BMI Part One: Uncovering the Truth About BMI

June 13th, 2012 by Sara Upson

I was asked to write a post about BMI (Body Mass Index).  It is popping up everywhere and we are hearing more and more about it.  Students are getting sent home with their BMI on their report card, states are conducting statewide screening and surveillance using BMI, health care facilities use it, and it is prevalent in research.  That’s great, but what does it really mean?  What information can you really get from knowing someone’s BMI?  To find out let’s play a game!  Think of the following individuals- their body shape/size- maybe even google them- and try to guess their BMI:

George Clooney, Arnold Schwarzeneggar, Brad Pitt, Johnny Depp, President Obama, Michelle Obama, Beyonce Knowles, Kirstie Alley, Cameron Diaz, Tyra Banks, Serena Williams. Go ahead, don’t be bashful.  I’ve included the BMI categories below to help you make your guesses.

Category BMI Range
Underweight <18.5
Normal 18.5-24.9
Overweight 24.9-29.9
Obese >30


How did you do?

George Clooney:  31- obese

Arnold Schwarzeneggar- 33- obese

Brad Pitt- 31- obese

Johnny Depp- 29.8- overweight (almost obese)

President Obama- 24 Normal

Michelle Obama- 22- Normal

Beyonce Knowles 21- Normal

Kirstie Alley- ~23- Normal

Cameron Diaz- 17.7- underweight

Tyra Banks- 25- overweight

Serena Williams- 21- Normal

Now, what did you learn about looking at these individuals and guessing their BMI?  I think the most important thing is you can’t know someone’s BMI just by looking at them, and that BMI does not always correlate to adiposity.  I find the most interesting thing is that all the fit, muscular men are overweight or obese, all the women who get slandered by the media for being overweight are normal, and all the women who appear “normal” are actually underweight.  I find it most shocking that the women we think of as curvy celebrities are actually on the lower end of the BMI for normal!  Now granted, this information was obtained from the internet, so it may not be 100% accurate.   The question is, what do we really learn, what information do we get from knowing BMI?  I would like to say nothing, however admittedly sometimes it can be helpful.  Before I write about how it is helpful, I would like to start with specifically what BMI is and how it was created.

BMI is a ratio of a person’s weight and height.  It literally is mass in kilograms divided by height in meters squared. (BMI= kg/ (m2)).  Although BMI seems relatively new, it has actually been around for quite a while.  BMI was originally developed in 1832 as the Quetelet Index by Adolphe Quetelet, a Belgian statistician, astronomer, and mathematician.  Originally he developed it to describe the ‘normal man’ and the distribution around the norm.  He observed that as people grew weight increased nearly as the square of the height.  Obviously in 1832 this was a very novel concept, however it was novel because of the statistical findings and nothing more.  When he developed it, he cared nothing about overweight or obesity.  Fast-forward to the years following World War II when insurance companies noticed that there was increased mortality in their overweight policyholders.  This lead to development of the Metropolitan Life Insurance Company Tables where individuals were classified into weight categories based on average weights for a given height.  The insurance company noticed that people seemed to come in different frame sizes and further classified height and weight into small, medium, and large frame categories.  You might recall seeing these tables before, however they are no longer used.  Later it was determined that the information that was collected was not very accurate as it was collected posthumously (after death).   It wasn’t until 1972 that Ancel Keys confirmed the validity of the Quetelet Index and renamed it- BMI.  When Keys renamed the Quetelet Index BMI he stated that it was appropriate for population studies, but inappropriate for individual diagnosis.

It seems like we have come a long way since Keys stated that BMI should not be applied to the individual.  Since then tables have been derived depicting set BMI categories.  In 1995 the categories were slightly different than they are today.  In fact, BMI above 30 and even 45 were classified as just overweight.  However, one morning in 1998 people woke up, the tables had changed, and suddenly they were classified as obese.  (This change in classification really plays an important part in the sky rocketing rates of obesity.)  What justification was there to change this classification from overweight to obese?  Not much.  The National Heart, Lung, and Blood Institute (NHLBI) published the changes correlating higher weight with increased deaths.  Since then many studies have questioned the evidence for changing the BMI categories to focus more on obesity.  In fact, generally the highest risks for death are associated with the underweight and the moderate to extremely obese (BMI greater than 35) category.  Additional studies are coming out that find individuals in the overweight and mildly obese category with the same relative risk of mortality as individuals in the normal weight category.   Now I am not denying that there is an increase in obesity.  People are becoming fatter.  However, the use of BMI to calculate prevalence of obesity is not entirely accurate.  I could probably go on about the development of BMI and its utilization for much longer, but there is still more information to cover!  The main thing to know about the development of BMI is that it was invented by an individual in the 1800s who was fascinated with statistics and proportions.  Since then it has evolved to its use today, however it was never intended for individual diagnosis.

