Body Mass Index, or BMI, has become somewhat of a catchphrase in a world of ever-increasing obesity. Doctors, dietitians, and personal trainers are just some of those who use your BMI to determine whether you are obese, or at risk of becoming obese. Even the tools we use for fitness – electronic scales, heart rate monitors, and even our smart phone apps – use BMI as a primary indicator of whether you are at your ‘ideal weight’. However, while everyone is busy calculating their BMI and worrying whether they’re outside the ‘golden range’, few seem to be asking what this number actually represents, and fewer still whether it actually has any validity.
What are we actually talking about?
Before trying to answer those questions, it’s important to define a few terms:
‘Weight’ – contrary to popular belief, your weight is not a measure of how fat you are. Your weight is a measure of the effect of gravity on your mass – that is, the resultant force that gravity is having on the total mass of matter that makes up you. Weight doesn’t care what this matter represents – fat, bone, muscle, fluid – it just tells you what gravity is doing to you. Because gravity is pretty much the same all over Earth, the only way to change your weight is to change your physical mass. By that reasoning it could be said that to be ‘overweight’ or ‘underweight’ means you either have too much or too little physical mass compared to an ideal value. However, in practice we do not define these terms in this way.
‘Overweight’ and ‘obesity’ are defined by the World Health Organisation (WHO) as, “abnormal or excessive fat accumulation that presents a risk to health.” Note that the thing that is being measured or compared in this definition is the amount of fat that makes up an individuals body composition. Therefore, to determine whether a person has an abnormal or excessive accumulation of fat, it stands to reason you need to know the amount of fat that makes up a person’s mass. WHO further defines obesity as a percentage of body fat equivalent to 25% total body weight for men, and 35% for women.
‘Underweight’ interestingly enough, is most often defined by a person’s BMI, which is discussed below.
‘BMI’ – again, contrary to popular belief, BMI is a mathematical calculation, not a measurement. BMI is calculated by dividing your weight (in kg) by the square of your height (in meters). So for example, a person weighing 70kgs at a height of 1.75m would have a BMI of: 70kg / (1.75m x 1.75m) = 22.86 kg/m2.
So what’s the problem with BMI?
Well, there are several problems which continue to be debated within the literature today, including concerns around its validity in reference to specific populations and its use as a diagnostic tool (which, by the admittance of its inventor, Belgian mathematician Lambert Adolphe Jacques Quetelet, it was never intended for). However the most glaring one should be obvious from these definitions. BMI tells us about the relationship between the height and weight of an individual, however it does not – in fact, it cannot – tell us anything about the composition of that individual’s body. Specifically, it does not – cannot – tell us anything about the fat composition of that individual’s body.
This in itself would not be a concern, if not for the fact that BMI is frequently used to categorise people as being underweight, overweight, or obese. The fact is, BMI cannot make this categorisation because it does not measure fat composition. At worst, it runs the risk of generating false positives and false negatives when attempting to categorise people based on a comparison of height and weight. This can be illustrated with a simple example using a muscular athlete, such as a competitive body-builder.
Meet Jay (Jason) Cutler, a four-time Mr Olympia winner from the United States. Jay is 5’9″ (1.75m) and at competition weight approximately 274 pounds (124kg). Based on his height and weight, Jay’s BMI is 40.49 kg/m2, which accordingly to WHO classification, puts him in the morbidly obese category. Looking at Jay at the 2009 Mr Olympia competition, you would be hard pressed to call this man “morbidly obese.” Why is his BMI so high? Simply because muscle has a higher density (approximately 18% greater) than fat, so for the same volume of tissue, muscle weighs more than fat. As such, Jay’s BMI is clearly a false positive result. (For interest, Jay’s weight outside of competition has been recorded at 310 pounds (140kg), which would place him in the super obese category!)
The problem of density differences between muscle and fat can also give false negative results as well. Consider someone with a low proportion of muscle mass (eg: sedentary-lifestyle) or whose muscle mass is reducing over time (eg: elderly, or active person becoming inactive). It is quite possible that this person will record a BMI that would be considered ‘normal’, yet physiologically have a body fat percentage higher than that considered ‘healthy’ by the WHO. This has been demonstrated in a number of scientific studies by comparing BMI with more accurate body composition recording techniques, such as dual energy x-ray absorptiometry (DXA or DEXA), including one study where approximately 1 in 4 men and 1 in 2 women were incorrectly classified by BMI . Can these error rates be considered acceptable when screening for risks to people’s health?
So why do we use it?
Despite its problems, BMI’s simplicity is the main reason why it continues to be used. It’s quick, cheap, and easy to do. If you have the ability to measure a person’s height and weight, and basic math skills, you can categorise someone as being within a ‘healthy’ weight range, or not. Apparent anomalies are often dismissed through subjective observation by the assessor. “No, Jay, of course you’re not obese,” is what we would expect Jay Cutler’s doctor to advise him, for example. Though in light of the incidence of false negatives BMI produces, it could be argued we should be questioning whether such subjective opinions are valid, or even putting individuals at risk of being miss- or undiagnosed.
