Why BMI is not accurate for taller populations

Lisa Wade (at SocImages) links to a really good, publicly accessible run-down on the various environmental reasons why human (and animal) populations have been growing heavier. The author makes a pretty quick equivalence between obesity and BMI by (effectively) just equating the two, thusly:

And so we appear to have a public consensus that excess body weight (defined as a Body Mass Index of 25 or above) and obesity (BMI of 30 or above) are consequences of individual choice.

Now, I believe that there are two germane debates when it comes to the issue of BMI, obesity, and health: (1) the validity of applying a population metric to individuals (which I contend is categorically, logically, and methodologically incorrect) and (2) the question of whether BMI is actually a useful metric of our population. I would contend that the author focuses mainly on studies that presume that #2 is true, which then allows him to focus on answering #1. (I will admit that it's far more juicy and fun to look into #1, but is epistemologically lazy.)

Before continuing, let me first state that we cannot disagree with the fact that the average weight-to-height ratio (otherwise known as the BMI) has been increasing over time. This is a fact. It has countless reams of documentation that support it. Arguing that it hasn't happened is akin to shouting imprecations of denial against the existence of the sun while standing in the middle of a desert at high noon. In short, the numbers are irrefutable: the BMI ratio is increasing. Indeed, the article agrees that this is happening, and does an excellent job of examining the potential causes of the increase in BMI. But what the article misses is the very question of whether BMI itself (and therefore the mechanical definition of "obesity is when BMI>30" that no doubt underlie most of the studies in this area) is actually a useful metric to use on our current population.

The reason why I bring this up (almost once each year on my blog) is that many people breeze past the fundamental question of whether BMI actually is doing the job we assume it is (supposed) to be doing, which is (at the very least) providing a description of the relative height-to-weight ratio of the study population. Indeed, to the issue of the misapplication of the BMI, I will only repeat my position that it's categorically, logically, and methodologically incorrect. (I'll let Devlin explain why.) Instead, below, I'm going to point out one fundamental flaw in the BMI itself: height.

Taller people have a bigger BMI than shorter people of the same build. This is an important (but unsurprising) statement. Also important (but also unsurprising) is that a person can gain weight (or lose weight) without gaining (or losing) height. All this means that height is the independent variable in the description of BMI.

Furthermore, looking at the formula for BMI (weight/[height^2]) it's easy to note that increases in height (say 6 inches/~15 cm) will have a far greater impact to the BMI than an increase in weight (say 6 lbs/~2.5 kg), since height is squared, while weight is not. This means that - as the population gets taller - in order to remain below a BMI of 30, the population must weigh relatively less than a shorter population. Why? It's because the relationship of height-to-weight that forms the BMI was derived from data about 19th century Belgians (who were 5'5" tall, and for whom the formula of wt/[ht^2] was adequate).

But things have changed - a lot - in the height department.

Specifically, humans have become taller than those 5'5"-tall, mid-19th century Belgians. Unfortunately, though, the BMI calculation we use is still that same one derived over 150 years ago (which, btw, was not meant to be a measure of health). But what does this mean? It means that - as a population gets taller - it must become relatively lighter in weight (i.e., far thinner) in order to remain at the same BMI. For example, let's see what happens when we compare the two 2008 US presidential contenders: Barack Obama and John McCain:

Barack Obama: 6' 1.5", 180 lbs = 23.4 BMI
John McCain: 5' 7", 165 lbs = 25.8 BMI

True, Obama was (and remains) thinner than McCain, but - to show how much height influences the BMI calculation, compare, McCain to (then) outgoing president, George W. Bush: at 5' 11" and 190 lbs, George W. Bush had a 26.5 BMI, which is 0.7 higher than McCain's, but was (without much argument) thinner than McCain. So, if Bush was thinner than McCain, but had a higher BMI, and Obama has a slightly lower BMI than McCain, but is a LOT thinner, why is this the case?

