History of ideal weight and BMI

When it comes to health, you often come across the terms ‘ideal weight’ and ‘Body Mass Index’ (BMI). This article discusses the historical background of these terms. Answers are provided to questions such as: In what ways has the ideal weight been calculated in the past, what does BMI mean and how did it originate, how have the cut-off points been determined and is a high BMI always detrimental to health? The question arises for parents: does the BMI say anything about my child’s health?

The ideal weight

The ideal weight is a concept that is used everywhere in our society. Advertisements about slimming products and books by many diet gurus revolve around this concept. This term is regularly used in health columns in newspapers and magazines and in television programs. The ideal weight appears on countless websites on the Internet. It is a seemingly obvious concept that few people will question. However, upon closer examination it appears that the ideal weight is not as clear a concept as it seems at first glance. The most obvious question is not easy to answer: how do you calculate the ideal weight of a specific person? If it is assumed that there is such a thing as the ideal weight then it should be possible to calculate it. Many attempts have been made in the past to devise a formula to calculate what the ideal weight should be.

Broca’s Index

The forerunner in this was the French physician Pierre Broca. He was looking for a method that would allow doctors and pharmacists to easily and quickly determine the amount of medication for a specific patient that should be calculated per kilogram of body weight. In 1871 he developed Broca’s Index, a formula that could be used to calculate how much someone should normally weigh based on body height. According to Broca, this was the body weight that matched a certain height, if there was no underweight or overweight. According to his formula, normal weight was calculated as follows:

  • Ideal weight (in kg) = height (in cm) – 100

Simply take the number of centimeters above 100 from the body height. That number should then be the ideal weight in kilos. Someone who is 180 centimeters tall should weigh 80 kilos according to Broca’s method. Broca’s principle of estimating weight based on body height has been used by many others to calculate ideal body weight. Over time, this resulted in a wide variety of formulas, all of which are based on Broca’s Index. Of unknown origin, the formula assumes the ideal weight for men should be 10% lower than Broca’s Index. According to this method, a man who is 180 centimeters tall should have an ideal weight of 72 kilos. According to this calculation, the percentage for women is even lower, at 15%. A woman with a height of 180 centimeters may not weigh more than 68 kilos. In addition to this relatively simple method, other formulas have also been devised over the years that were not based on this index.

The Hamwi formula

An example of a more complicated calculation is the formula of George Hamwi who published two formulas in 1964, one for men and one for women. Converted to the metric system, the Hamwi formulas for the ideal weight are as follows:

  • For men= 48.0 + (height – 152.40): 2.54 x 2.7
  • For women= 45.5 + (height – 152.40): 2.54 x 2.2

The ideal weight of a woman of 180 centimeters should not be 68, but 70 kilos, and that of a man of the same height should not be 72, but 77 kilos. In other calculations, other body characteristics are taken into account in addition to height. Such as the variation on the Hamwi formula, which can be found on various websites. The physique is taken into account. For thin and heavy people, body weight may deviate 10% from the average. The ideal weight of a man with a narrow build and a height of 180 centimeters may be 69 kilos and for a wide build 85 kilos. For narrow women of the same height, 62 kilos is the ideal weight and for wide women 76 kilos.

Differences between the calculations

The number of kilos that a man of 180 cm may weigh according to these two calculations of the ideal weight varies considerably: from 69 kilos to 85 kilos. For a woman of 180 cm it even varies from 62 kg to 80 kg. These are strikingly large differences. It appears that the inventors of the calculations and their users did not rely on the same information or insights. For example, the purpose for which the calculation was made sometimes does not correspond to what it is used for. For example, as described above, Broca’s Index was not developed to calculate someone’s ideal weight. The basis on which the calculations are based is missing. For example, with the variation on the Hamwi formula, it is not clear why a 10% deviation up and down is chosen and not 5% or 15%. There is no sound scientific basis for all these formulas.

The ideal weight is the statistical average

A calculation method that has been extensively researched is called the Body Mass Index, better known as BMI. Anyone who searches the internet will find hundreds of websites with calculation modules that use two pieces of data, height and weight, to calculate what the ideal weight should be based on BMI. The calculation method used is based on the Quetelet Index. This formula was developed by Lambert Adolphe Jacques Quetelet, an influential Belgian scientist from the nineteenth century. He was involved in the first Dutch census in 1830 and conducted research into, among other things, the development of the height of children and adults. In his work he attempted to describe the average person, a model of a non-existent person who was in all respects the average of a certain group of people. The groups he researched were, for example: all people who had come into contact with the Belgian police, or all soldiers in the French army or all people in a certain hospital. The average person of a certain group was exactly the average of all measurements of that group, an average person of that group. This not only involved all kinds of body measurements (skull circumference, height, weight, arm strength, etc.) but also data about income, education, drinking, age at marriage, etc. Using the model of the average person, the deviation from the average could also be made clear. When finding the average of a relationship between the data of a certain group, it is important that there is also a real average. This becomes clear when the two data between which the connection is sought are compared in a graph. If the connection exists, it is expected that the data of most people will be around a certain value and that the more the value deviates from this, the less it occurs in the group. Only then can it be properly determined what the average is. When he analyzed the body measurement data, he hoped to find a clear relationship between height and weight and the graph that matched it:

