The Science of BMI - Is it reliable?
Where did it come from and should it still be used?
If you paid attention in health class, you would have been taught that every person can be categorised according to their body mass index, also known as BMI. Using a simple calculation, people are allocated to one of four categories – “underweight”, “healthy”, “overweight” or “obese”.
For decades, BMI has been used by health professionals and researchers to indicate the health and future health risks of individuals. However, in recent years, the criticisms for this practice have grown as the usefulness of BMI is called into question.
In this article, we’ll talk you through the history and evidence surrounding BMI, and unpack the arguments for and against its use.
Where did BMI come from?
The equation used to calculate BMI was created by Belgian academic, Lambert Adolphe Jacques Quetelet nearly 200 years ago. Quetelet studied astrology, mathematics and statistics – NOT medicine or the human body.
On this note, it is important to highlight that BMI was never supposed to be used to measure individual health – instead, it was designed for the statistical analysis of populations. So essentially, we’re criticising BMI for doing a poor job at a task it was never intended for! But more on its downfalls later…
This article by writer and absolute icon Aubrey Gordon (aka Your Fat Friend) gives a more in-depth background on the development of BMI and how it came to be used in our daily lives.
Why shouldn’t BMI be used?
ONE: Healthy bodies come in all shapes and sizes.
Despite the frequency at which it’s used, BMI is actually a pretty unreliable indicator of individual health.
One study that demonstrates this is a 2016 study into the cardiometabolic health of 40,000 people with a range of BMIs. Their results found that 30% of participants in the “healthy” BMI range were actually metabolically unhealthy based on their blood pressure, cholesterol levels and inflammatory markers. Additionally, almost 50% of “overweight” participants and almost 30% of “obese” participants were metabolically healthy based on the same measures.1
On top of this, another study of more than 600,000 people found that the lowest rates of death were in those with a BMI of between 23 and 29, which is considered the “overweight” category.2
This shows us that BMI is not the accurate predictor of health that we have been led to believe it is!
TWO: The equation is much too simple
When the BMI equation was created, calculators were not commercially available – consequently, Quetelet’s formula had to be simple and easy to use.
However, its simplicity is one of its key downfalls!
Firstly, the equation was developed based on the measurements of white European men, without taking into account the huge variations in the healthy body weights of different genders and ethnic backgrounds.
For example, research has found that BMI overestimates the health risks in Black populations underestimates them in Asian populations.3,4 It doesn’t even work well for people of different heights, as it divides weight by more in shorter people than in taller people, giving the impression that taller people are larger than in actuality.
Secondly, BMI doesn’t distinguish between different body components, assuming that all bodies are high in fat and low in muscle mass. This is problematic as different components of the body weigh different amounts.
For example, bone is much denser than muscle and muscle is much denser than fat. The denser the material, the more it weighs! Consequently, someone with strong bones and a high muscle mass will have a higher BMI than someone without, despite both of these qualities being objectively healthy.
THREE: The categories are arbitrary
In 1995, the World Health Organisation published a report recommending that the BMI thresholds for what was considered “overweight” and “obese” be lowered. This report was primarily written by another organisation called the International Obesity Task Force, who were funded by two pharmaceutical companies that produced weight loss drugs.
When the cut-offs were lowered in 1998, the number of people who were categorised as “overweight” or “obese” increased by the millions overnight, leading to (you guessed it) an increase in sales of weight loss drugs.
We think it’s pretty clear that this change was not merely evidence-based!
The other issue with these thresholds is their definitive nature; gaining or losing just one kilogram can push you from one category into another, from “healthy” to “unhealthy”. Since a person’s body weight can vary much more than this on a day-to-day basis, due to factors like hydration, food intake and exercise, this much too specific to be useful.
So if BMI is not accurate, what can we use instead?
Since these criticisms of BMI have come to light, there has been much discussion in health and research circles about what might be used in its place.
At Embody Health London, we practice from a Health At Every Size perspective. One aspect of this is that we don’t evaluate health based on weight; instead, we look at things like behaviours, blood tests, family history and lifestyle, to give a more accurate and holistic view.
We also think it’s important to question if there is a true need to categorise people by their weight. When we know that weight is a poor indicator of health, why is it necessary or (more importantly) kind?
While BMI may be able to give an indication of body size, it can tell us nothing about a person’s health (current or future), lifestyle or character. Let’s leave BMI in the past where it belongs.
Karli Battaglia, MSc, APD
EHL Team x
- Tomiyama A, Hunger J, Nguyen-Cuu J, Wells C. Misclassification of cardiometabolic health when using body mass index categories in NHANES 2005–2012. International Journal of Obesity. 2016;40(5):883-886.
- Campos P, Saguy A, Ernsberger P, Oliver E, Gaesser G. The epidemiology of overweight and obesity: public health crisis or moral panic?. International Journal of Epidemiology. 2005;35(1):55-60.
- The Endocrine Society. Widely Used Body Fat Measurements Overestimate Fatness In African-Americans, Study Finds [Internet]. ScienceDaily; 2009. Available from: https://www.sciencedaily.com/releases/2009/06/090611142407.htm
- Racette S, Deusinger S, Deusinger R. Obesity: Overview of Prevalence, Etiology, and Treatment. Physical Therapy. 2003;83(3):276-288.