The reason why the BMI fails at predicting taller populations is that BMI is fundamentally measuring the relationship incorrectly. For people who are around 5'5" tall (like those mid-19th century Belgians for whom the BMI was calculated originally), then the relationship is actually pretty good. However, with significantly taller (and significantly shorter) populations, the BMI falls apart, because - at these heights - the BMI is no longer acting as an accurate measure.

As Devlin (over at Devlin's Angle) writes:

The BMI was formulated, by a mathematician, not a medical physician, to provide a simple, easy-to-apply mathematical formula to give a broad, society-level measure of weight issues. It has absolutely no scientific or medical basis. It is based purely on a crude statistical analysis. It measures a general society trend, it does not predict.

I would put it one further: saying that "it measures a general society trend, based on the assumption that we can extrapolate from a statistical relationship found among mid-19th century Belgians; it does not predict."

And - in general - populations have been growing taller than mid-19th century Belgians. Especially in Asia. For an example of the change of Japanese average height over time, see here. And this means that these populations are moving away from that range of heights at which BMI was calculated, and for which - at the population level - BMI is/was relatively descriptive.

I'm not even going to get into a discussion about how screwed up BMI gets when you look at populations of athletes. (Hint: the US Olympic team is almost definitely in the "overweight" or "obese" categories, since most of them have a higher-than-average muscle density.)

My other entries on BMI (many of which were motivated from reading SocImages) are here

History and misuse of BMI

In response to yet another post over at SocImages, I delved even deeper into the coining of the term BMI and how it became used in the health world. I have previously written about BMI (here, here, and here), but I never really delved into how this unit had become a part of looking at population health (specifically obesity). Well, it was because I like participating in the commentary in SocImages that I started to look even further into the history of the thing. Using a lead from Wikipedia's entry on BMI, I found the 1972 document ("Indices of Relative Weight and Obesity") that first coined the phrase "body mass index" (the authors didn't even use the intialism in the article). I was pleased to note that the authors recognized that the body mass index should not be used to describe the individual:

What we here call the body mass index, W/H^2, has a long history. Because Quetelet was the first to calculate that ratio, W/H^2 has sometimes been called Quetelet’s index. But Quetelet himself did not actually advocate that ratio as the general measure of ‘build’ or of adiposity; he merely noted that in young adults W/H^2 was more stable than W/H^3 or W/H with increasing height. ... [No] proponents [prior to this paper] offered a convincing objective analysis in favor of the [Quetelet] index. Further, it should be observed that the greatest emphasis in almost all of the index making of the anthropometrists was on growth with relatively little consideration of the evaluation of body composition, nutritional status or adiposity.
...
As noted elsewhere [43], the use of ideal or recommended weight confounds age and weight because on the average weight increases with age until the fifties while increase in height is over by the early twenties at the latest. The general trend to continue growth in weight may be undesirable but it has no relevance to the question of providing an objective description of relative body mass; it is scientifically indefensible to include a value judgement in that description. The characterization of persons in terms of desirable weight percentage has resulted in attributing to ‘overweight’ some tendencies to ill health and death that are actually only related to age [43].

Of course, the disappointing thing about some of the commentators is that they don't understand that some terminology is quite old. The BMI was originally referred to as the Quetelet index (named after the man who conceived it), which was formulated in the early-middle 1800s; about 150 years before the paper that first coined "body mass index". Of course, some people apparently didn't know this. "EschewObfuscation" wrote (and at least 1 person "liked"):

But the whole purpose of BMI was to define fatness and thinness of a population. And why do that? So that doctors could talk to patients about their weight. Why? Because you shouldn't fall out side the norm or desired BMI number/category. Who decided normal? Who decided desired? How were those decisions made?

and

It is a social construction because the names we put on those numbers have social meaning. Why not call them "flower", "blue", "headphones", "towel" and "tricycle" instead of “underweight,” “normal,” “overweight,” “obese,” and “morbidly obese"? Or why provide categories at all and just use the number? Really what use is BMI anyway?


