to be superior to DEXA. It just needs to be better than BMI, which the study shows that it clearly is.

Problems with BMI have been well-documented by researchers. For example, the method can classify some people with a “normal” BMI even when they have a high percentage of body fat, which is known as normal-weight obesity. Body fat percentage can also vary by age.

BMI dates to the 19th century. Adolphe Quetelet, a Belgian statistician, astronomer and mathematician, created “Quetelet’s index” in 1832. He would no doubt fail to recognize the uses being put to his index in the modern world.

Quetelet’s interest instead focused on a quixotic statistical quest to describe the characteristics of the “normal man.” Later generations, starting with insurance actuaries, began using his index to assess health risk.

Today, BMI categorizes an individual as either underweight (BMI under 18.5), normal weight (18.5 to 24.9), overweight (25 to 29.9) or obese (30 and above). The formula does not change based on gender, another flaw cited by critics.

“BMI is just so ingrained in how we think about body fat,” Mainous said. “I think the study shows it’s time to go to an alternative that is now proven to be far better at the job.”

A new University of Florida Health study shows that body mass index, or BMI—commonly used to measure obesity and health risk—fails in predicting the risk of future death, suggesting the calculation is deeply flawed.

The study, published June 24 in the The Annals of Family Medicine, concludes that a direct measure of body fat using a relatively inexpensive device that notes the resistance of body tissue to a small electrical current is far more accurate.

This method, bioelectrical impedance analysis, is increasingly found in medical practices.

Individuals who have high body fat as measured directly by bioelectric impedance are 78% more likely to die of any cause than those with healthy body fat levels during the 15 years they were tracked, the study found. They are also about three-and-a-half times more likely to die of heart disease.

By contrast, BMI proved entirely unreliable. The study showed no statistically significant association with 15-year mortality risk from any cause, including heart disease, when BMI is used.

BMI and bioelectrical impedance data were analyzed in a way that removes the effects of age, race and income.

Researchers in the UF College of Medicine’s Department of Community Health and Family Medicine who conducted the study urged physicians to move away from BMI as the standard body composition measure.

The study’s authors likened their research to a test between two competing means of measuring body fat, much like a product comparison.

“This study is a game-changer,” said Arch Mainous, Ph.D., lead author of the study and a professor and vice chair of research in the department. “This is the ultimate Coke versus Pepsi test. And BMI failed.”

The study pulled data from the National Health and Nutrition Examination Survey, which is linked to the national death index. The analysis involved 4,252 people, a representative sample of the U.S. population.

The study’s senior author, Frank Orlando, M.D., medical director of UF Health Family Medicine—Springhill, said BMI’s use as a “vital sign” of health isn’t supported by good evidence.

“I’m a family physician, and on a regular basis, we’re faced with patients who have diabetes, heart disease, obesity and other conditions that are related to obesity,” said Orlando, an associate professor in the department.

“One of the routine measures we take alongside traditional vital signs is BMI. We use BMI to sort of screen for a person having an issue with their body composition, but it’s not accurate for everyone like vital signs are,” he added.

BMI is calculated using an individual’s height and weight—the formula is weight divided by height, squared. But the measure cannot distinguish between muscle and fat mass, only providing an indirect indication.

“For example, people who are bodybuilders can really elevate their body mass index,” Orlando said. “But they’re healthy even with a BMI indicating they’re obese.”

Mainous said BMI gained popularity because of its simplicity. It requires no equipment, just a patient’s weight and height.

“It’s very easy to get,” he said. “The problem was, we didn’t have a practical alternative.”

Devices that measure bioelectrical impedance, however, are increasingly affordable, with many models under $300. They are now found in some primary care practices. In fact, Orlando noted, one is available at the UF Health Springhill family practice.

Mainous and Orlando said some researchers have criticized the devices for not being as accurate as a DEXA scan, or a dual-energy X-ray absorptiometry. That scan uses low-dose X-rays to measure bone, fat and lean muscle mass. DEXA, of course, requires expensive equipment, although it is considered the gold standard to measure body fat.

“If you talk to obesity researchers, they’re going to say you have to use the DEXA scan because it’s the most accurate,” Mainous said. “And that’s probably true. But it’s never going to be viable in a doctor’s office or family practice.”

Bioelectrical impedance, Mainous said, doesn’t need to be superior to DEXA. It just needs to be better than BMI, which the study shows that it clearly is.

Problems with BMI have been well-documented by researchers. For example, the method can classify some people with a “normal” BMI even when they have a high percentage of body fat, which is known as normal-weight obesity. Body fat percentage can also vary by age.

BMI dates to the 19th century. Adolphe Quetelet, a Belgian statistician, astronomer and mathematician, created “Quetelet’s index” in 1832. He would no doubt fail to recognize the uses being put to his index in the modern world.

Quetelet’s interest instead focused on a quixotic statistical quest to describe the characteristics of the “normal man.” Later generations, starting with insurance actuaries, began using his index to assess health risk.

Today, BMI categorizes an individual as either underweight (BMI under 18.5), normal weight (18.5 to 24.9), overweight (25 to 29.9) or obese (30 and above). The formula does not change based on gender, another flaw cited by critics.

“BMI is just so ingrained in how we think about body fat,” Mainous said. “I think the study shows it’s time to go to an alternative that is now proven to be far better at the job.”