Weighing In on Obesity—What BMI Can Tell You

, food systems & health analyst | December 16, 2014, 4:09 pm EST
Bookmark and Share

In public health, one of the first steps to addressing a problem is defining it. The second, measuring it. Most Americans are aware that our country has an obesity problem—in adults as well as children—in part because of our unhealthy food system. But how do researchers gauge the extent of the problem, and what do we know about whether it is increasing?

I recently began thinking about this in my veterinarian’s exam room, when I heard words that no obesity-prevention researcher wants to hear: “Your cat is heavy for her size.” Embarrassed, I quickly countered, “Based on what? Her body mass index? Where does she rank on the chart?” My vet simply chuckled and shook her head, “Cats don’t have a body mass index chart.” It turns out that obesity in cats is determined by placing your hands around their sides and trying to feel their ribs. If you can’t feel them—the cat is overweight.

Defining obesity in humans is complex

Defining and measuring obesity in humans is a bit more complex. It requires both researchers and doctors deciding how to measure body fat and then establishing body fat thresholds—healthy weight, overweight, and obese. Just as I felt my cat’s “heavy” diagnosis was biased, we also deal with bias in human science. When measuring body fat in humans, several factors that may bias weight thresholds are age, sex, race/ethnicity, body fat distribution, height, and weight.

According to the World Health Organization obesity is excessive body fat that may impair health.

Figure 1: BMI and Mortality Ratio*

Currently, the most widely-accepted method to measure excess fat is body mass index (BMI). BMI is obtained by dividing weight (measured in kilograms) by height squared (measured in meters). Generally, most professional organizations (i.e., the Centers for Disease Control and Prevention) consider persons obese if their BMI is 30 or greater. This threshold is largely based on the positive relationship between BMI and mortality (see Figure 1). That is, as BMI increases, mortality rates increase. BMI is also positively associated with Type II diabetes and heart disease.

Options for measuring body fat

BMI is the most popular method for measuring obesity. It’s quickly measured, easily calculated, and cost-effective. Unfortunately, since BMI is based on height and weight, it can’t account for differences between fat mass and lean body mass (i.e., muscle). Additionally, it has been suggested that BMI threshold categories are based predominately on Caucasians living in Europe and the United States. This can be problematic since some races/ethnicities have higher percentages of body fat than do Caucasians. Therefore they might have a normal BMI, but still classified as overweight or obese. While BMI is one of the most widely-used methods for measuring body fat, there exist more precise methods. Three other options for measuring body fat are: bioelectrical, densitometry, and imaging techniques.

  1. Bioelectrical Impedance Analysis. Bioelectrical impedance analysis (BIA) measures body fat by sending low electrical currents through the body with the idea that a leaner person will have lower resistance to the electrical current. However, BIA’s precision is influenced by the amount of water in the body, which can be influenced by body structure (i.e., short person vs. tall) and hydration status. For example, someone who just drank a glass of water will test differently than someone who hasn’t had anything to drink in the past couple hours.
  1. Densitometry. Densitometry, also known as “underwater weighing” is based on the idea that fat is less dense than water. Therefore a person with lower body fat will have higher body density. Densitometry is often considered the “gold standard” for accurately measuring body composition. Unfortunately, the test is time consuming, expensive and not easily performed on children and older adults since it requires a person’s attention and cooperation. If you’ve ever experienced a child’s bath time, you know “attention” and “cooperation” aren’t usually words used to describe it.
  1. Imaging techniques, such as dual energy x-ray absorptiometry (DXA), CT scans, and MRIs are very accurate when measuring body composition. Unfortunately, they are also very time-consuming, complex, and expensive. The tests can take up to an hour and technicians operating the machines must be highly trained. Additionally, since DXA uses X-ray exposure, it is not suitable for pregnant women. And good luck getting a health insurance company to cover the cost!

Is BMI a good measure for obesity?

While BMI may not be as precise as BIA, densitometry, and imaging, it is easy and cost-effective. That makes BMI a great tool for doctors and researchers. It gives doctors the opportunity to discuss with patients how they might be at risk for becoming overweight or obese. And it gives researchers the ability to track obesity trends from year to year. In the latest report from America’s Health Rankings, released just last week, results show that adult obesity rates in the U.S. increased 7 percent from 27.6% to 29.4%. With almost 30 percent of adults being obese, America is in for a rude awakening. Increased healthcare costs due to obesity-related chronic diseases such as type 2 diabetes, heart disease, dementia, osteoarthritis, respiratory illness, cancer AND decreased employee productivity from missed work days due to illness.

So, can you really trust BMI as a measure of weight status? Yes!—BMI is a great indicator for understanding population trends in obesity and goal-setting for policymakers who want to reverse the current trend.

Special Note for Cat Owners: Prescription diet cat food is approximately three times more expensive than regular cat food.

Posted in: Food and Agriculture Tags: ,

Support from UCS members make work like this possible. Will you join us? Help UCS advance independent science for a healthy environment and a safer world.

Show Comments


Comment Policy

UCS welcomes comments that foster civil conversation and debate. To help maintain a healthy, respectful discussion, please focus comments on the issues, topics, and facts at hand, and refrain from personal attacks. Posts that are commercial, self-promotional, obscene, rude, or disruptive will be removed.

Please note that comments are open for two weeks following each blog post. UCS respects your privacy and will not display, lend, or sell your email address for any reason.