The words attached to numbers have social meaning that affects the society. Only one group is called 'normal' and that communicates a lot to the rest who don't fall in that category. Morbidly obese means "OMG!!! You are going to DIE!!!!11!!!!!!11" It is not a phrase that is purely objective.

Wow. That really did anything but eschewing obfuscation. It failed to approach and clear obfuscation in favor of allowing obfuscation to remain in order to justify what appears to be a normalized rationalization of what BMI is: an unjust, unobjective, and possibly something with an agenda to hurt people.

Now, it is, admittedly, a slightly difficult problem with regard to the idea of "objectivity", since its use for determining an individual's health condition is not objective, but many people fail to understand (quite apparently) the scientific use of BMI as it is used in public health research. My attempt to show why it is both an objective measure and a non-objective measure:

Also, BMI is "Completely NOT objective"? Remember, "objective" here means that it is not influenced by personal feelings, interpretations, or prejudice. Is measurement using standardized lengths (such as feet, meters, pounds, kilograms) not objective? Is dividing weight by the square root of the height (and applying the units correction of 703 if you did your measurements using feet and pounds) not give you an objective output? The measurement and calculation of BMI is objective.

However, is it useful for interpreting individual health conditions? No. (In this way it suffers from interpretive bias, is layered with social interpretation, has caused massive influence to personal feelings, and is - in this way only - not objective.)

In sum: BMI wasn't developed for doctors to talk to their patients about their weight. BMI wasn't used for studying obesity in a population until the 1970s (roughly 150 years after it was developed). BMI is a number calculated objectively based on objective measurements. BMI is next to useless in determining individual health conditions and suffers (for whatever reasons) a lot of social interpretations and personal feelings and is (in this final, falsely applied manner) not objective. Therefore, BMI is an objective, mathematical measurement and BMI - as it is come to be used - is non-objective social construction. 

The blog entry itself is a great article about a book (How Much Do You Weigh?) that shows women of various different body types (heights and weights) and different BMI values. It can, hopefully, diminish the amount of negative pressure that surrounds the (ab)use of BMI in the public by showing women the vast panoply of what it means to be a "21" or a "27.4". (There was also a link in a comment by Tracy Rohlin to a posting at Jezebel that looks like a great website that does a similar thing.)

Now, in the end, do I think that BMI is an evil, unobjective, completely useless measurement? No, but that's not what it is supposed to be used for. At the same time, I don't think that BMI the best thing in the world for all population level analyses (nor the best thing in the world for population level analysis). However, it is a useful population measurement that can show trends over time, and looking at current, historical, and ongoing trends is a very important part of a lot of the sciences that we have. BMI is a next-to-useless metric to help all individuals understand - by the BMI alone - what their overall health condition is, not because it is a "bad thing," but because it is fundamentally not meant to be used in that manner. (Although it is probably a little more precise than one's daily horoscope.)

All that being said, the use of obesity to punish yourself or others is just wrong. Partly because it is based on a fundamentally flawed understanding of the application of a population level index, but mostly because using it to punish people is morally wrong.

On BMI

I have a rough BMI of 30. What does this mean? It means that I'm in the "Obese" category, as defined by the CDC. There is no category above mine.

But wait, am I obese? I don't think so... I cycle 8 miles each day, do manual labor, and eat (relatively) healthily. I haven't checked my resting heart rate recently, but I'm guessing that it's between 60 and 70. I have a 36" waist, and 48" shoulders. I can benchpress 200 pounds, easily leg press 500 pounds, and do calf extensions at 250 pounds per side. How is this obese?

Recently, Devlin's Angle did a piece on BMI - the history, and stupidity of it. On the one hand, the usage of the BMI to indicate a single person's relative health is a good example of misuing a [simplistic] formula meant to determine population level characteristics, not individual ones. On another hand, this is also a story of how numbers and scientific wrappings seem to hold social significance. On yet another hand (this is turning into a Vishnu-statue of "other hands"), the BMI is a major tool in looking at trends in obesity. Finally, on the remaining hand (of Vishnu), it's completely meaningless, mathematically speaking.

