The Shape of Health

The Shape of Health

Researchers examine alternatives to the BMI tool and its outdated assumptions.
By Dara Chadwick

If you’ve ever sorted through an old toolbox or your grandmother’s kitchen drawer, you’ve likely encountered tools that—at first glance, anyway—seem to have outlived their usefulness. Once at the leading edge of advancement, these items may seem quaint as our thinking has evolved.

For some researchers who study obesity and metabolic disease, body mass index (BMI) falls into that category. This simple equation, BMI = weight (lb)/height (in)2 x 703, is a tool that’s long been used to make assumptions about a person’s health.

But those assumptions are often wrong, according to Fatima Cody Stanford, MD, MPH, MPA, MBA, a fellowship-trained obesity medicine physician-scientist and associate professor of medicine and pediatrics at Massachusetts General Hospital and Harvard Medical School in Boston. “BMI is a screening tool, not a diagnostic tool,” she says. “We—meaning the CDC and World Health Organization—use it as a health measure, but BMI doesn’t tell me how healthy or unhealthy one person is. It gives me a sense of their weight relative to others based on this measure, but the metric doesn’t come from science.”

In fact, BMI originated in the statistical theories of a 19th century Belgian mathematician who noted the relationship between a person’s weight and height. Adolphe Quetelet created an equation that calculated a person’s ideal weight based on height. Known as the Quetelet index, this measurement scale was renamed the body mass index in 1972 by Minnesota physiologist Ancel Keys, PhD, according to a June 2014 International Journal of Epidemiology commentary, “Origins and Evolution of Body Mass Index (BMI): Continuing Saga.”

As interest in obesity and its role as a contributor to disease deepened, researchers examined the relationship between BMI and mortality, according to Steven Heymsfield, MD, professor of metabolism and body composition at Pennington Biomedical Research Center at Louisiana State University in Baton Rouge. The BMI chart became a tool to quickly categorize a person’s weight status. They “sliced up the curve to underweight, normal weight, overweight and obese,” Heymsfield says. “But I don’t think there was a super mathematical approach to saying where these cut points are.”

Not the Whole Story

Insurers have long used height and weight ratios to predict beneficiary health, Stanford notes. She cites the Metropolitan Life Insurance Company actuarial tables used in the 1940s to determine your risk of dying if you were a white man or woman. But these ratios don’t tell the whole story, she says.

Stanford and her colleagues analyzed BMI cutoffs and sought to redefine BMI thresholds by sex, race and ethnicity. They examined current National Health and Nutrition Examination Survey data in relation to hypertension, diabetes and dyslipidemia. Their research, published in the February 2019 issue of Mayo Clinic Proceedings, found that when obesity is correlated with these metabolic risk factors, BMI cutoffs that define obesity change for specific racial, ethnic and sex subgroups. 

“What was particularly interesting is that for most groups, particularly for all men, the cutoff for BMI to determine obesity drops below 30 across the board,” Stanford says. “And for Black women, the group known to have the highest rates of overweight and obesity, it shifted up across the board to between 31 and 33. That shows that different gender, racial and ethnic groups may need different cutoffs if we try to use BMI as a measure for obesity.”

Stanford, who is Black, says the historical focus on white bodies was the impetus for her wanting to redraw the charts, which “did not capture racial minorities like myself.” She notes that BMI and other height and weight charts didn’t capture differences in Asian, Latino or Native populations either. 

Age and athleticism are additional factors that complicate the BMI story. A muscular athlete may have a high BMI with greater lean muscle mass, while an older adult may have a lower BMI, but greater fat mass. “At the same BMI, an old person will have higher body fat than a young person,” Heymsfield says. “Let’s say you maintain the same weight and height between age 20 and 80. At 80, you’ll have a higher percentage of fat and a lower percentage of muscle mass.”

Where Do You Carry Fat?

Michael Czech, PhD, is the Isadore and Fannie Foxman Chair of Medical Research in the Program in Molecular Medicine at the UMass Chan Medical School in Worcester, Massachusetts. He studies molecular changes that contribute to the development of diseases related to obesity. He notes that BMI is useful in the broadest sense. 

“Most scientists and physicians agree that BMI cannot be directly linked to metabolic health,” Czech says, adding that in his lab, researchers use a variety of methodologies to investigate the molecular factors and elements that are causative in the transition between obesity and type 2 diabetes.

“The BMI value itself is only a very rough estimate of adiposity,” he says. “The heterogeneity of adipose tissue confounds the use of BMI strictly as an indicator. Taking any individual patient, getting a BMI value and trying to make an association with metabolic health just doesn’t work anymore.”

Yet sources agree that BMI remains useful as a screening tool. “We know that the higher one’s weight gets, there’s a higher risk for death,” Stanford says. “But it doesn’t tell me about one’s health, and I think that’s really important.”
Some of Stanford’s patients have BMIs that indicate they have severe obesity. “You have to look beneath the surface. Maybe they weigh 300 pounds but they have better overall health than patients at 150 pounds with a BMI of 24.8,” she says. “So yes, screen. But we must look at that person’s 300 pounds and see what’s going on.”

She notes that measures such as lipid profile, glucose levels and evidence of sleep apnea offer more predictive health information. “They may be very healthy,” she says of people with higher BMI. “They just happen to carry excess adipose tissue.”

Racial, ethnic and gender differences play a role here, too. “It’s really the distribution of adipose tissue that’s most closely correlated to a disease,” Czech says. He notes differences in the relationship between BMI and glucose intolerance in many parts of the world. “BMIs in some Asian nations and in India, for example, are much lower associated with Type 2 diabetes than in European populations,” he says. “That’s likely related to the fact that the distribution of adipose tissue throughout the body is different in different regions of the world.”

