The Failure of BMI as a Medical Metric

Aug 19, 2015

standard Body Mass Index (BMI) calculator puts LeBron James (6’8”, 250lbs) as close to being “obese" as he is to being “normal” weight. The use of BMI has also kept individuals known to have bulimia or anorexia nervosa from getting treatment because their BMI isn’t low enough.

These two examples demonstrate that BMI has limited value for determining a “healthy” weight. And I’m not the first to point this outNotEvenClose. But that’s not the end of the story. What, sense should we make of studies showing an increased risk of death for BMI-overweight individuals?  Obviously, we have to take seriously any robustly demonstrated correlation. So we should take these studies at face value and, when feasible, decrease our risk by aiming to lower our BMI. 

At the same time, we should recognize that this kind of study doesn’t validate BMI as a criterion for health any more than it would also validate “200 lbs” as a criterion for health. That is, if we did a study of people in the United States who weighed more than 200 lbs and compared them with people who weighed less than 200 lbs, the people who weighed more than 200 lbs would, on average, have poorer health outcomes. That is, weighing more than 200 lbs will be, in general, correlated to poorer health outcomes. 

Does this tell us anything about whether a particular person should weigh less than 200 lbs? Of course not. It could only demonstrate that, in general, it’s better for people’s health to weigh less than 200 lbs. In the same way, BMI demonstrates that, in general, individuals are less likely to have health problems if their BMI is normal. Does this mean normal BMI is generally correlated to better health in general? Yes. Does it mean that individuals with a normal BMI are healthier than other individuals? No. 

Perhaps most importantly of all, BMI wasn’t created as a means to distinguish healthy versus unhealthy characteristics. It was produced in 1832 by a Belgian polymath who analyzed variations in weight and height across individuals whose data was available. Were the heights and weights of these individuals the heights and weights of “healthy” individuals? We couldn’t know. And yet, in my doctor’s office, what do they use to determine if they need to say something to me? BMI. When studies talk about best strategies to achieve healthy weight, what do they use? BMI.

This is where medicine is failing us. How can institutions aimed at the better health, continue to use a metric of assessment that has an origin of limited applicability and that provides erratic guidance on what counts as a “healthy” weight? Perhaps for this reason, the National Institutes of Health have now changed questions of healthy weight to BMI combined with waist size (men should be no more than 40” and women no more than 35”) and other health problems. But I would suggest a more radical evaluation. The question should not be “can we get some information from BMI?” but “What reason is there to use BMI?” 

My conclusion is unsurprising—limit the weight you put into BMI as a way to determine “health.” But this conclusion doesn’t really help me, or perhaps even you. For a number of years I didn’t really struggle with my weight—I just put on a bunch of it. Over the last year, I’ve lost 1/6 of my body weight. To get to my target weight, I want to lose another 1/6. At that point, according to BMI, I’ll still be 20 lbs overweight. So, should I keep my target weight or should I listen to the BMI recommendation? I guess I’ll see how I feel. If my target weight seems healthy enough, I’ll stop there. At this point, I’ll put as much stock in my intuitive judgment and how I feel as I would in BMI. 

Abraham Schwab is an associate professor of philosophy and a medical ethicist at IPFW.  

Opinions expressed in this column are those of the individual writer and do not necessarily reflect the opinions of the staff, management or board of Northeast Indiana Public Radio.  If you want to join the conversation, head over to our Facebook page and comment on the post featuring this column.