Figure 1

However, the graph he got was not very symmetrical. That happens more often. The problem that then arises is that the average cannot be properly determined. The graph in Figure 2 makes this clear:

Image 2

This shows that the average cannot be determined properly due to the distribution of the results. He experimented with various formulas, but it turned out that calculations such as: weight divided by height and weight divided by length times length times length (so weight divided by length3) did not give a good average either. However, if the formula weight divided by length times length was applied, then there was a clear average:

  • Quetelet Index: weight in kilograms : length in meters x length in meters

He thus received a graph that looked very much like Figure 1. The height of the curved line indicates how many people have a certain outcome. The higher the line, the more people have that outcome. This not only shows what the average is, but also to what extent a random outcome deviates from it. The more the value of a given person’s index is to the left or right from the average, the greater the difference with that of the average person. Further and further to the left the people are increasingly thinner for their height and further and further to the right they are increasingly fatter for their height. Quetelet examined data on the weight and height of a number of groups of schoolchildren, adults, elderly people and hospital patients near Brussels and came to the conclusion that on average a full-grown person

  • weighed just over 45 kilograms
  • was slightly longer than 1.60 meters
  • Quetelet Index of 17.58

According to Quetelet, the average person belonged to the class of the ideal type of person. He saw any deviation from this as ugly or undesirable.

If you are at the ideal weight, you will live longer

In Quetelet’s time, medical scientists suspected that being fat was probably not healthy, but no convincing evidence had been provided. Scientific interest in the relationship between body weight, height and health was aroused in the early 20th century by American life insurance companies looking for new criteria for new customers to better assess financial risks. They wanted to calculate the average lifespan using objective measures, such as height, weight, blood pressure, occupation, etc. Louis Dublin was a statistician who worked for the major life insurance company Metropolitan Life Insurance Company . He created The Dublin Standard Table of Heights and Weights in 1908 , which for the first time linked people’s height and weight to their life expectancy. In 1943, this insurance company developed different tables for men and women for the first time. These tables showed that people who were overweight for their height lived shorter lives than people who were not overweight. At that time, the term ideal weight was associated with these tables. The life insurance companies meant the weight at a certain height that predicted the longest life expectancy according to the tables. In the period after 1908, the tables were quickly used as a standard in American medical practice to determine ideal weight. This is striking because they were not initially intended for that purpose. For example, differences in lifestyle were not taken into account. Furthermore, the accuracy of the measurement data was doubted because the weight was not determined by independent persons, but by the policyholder himself. By using self-reporting, there was a risk that the policyholder lied about his data. In his 2007 study, Connor Gorber summarizes a trend that emerges from 47 studies: when self-reporting, women in particular tend to report a lower weight and slightly exaggerate their height. Furthermore, the tables were only intended to estimate life expectancy.

Methods to determine the healthy amount of body fat

Medical science’s search for the ideal body weight required a different approach to research, one that focused on health rather than longevity. Over time, it became increasingly suspected that the cause of the shorter life expectancy could lie in the amount of fat stored in the body: the more body fat, the greater the risk of premature death. Because fat people were more likely to develop diseases such as heart disease, high blood pressure and diabetes, it was suspected that these were caused by excess body fat. In order to conduct further research into the effects of excess body fat, it was important to develop a simple and, above all, cheap way to determine the amount of body fat. At the time, several methods were available to measure this.

When measuring skinfold, the thickness of the skinfold must be accurately determined at a number of places on the body using a special measuring device (a skinfold meter). The number of places to measure depends on the measurement protocol and varies from a minimum of 3 to a maximum of 7. With these measurements the amount of body fat could be calculated quite accurately. In order to properly determine the amount of body fat, accurately measuring the skinfold thickness is crucial. The skinfold caliper should be used with the same pressure and exactly on the prescribed spot. This method of measuring can only be done successfully if measured by experienced and trained employees. The costs of an investigation can therefore increase considerably, especially if it is a large investigation. Another method is to compare body weight when the person is underwater and normal weighing. With this data the amount of body fat can be determined quite precisely. This method is also very expensive due to the required measuring instruments.

Ancel Keys: BMI as a measure of the amount of body fat

Ancel Keys, a professor at the University of Minnesota, came up with a much cheaper solution. In 1972, a major American study was conducted under his leadership to investigate the accuracy of a number of calculation methods that could be used to estimate the amount of fat in a body based on height and weight, including that of Quetelet. The amount of body fat was determined in 7,424 people from five countries using direct measuring methods. These data were compared with the results produced by the different calculation methods. Quetelet’s index appeared to correspond most accurately with the direct measurement methods. The name of the index was renamed the Body Mass Index (BMI) by Keys, the formula remained identical to that of Quetelet. Keys explicitly warned that the index cannot be used to say anything about a specific person. The BMI was solely intended to compare data from groups of people in scientific research.