Here's the equation:

BMI = weight in pounds/(height in inches^2) x 703

Devlin's Angle goes on to explain where the 703 comes from, and ponders the question of why the height is squared...

Devlin's Angle outlines two reasons why the BMI is useless as a individual measure: it's a population-based measure, and populations are made up of sedentary individuals (not atheletes) and it was derived at an early time in the statistic-ization of sociology. On the first point, the formula assumes that all "extra" weight on an individual is from fat. On the second point, the formula is derived to measure the trend of the majority of the population.

I would like to outline two other reasons why the BMI sucks for people like me: height and body proportion. Devlin's Angle points out that the BMI was derived in the early 1900s in Belgium. Looking at Wikipedia's entry on human height, one finds that in the mid-nineteenth century the average height in the Netherlands and France (they don't list Belgium) was164 cm and 165 cm, respectively. Looking at these measures "today", one sees a "slight" difference: 182 cm (169.7 cm) and 177 cm (164.6 cm), respectively (female heights in parenthesis). I assume that female heights weren't included in the mid-nineteenth century measurements, but we see that after 150 years, Dutch females are taller (males much taller) and French females as tall (males much taller) than their ancestors. What is so important about this? Well, remember that the equation for BMI was based on the average person. The average person being (among males) as much as 12-18 cm taller than the people measured to derive the BMI.

Let's try and make two individuals who matches the criteria of different BMI groupings; one from the mid-nineteenth century, and one from today. Since BMI is only a relationship between height and weight, this shouldn't be difficult to figure out. Therefore, a mid-nineteenth century Dutchman of average height (roughly 65 inches), would have the following BMI table:

Underweight (below 18.5): below 110 lbs.
Ideal (18.5 to 24.9): 110 lbs. to 148 lbs.
Overweight (25.0 to 29.9): 148 lbs. to 178 lbs.
Obese (30.0 and above): 178 lbs. and above.

We can imagine the "average build" 5'5" person and think, "Okay, that works." However, now let's look at what the BMI chart would mean for the average height Duchman of 6'0":

Underweight (below 18.5): below 135 lbs.
Ideal (18.5 to 24.9): 135 lbs. to 183 lbs.
Overweight (25.0 to 29.9): 183 lbs. to 219 lbs.
Obese (30.0 and above): 219 lbs. and above.

That man would have to be one skinny person. Imagine a person who is 6'0" and 135 lbs. Jim Carrey - someone we might think of as tall and skinny - is according to this site - 6'1" and 180 lbs. That's on the upper end of "ideal". President Barack Obama is - according to this site - 6'1.5" and 180 lbs. (Which, strangely, gives him a slightly higher BMI of 23.4 compared to John Kerry's 22.5, even though I would say that Obama's more atheltic now than Kerry was in 2004).

So, height is a determinant. Therefore, BMI was a good measure of estimating the height/weight relationship of early 1900s Belgians, not necessarily early 21st century Americans.

Also, there is body proportion. Having a relatively long torso, I have been blessed with not having to worry about too little leg room on aircrafts (yet), but I am annoyed at how low the backs of seats cut me (usually well below the shoulder, even on "tall" chairs). There are trends on body proportions, both in terms of proportions of height (i.e., long torso vs. long legs) and width (e.g., hip-to-waist ratios). Neither of these are included in the BMI. True: certain regions have slightly different cut-off points for BMI measurements, such as in SE Asia where there is a relatively consistent height body proportion, however, in the United States - where there is such a wide range of different height and width body proportions that BMI loses its meaning even faster. But why?