So, if BMI was not invented to be applied to the individual, what is it good for?  Well, it is an easy calculation that can quickly look at a ratio of height to weight.  It is an internationally standardized and accepted measurement.  It can be a simple way to estimate adiposity of a given population, and it can be used to predict risk of death or disease within populations based on correlation data.  I must admit, when I worked in the hospital we documented BMI for every patient and sometimes it was helpful to know the patient’s ratio of height and weight.

As much as I hate to admit it, BMI does have some utility, however it also has some major limitations.  Again, it is not valid for evaluation of individuals.  This is because it does not consider cultural and/or genetic variations, and it does not distinguish between body composition (fat versus muscle mass).  Different cultures tend to have different body compositions.  In some populations BMI will underestimate levels of adiposity and in some populations it will overestimate levels of adiposity.  Let me give you an example, in Asian and Pacific populations there has been increasing evidence that health risks increase below a BMI of 25, which is still within the normal BMI category.  On the flip side, some studies show that African American populations are not at increased risk when their BMI is between 25-30, which is technically considered overweight.

BMI categories are applied universally, to everyone, and we’ve looked at how they do not apply universally to specific populations, however lets also consider gender, age, and genetics.  Men generally have more muscle mass than women.  Muscle mass weighs more than fat, so many times BMI does not accurately depict adiposity.  This can be seen in the above pictures of Arnold Schwarzeneggar, Johnny Depp, George Clooney, and Brad Pitt, all technically classified as overweight or obese based solely on their BMI.   With age we also tend to see changes in body composition.  As age increases it is normal to see a decrease in muscle mass.  This means that BMI could potentially underestimate adiposity in older individuals.  However, there is mounting evidence that above the age of 70 additional weight actually has protective effects and that the relative risk of death is similar to normal weight individuals.  BMI also does not distinguish between genetic variations and frame size.  I don’t think you have to look far to realize that people have individual variations in frame size that vary from small to large.  I’m not just talking about weight, but about frame.  This is just the way we are made.  People with smaller frames could potentially be in the normal BMI category and be at a higher health risk.  People with larger frames could potentially be in the overweight or even obese category and be at a lower health risk.

Really this all boils down to the fact that BMI estimates adiposity.  It does not distinguish between muscle mass or fat mass.  This is why any person with a large percentage of muscle mass will be classified as overweight or obese.  If you look at an entire population, as BMI was intended, this variation will cancel out.  On the population level, variation in culture, genetics, gender, and age will also cancel out.

Too much emphasis is placed BMI and BMI category especially on the upper end of the spectrum.  Yes, there is a correlation for populations; health risks on the population level do increase as BMI increases.  However, what about the opposite end of the spectrum?  The underweight category?  It seems like, per society, you can’t be too thin.  When we look at health risks of the underweight category, they are actually higher than the normal weight category and, in some cases, higher than the overweight and obese categories.  Too much pressure is placed on weight loss and low BMI.  This may have unattended consequences of supporting efforts to achieve and maintain an unrealistic lean appearance.  Emphasizing the importance of low BMI may even increase the prevalence of eating disorders such as anorexia and bulimia.  The serious health hazards of being underweight are minimized.  It is likely the number of individuals subject to these hazards will increase as children, young adults, and really, all age groups are more motivated to maintain a dangerously low BMI.  Achieving and maintaining an unnaturally low BMI can cause serious health consequences including increased risk of eating disorders, osteoporosis, and malnutrition.

This focus on achieving a low BMI really puts children at risk.  Of course, if your child comes high with a BMI higher than the normal category on their report card, you are going to be concerned.  The question is… is this concern validated?  Should you really be concerned?  Should you put your little honey on a diet?  The overall answer is no.  Children and BMI is a whole different can of worms!  I think the most important thing to remember is that your child is still growing and needs the appropriate nutrition for growth.  On my next post I will cover BMI and children, and what to do if your child is not in the normal category.  The following post after that will cover additional ways to assess health and health risk, besides using BMI.

The most important concept to take away from this post is that BMI is simply a ratio of height and weight.  It cannot differentiate muscle mass from fat mass.  It does not distinguish between cultural, gender, genetic, and age variations.  For this reason the categories developed for BMI are not cut and dry.  On the population level, generally, health risk does increase as BMI increases, but this is not always the case.  For individuals, just because your BMI increases, it does not necessarily mean that your health risk increases.  The most important thing to look at is truly your behaviors.  Are you exercising?  Are you eating a balanced diet?  Are you taking care of yourself?

Leave a Reply

Your email address will not be published. Required fields are marked *