Inaccuracies in BMI as a measure of body fat

If this index is used to estimate the amount of body fat of a given person, other important aspects must also be taken into account. When calculating the BMI, it is not taken into account whether the person is a man or a woman, what the age is, to what extent the person is muscular and from which population group the person comes. And these data are important because they influence the amount of body fat:

  • A man’s body differs in composition from a woman’s: for example, a woman has more body fat than a man.
  • In children, body fat fluctuates from period to period and as a person gets older, the fat percentage usually increases somewhat.
  • People of Asian descent have a lower BMI, but more body fat.
  • Because muscle tissue is heavier than body fat, highly muscular people weigh significantly more than people of the same size but with less muscle.

For a significant number of people, the calculation incorrectly reflects the amount of fat in the body. However, BMI became the most popular measurement method that is still widely used today in science, healthcare, government and beyond.

When there is too much body fat: the origin of the cut-off points

Keys, like Quetelet, has not done any research into the limits within which a certain body weight at a certain height is good or not good for health. Much research into body weight and health conducted in the last century was based on the limits that emerged from the data of the American life insurers from 1943 and 1983. A man of 180 centimeters with an average build was considered to be overweight according to these limits. this one weighed more than 83 kilos. A woman of the same height and build was classified as overweight if her weight was 80 kilos or more. More recent research uses cutoffs derived from a series of comprehensive health surveys of the U.S. population conducted from 1976 to 1980, the National Health and Nutrition Examination Surveys (NHANES) . It states that:

  • an adult woman is overweight if her BMI is greater than or equal to 27.3. A woman of 180 centimeters is overweight at this limit if she weighs 89 kilos or more.
  • a man is overweight with a BMI of 27.8. A man of 180 centimeters is overweight from 91 kilos.
  • In addition, limits are also given for seriously overweight BMI higher than 31.1 for men and BMI higher than 32.2 for women.

In 1995, the World Health Organization (WHO ) proposed lower limits: a BMI above 25 meant overweight, above 30 extremely overweight. In 1998, the National Institutes of Health (NIH) , America’s leading health institute, formed a forum of experts to advise on the limits of BMI. This forum concluded that the limits should be set at 25 and 30, because research showed that in people with a BMI above 25 there was a clear increase in the health risks related to being overweight. Another argument was that these boundaries were easier to remember by the professionals who would be working with them . At these last cut-off points, no further distinction was made between men and women. These limits are currently the most widely accepted standard used in healthcare and scientific circles to determine overweight on the basis of BMI. The consequence of accepting these new cut-off standards is that men and women who are 180 centimeters tall may not weigh more than 81 kilos.

The BMI: not always reliable

Not all scientists agree with the decision to maintain these limits. Failure to take into account the above factors that influence the accuracy of body fat estimation can sometimes lead to very remarkable results. Information about the height and weight of many American celebrities is available on various websites of and about movie stars, in sports magazines and in promotional materials. If the BMI is calculated, it appears that quite a few American celebrities are overweight according to the set limits, which raises the question whether the fat percentage is indeed that high. Examples of this are:

  • ex-president George Bush (BMI 26),
  • Johnny Depp (BMI 27)
  • Brad Pitt (BMI 28)

The following Americans should be extremely overweight:

  • Mel Gibson (BMI 32)
  • Arnold Schwarzenegger (BMI 33)
  • Mike Tyson (BMI 32)

A number of scientists have doubts about the boundary between the category of light (BMI 25 to 30) and extreme (>30) overweight. Daniel McGee, professor of statistics at Florida State University, has analyzed the death data of more than 15% of a total of 388,622 people from different countries (including America) from 26 scientific studies. His research shows that more people who are extremely overweight die earlier than people who are not overweight. However, he could not demonstrate a link between premature death and mild obesity. In 2008, a team of scientists led by Gregg Fonarow, professor of medicine and director of the Center for Heart Disease at the University of California Los Angeles, analyzed research on lifespan and obesity in healthy people and people with heart disease. The results of the 9 studies, which involved 28,209 people, show that overweight (including extremely overweight) people with heart disease had a greater chance of survival than non-overweight people.

What should you do as a parent with the BMI?

In the Netherlands, adjustments are made to the cut-off points for the BMI of children from 2 years of age. These can be found on various websites. Especially given what has been described above, as a parent you can never rely solely on this when it comes to the health of your child. It is therefore unwise to put your child on a diet based on BMI. If you, as a parent, are concerned about your child’s weight, everything important will have to be examined, including:

  • What is the growth curve: what data is known about previous measurements?
  • Is there an accelerated increase in BMI from age 6?
  • Is there a genetic predisposition?
  • How much discomfort does the child have from being overweight?

It is best to always discuss this with your GP or school doctor. This person can determine the next steps together with you.

read more

  • Governments, obesity and children
  • Can you change eating habits?
  • How does taste preference arise in children?
  • Parents, children and obesity