Critics of BMI have cited more potentially malicious reasons why the use of BMI persists. For example, some health insurance companies adjust their customer’s premiums based on their BMI – the higher your BMI, the more you pay – because they are considered to be at higher risk for developing problems with their health. Is this fair for the professional athlete, a person considered to be at peak physical fitness, and therefore healthy compared to the rest of the population? What about the 29 million Americans who suddenly “became fat” in 1998 when the U.S. National Institutes of Health and the Center for Disease Control (CDC) lowered the U.S. cutoff for ‘normal’ BMI from 27.8 to 25 ?
What’s the alternative?
There are many different ways body composition can be measured. In relation to body fat, hydrostatic weighing (weighing underwater) and whole-body air displacement plethysmography (ADP) are considered the gold standards of getting an accurate value. DXA or DEXA is also highly regarded, though primarily used in research. However, these procedures require specialised equipment and operators, and in some cases can be quite expensive to undertake.
Bioelectrical impedance analysis (BIA) is a cheaper alternative which is gaining popularity, particularly as an ‘in-home’ means of determining body composition through the use of body composition analysis scales. BIA has been demonstrated to be useful for predicting the body fat composition of groups, however shows poorer accuracy for individuals, likely due to the fact that readings can vary depending on an individual’s level of hydration (or dehydration) .
Skinfold calipers are another inexpensive and time-honoured method of determining body fat percentage, and chances are if you’ve been involved in college/university level sports, or been part of a sports institute, you’ve felt the skinfold pinch. There has been much ongoing debate in the scientific literature about the accuracy and usefulness of skinfold measurements in determining body fat, with issues such as inter- and intra-tester reliability (ie: whether the same person or different people can produce the same results with each test), the sites used, the methods of calculation, and again, the hydration level of the subject, identified as potential problems. At best, it seems skinfold measurements are accurate at determining body fat percentage on lean athletes, but less so for those with excess body fat or loose connective tissue (ie: the elderly) .
In the last two decades, waist circumference measurement has been increasingly used as a screening tool for assessing individual’s risk of developing diseases associated with excess body fat, such as diabetes and cardiovascular disease. Like BMI it is quick, easy, and inexpensive to administer, requiring only a tape measure. Current Australian guidelines state that waist measurements above 94cm for adult males and 80cm for adult females are indicator of excessive internal fat deposits, and increase the risk of chronic disease. Waist circumference has been demonstrated to be a good predictor of visceral fat (fat around the organs)  and a better predictor of risk of cardiovascular disease [4,5,6], type II diabetes [4,5], and metabolic syndrome  than BMI.
Is it time to say “bye” to BMI?
It is apparent that BMI tells us nothing about the composition of the body, and that there are problems with its use as an indicator of diseases such as obesity, and with its accuracy at identifying individuals who may be at risk of further health complications based on their body composition. It’s also apparent that there are more accurate ways of determining body composition, and these may be better indicators for people’s risk of developing diseases related to unhealthy levels of body fat . Is it therefore time we stopped using BMI?
It’s not a simple question to answer. BMI still has potential in screening people who are underweight compared to a normal population, and flag the need for further examination, to determine if they may be suffering from diseases such as malnutrition or anorexia. However, while BMI can determine whether an individual is ‘overweight’ compared to a normal population, it cannot give any indication as to why that is the case. Therefore it would seem justifiable that BMI should no longer be used as a means of categorising people as being ‘obese’, as it cannot indicate what the body composition of an individual is.
Further, given that other methods of body composition analysis are better predictors of the risk of disease and other complications associated with unhealthy levels of body fat, it seems justifiable to suggest that BMI should be the least favoured tool utilised by clinicians and individuals. This is particularly true where other methods, such as waist circumference measurement, have been demonstrated to be more reliable indicators of risk of associated health problems, such as cardiovascular disease, and are just as quick, easy and cost-effective to administer as BMI.
Why, then, should we continue to use BMI?
Do you know your BMI? Do you agree with the category it places you in? Would you rather know your BMI, or your Body Fat Percentage, in terms of making decisions about your own health? Please feel free to share your thoughts in the comments section below.
1. Shah, N. R. and Braverman, E. R. (2012). Measuring adiposity in patients: the utility of Body Mass Index (BMI), percent body fat, and leptin. PLoSOne. 7(4): e33308 1-8.
2. Cohen, E. and McDermott, A. (1998). Who’s fat? New definition adopted. CNN: 17 June 1998. Retrieved 26 June 2014 at: http://edition.cnn.com/HEALTH/9806/17/weight.guidelines/
3. Houtkooper, L. B. et al. (1996). Why bioelectrical impedence analysis should be used for estimating adiposity. American Journal of Clinical Nutrition. 64(suppl.): 436S-448S.
4. Wagner, D. R. and Heyward, V. H. (1999). Techniques of body composition assessment: a review of laboratory and field methods. Research Quarterly for Exercise and Sport. 70(2): 135-149.
5. National Health and Medical Research Council (2013). Clinical practice guidelines for the management of overweight and obesity in adults, adolescents and children in Australia. Melbourne: National Health and Medical Research Council.
6. Siavash, M. et al. (2008). Comparison of Body Mass Index and Weight/Height Ratio in predicting definite coronary artery disease. Annals of Nutrition and Metabolism. 53: 162 – 166.