  • Michael

    I feel like the article says that BMI is kind of mediocre as a measure of overweight, but it’s the best we got. First, I’m not positive why this a UCS issue, given that this article advocates exactly what doctors already do. But second, I’m not sure it’s very accurate or useful.

    For the individual: In the comments, the author has pointed out how a doctor will look at BMI and then override the results if the patient appears to be obese (or not) when the BMI says otherwise. This tells me the BMI is useless: the doctor knows by looking. In fact, to the extent that BMI suggests something the doctor does not see, we’ve already established that BMI might be high because of muscle or fat or something else, so again BMI doesn’t help.

    For the population: First, different population groups can have different typical BMIs, so as we have immigration, BMI may change over time for that reason alone. Beyond that, if the country goes on a health craze where we lose on average 5 lbs of fat and gain 10 lbs of muscle, BMI will completely mislead as to in which direction the country is headed. Likewise if we all lose a little muscle mass without gaining fat, the BMI will likewise mislead. Finally, of course, BMI produces a larger number for taller people (the formula logically should divide by height cubed but it only divides by height squared), so merely a population growing taller will produce a growing BMI. It’s a lousy and misleading measure for populations as well.

    If we were just stuck with BMI as a proxy for body fat percentage because there was no other practical option, then it would have to do. But that’s not the case. The author neglected to discuss a well-known methodology that is equally cost effective but vastly more informative. The US armed forces have been measuring body fat percentage for years with a tape measure, e.g. http://www.calculator.net/army-body-fat-calculator.html. For a doctor’s office to perform the extra measurements and put them in the computer would take possibly 30 extra seconds of a nurse’s time and from both an individual and population standpoint the data produced would have substantially more value.

    I posit that standardizing on a simple approach like this would give us much better data and would remain much more robust to variations in body type, for essentially no additional cost.

    • Lindsey Haynes-Maslow

      Michael, thanks for your comment! In response to your first question as to why UCS is even covering this topic, it’s because in the upcoming months we’ll be discussing the relationship between our current food system, obesity, and obesity-related chronic diseases. As I mention in my post, the first step to addressing a problem is defining it, then measuring it. For UCS to discuss obesity, it requires that we have a conversation about how it’s currently defined.

      In terms of physicians using BMI with patients, BMI is a well accepted tool throughout medicine that is associated with elevated
      risks of poor health outcomes. A physician who counsels a patient on
      their BMI is following the current standard of care.

      While you bring up another option for measuring body fat that the U.S. Armed Forces uses, it unfortunately only takes into account neck and waist measurements. This method fails to capture the fact that different bodies store fat in different areas. Additionally, each time a new measure is introduced into an equation, there is a possibility of introducing measurement error — which is even more of an issue if the measurement is physically obtained by a person.

  • Jay Wiener

    I’m very disappointed to see so many inaccuracies about the BMI published by a site designed to be read by scientists. The BMI is a useful tool for describing large populations, but it is a completely useless tool for describing individuals. It should never be used by a doctor to give advice to a patient; that borders on malpractice.

    Think of a broad, muscular athlete. Now think of his sedentary, overweight, chain smoking, beer drinking grandmother. They might weigh the same, although he is solid muscle and she is… not. If they were both, say, 5′ 10″, they would have the same BMI. Ludicrous. More important, meaningless.

    I’m a mathematician and have studied obesity for years. I wrote an algorithm that uses 25 factors (body stats, health history, and exercise history), and then creates a unique, healthy range of weight for each individual. I put it up for free at http://www.weightzonefactor.com because too many health professionals use the BMI, claiming ‘It is all we have.” It is not. You are welcome to use my algorithm for yourself and for your patients; if you check the site you will see it is ad-free and 100% safe.

    • Lindsey Haynes-Maslow

      Thank you for your well thought-out response. As I mention in my post, BMI is great population-based measure and screening tool for healthcare practitioners and their patients. I’ve worked with healthcare practitioners that are uncomfortable discussing weight with patients, and BMI is a good starting point to begin the conversation. Additionally, I have worked on qualitative studies that show that physicians do acknowledge the difference between an athlete and a heavier set individual, the latter which they might consult about weight management. I actually prefer color-coded BMI charts, since the “overweight” threshold is highlighted in yellow, signaling to patients that they should be aware of their weight.

      • Jay Wiener

        Thanks for replying, Lindsay. I live in Northern California
        and belong to Kaiser, which has placed color-coded BMI charts in almost every room of the facility. The charts are
        nonsense; they were created in the 1950’s, by actuaries working for Metropolitan Life Insurance Company. The actuaries had no interest in promoting good health; their goal was to increase the number of men who would be rated as ‘overweight’ during a physical exam so that they could increase the premiums. The result: the charts have
        given misinformation for sixty years. I have written about the BMI here: http://www.weightzonefactor.com/blog/why-do-doctors-still-use-the-bmi/

        A personal story: I am 5’ 10” with a very broad frame. I survived a heart attack while very young, lost over 100 pounds, and began to exercise regularly – both cardio and strength training.

        Today I weigh 240. The BMI charts tell me to weigh 170, but that would put me in the hospital. My own algorithm (www.weightzonefactor.com) tells me to weigh between 204 and 218, which is reasonable. The upper end would be higher if I did not have that heart attack, and if my father and brother were not diabetic. This is how medical science should work: with unbiased mathematics that use as much raw data as is feasible. The BMI charts fail that test.