Well, take me for example. I have a long torso. What does that mean, though. Well, a single inch of height that is comprised of torso has more mass than an inch of height comprised of leg; there are more organs, more girth, and more water in the torso than in the leg. Therefore, if someone has a relatively short torso compared to me (i.e., they are my height, but have much longer legs than me), having a 240 lbs. somewhat athetic build would mean that individual would have legs even more massive than mine (or a torso that was gigantic). In other words, body proportions will have a direct bearing on the amount of mass you are carrying around (and therefore, your weight), irregardless of how much muscle or fat you have.

Related to the issue of body proportion is the issue of amputation. If you amputated a leg, you could lower your total body weight by 20-40 pounds (depending, obviously, on the weight of your legs). If I cut off legs, then my BMI would drop to about 25 (still "overweight", but not "obese"). However, if you had both your legs amputated (or weren't born with both legs), then you could shorten your height by 27" (plus or minus) and your weight by 40-80 lbs. However, due to the formula, my BMI would actually increase to 44 if that happened to me.

What to do? Well, I have a novel idea of doing a BMI survey of a cohort of incoming students, combined with a number of other measures that will give an indication of body proportion, muscle mass, and fitness. Then, look at doing statistics on the whole set to see what sort of relationship one would get for "modern" BMI cut-offs, and determine if there is any easier way of determining a general figure of fitness that can be used by an individual (as opposed to being a proxy for an entire population, mostly made up of sedentary individuals).

My comment (on another blog) about obesity, BMI, and "blame"

Over at Sociological Images, there is a post showing a 1967 public service announcement (PSA) called "To Cure a Fat Child is not a Simple Matter".

One of the commenters there - attentie - wrote the following:

I think the focus should lie on promoting a healthier life style rather than thinner bodies at any costs. The real epidemic is not the fact that people are getting less atractive according to the modern beauty ideal (thin= beautiful) but that people are less active and eat more junk food and candy. Unhealthy diets or excercise routines (too much excercise can be very bad for your joints etc) are no solution. Little things, like walking or taking the bike to work instead of the car or eating vegetables and fruits every day, it all helps to stay healthier. If you lose some weight doing that, that is great, but it should not be the focus point IMHO.

What follows is my response to that comment (and, to a lesser degree, the issue of who is to blame about a fat child).

IMHO, too. Metrics of “thinness” don’t work for me; even though I might think of myself as not being “obese”, that is what I am medically classified as being (6′3″, 243 lbs = 30.4 BMI). (But then again, many athletes have very high BMIs: when he was still in pro-wrestling Dwayne Johnson – aka “The Rock” – was billed as being 6′5″, 275 lbs = 32.6 BMI.)

I was – a year ago – feeling that I was getting a little too chunky for my jeans (was up at ~260lbs), and so (instead of buying new jeans), I bought a new bike, and started cycling my commute instead of busing. I’ve not gone to the gym, except to do yoga (but stopped doing that about 5 months ago), and the only thing I do for “exercise” is ride my bike and go for walks.

I also changed up my diet, and eat out less often than before, but still enjoy a weekly beer night with my friends (don’t worry — I have one or two beers over four hours and never cycle home if I don’t feel safe). The eating out less has not only saved my budget (letting me pay off my bike a lot sooner), but it (I believe) helped slim me down, since I don’t make the heavy cream sauces that I am so partial to when dining out. In addition, since I make a lot of my own food, I am very critical about what goes into meals, and try to be more balanced with my ingredients.

Technically, I’m still “obese,” and in order to move all the way down through “overweight”, I have to either become shorter (not likely happening) or I have to get down to 200 lbs (something that didn’t happen, even when I was in peak physical condition in my early 20s). Despite all of this, however, I still feel that I am “healthy.” I am physically able to do a lot of the things that I like doing, I don’t have to wait several minutes to catch my breath after exertion, and I am happy with my own body image.

Why all that discussion? Well, I wrote up a blog entry about BMI a little while ago, looking at what its implications really mean, and came out with the conclusion that I don’t fit inside the general population parameters for which the BMI works: I’m either too tall (I am much taller than the original subject population from which the BMI was derived), of the wrong body proportions (I have an Asian build of a relatively longer torso and relatively shorter legs, meaning that my weight distribution is different than the original subject population), or more athletic (I’m not really a sedentary person, but tend toward athletic).