Another interesting note: Some people with the highest BMIs seem to be more protected from glucose intolerance and diabetes than people with lower BMIs, according to Czech. Again, fat distribution matters.

Czech says it’s important to consider a person’s phenotype. “At this point, at the physician’s office, there’s not a lot of sophistication that goes into this. But human body phenotype is important to look at carefully because it can give clues and hints about adipose distribution.”

“The hypothesis is that if you are able to expand your subcutaneous adipose tissue as you gain weight rather than your visceral adipose tissue, your subcutaneous adipose tissue is able to sequester the fat that’s being taken in into subcutaneous depots,” Czech says. How is this protective against disease? Fatty acids, triglycerides and other fats taken in with new calories are stored away in these depots. This prevents them from circulating in the blood or being deposited in the liver, beta cells, islets and other tissues.

“Healthy adipose tissue can vigorously and robustly store fat and keep it from circulating in your blood and other tissue,” Czech says. “This can protect other tissues from insulin resistance because we know that fatty acids can, under certain conditions, be disruptive. That’s the canonical answer. However, it may well be more complicated than that. There may be additional mechanisms at play.”

There’s some evidence that adipocytes in subcutaneous adipose tissue—particularly beige and brown adipose tissues—secrete hormones that act on distant tissues in liver and muscle.

“These factors are in two classes—beneficial factors and deleterious factors,” Czech says. “And maybe individuals who are protected with higher amounts of subcutaneous fat secrete more beneficial factors that operate on distant tissue in a positive way. So that’s another way of thinking about how adipose tissue can be a master regulator of whole-body health.”

Czech also raises the theory—less discussed and with less data to support it—that adipose tissue may be able to communicate with other cells and tissues in the body through modulating the overall neuronal system. “That’s a possible way one could imagine adipose tissue affecting other tissues and making them more insulin sensitive,” he says.

Fat may also support health because active beige and brown adipose tissue may burn fat and oxidize it away from the body. This oxidation may help keep energy expenditure high and lower the fat content in circulation. While these theories are “less studied and appreciated,” Czech notes, they support the idea that some people may be genetically predestined to have more adipose tissue that can modify the rest of their body.

Beyond BMI

Variability in the distribution of adipose tissue is one reason Stanford doesn’t give her patients a target weight to work toward. “I don’t believe in giving them a target BMI because it doesn’t tell me anything about their health,” she says. “Things beneath the hood determine their health. These include their cholesterol, blood sugar, blood pressure and other factors that give me a sense of their metabolic life. That’s based on measurements I can’t get just by looking at them.”

She also notes the demotivating nature of trying to meet a BMI standard that wasn’t created for your specific body type. It’s problematic when health care professionals make assumptions about a person based on their BMI. “It leads to avoidance of health care,” Stanford says. “It leads to poor treatment adherence. And it leads to physiologic changes. When people experience weight stigma, studies have shown increased cortisol levels in those individuals. We have higher hemoglobin A1Cs in those who experience and internalize weight stigma.”

While sources agree that BMI has value as a screening tool, they also agree it’s important to dig deeper when assessing metabolic health. Heymsfield notes that researchers who study metabolism and physiology use a variety of technologies and techniques to learn more about a person’s body composition, particularly adipose tissue. These methods include dual-energy X-ray absorptiometry (DXA), a quick and pain-free scan that analyzes adipose tissue, lean muscle mass and bone density. Researchers also use MRI and hydrostatic weighing for a more accurate assessment of adipose tissue distribution. 

At the clinical level, one simple measure can give a clearer picture of adipose tissue distribution: waist circumference, Heymsfield says. “BMI just makes you into a cylinder. When you add waist circumference, you can learn more about their phenotype, such as whether they’re muscular with a small waist even though their BMI might be high,” he says.

Czech says it’s important to consider a person’s phenotype. “At this point, at the physician’s office, there’s not a lot of sophistication that goes into this,” he says. “But human body phenotype is important to look at carefully because it can give clues and hints about adipose distribution.” For example, adipose tissue distribution below the waist may be protective, while fat in the visceral area may be less protective.

As for assessing metabolic health and predicting the likelihood of disease, Czech notes that glucose tolerance tests are standard. Some researchers also use the homeostatic model assessment for insulin resistance (HOMA-IR). “It’s a measure of circulating insulin levels versus glucose levels, which gives a rough indication of insulin resistance and sensitivity,” he says. “Those methods are much more diagnostic in terms of metabolic health.”

Other researchers are using more experimental, increasingly sophisticated measurements such as the hyperglycemic clamp technique and euglycemic insulin clamp technique. “Hyperinsulinemic glucose clamp studies done in the research laboratory can assess adipose tissue and insulin sensitivity, as well as liver and muscle, with a pretty high degree of reliability,” Czech says.

The research question is whether overall metabolic health can be correlated to the distribution of adipose tissue in different depots in the body. “In clinical studies and in model studies, we can define clearly not only the level of insulin resistance in the whole body, but also the level of insulin resistance or sensitivity in individual tissues, including adipose tissue.”

So, what’s the future of BMI? For now, sources agree it’s here to stay as a screening mechanism.

“One of the ways BMI continues to be of interest and roughly useful is as an indication of weight gain,” Czech says. “We do know that in general, a more lean phenotype in a given individual is often more healthy than a high BMI condition. But once you get away from this gross indicator and start delving down into the details, you realize that it’s so much more complicated.” 

This article was originally published in the November 2022 issue of The Physiologist Magazine.