Furthermore, although the BMI seems a “scientific” and “objective” measure of obesity because it produces a number (and we tend to have social relationships with numbers that imbue them with an air of objectivity), it really isn’t. It’s based on a generalized trend of height and weight seen among sedentary male Belgians from the mid-19th Century. In addition, that relationship is a description of the general trend of that population, and not a predictive statement about any one person who is either in that population or not in that population.

The BMI still persists, though. Why? Well, because population trends that have a predictive ability at the population level tend to be useful (and BMI still is a good population-level predictor of health risk). It’s also useful because it is easy to calculate, requiring very little in terms of time and money to collect the data (as opposed to lipid panels, %fat calculations, etc.). And (perhaps most annoyingly) it remains a standard, and is thus something that is difficult to overturn.

What does YOUR BMI calculation mean, though? It could be spot-on (e.g., if you are over 30, you may well be obese), or it may be highly deviant (as in the case with Dwayne Johnson). This is called variability, and is one of the reasons why back-calculating population-level statistics to an individual is a statistical no-no (no matter how often we do it).

At the end of the day, is my mother to blame for me being 6′3″ and 243lbs? Well… since I’ve not been living at home for over 14 years, I would say… no. Also, since I’ve (hopefully) shown that BMI categories are socially constructed and have very statistical relevance when assigned to an individual, then I would say that even when I was 18 and going off to college (all 246 lbs of me), she still wasn’t “at fault” for making me “obese.”

UPDATE (11/17/2011 @ 11:28PM): The property of "objectivity" is a little difficult in the case of BMI. BMI is an objective measurement that it isn't subject to personal feelings, interpretations, or prejudice. The measurement of height and weight (if taken properly) is objective. The calculation of BMI (using the formula) is also objective. The collation of height, weight, and BMI data from large groups of people is also objective. The analysis of BMI data as population data is also objective. However, the interpretation of the the population level metric to that of the individual (which is what I'm talking about here) isn't objective, specifically because it is subject to personal feelings, interpretations, and prejudice.

Processed food, obesity, health care, and societal costs

I suppose food is on my mind today (although I don't really know why, since I've eaten enough)... However, this graphic from Next Generation Food got me thinking about something else in the health care debate - something that isn't discussed: the health cost of the food we eat.

A 2005 (I believe) publication from the American Heart Association showed that 30.4% of adults in the United States were considered "obese," based on BMI. (Now, while I do have an issue with the use of BMI as a metric for individuals, at a population level, its trend does seem to work better.) What I want to draw attention to, though, is the link that the AHA makes between obesity and health and financial consequences:

The publication lists correlations with life expectancy and obesity, and has a nice table (on page 14) of increased likelihoods of diseases, too. For example, for people with a BMI greater than 35, there is a:

• 6.16 times greater possibility of developing type-II diabetes,
• 5.48 times greater possibility of gallstones,
• 3.77 times greater possibility of having hypertension,
• 2.39 times greater possibility of arthritis,
• 1.75 times greater possibility of stroke, and
• 1.67 times greater possibility of heart attack..

With regard to financial costs, the publication cites a 1999 paper ("The costs of body mass index levels in an employed population") that showed that as BMI increased, so did the number of sick days, medical claims and insurance costs. Furthermore, they cite a 2002 paper ("The Effects Of Obesity, Smoking, And Drinking On Medical Problems And Costs") that showed:

Obesity is associated with a 36 percent increase in inpatient and outpatient spending and a 77 percent increase in medications, compared with a 21 percent increase in inpatient and outpatient spending and a 28 percent increase in medications for current smokers and smaller effects for problem drinkers.

How does work in with the health care debate currently going on in the United States? I can see it working itself into the conversation in two ways: current denials to those who are BMI-obese and future population-level costs if it isn't effectively addressed in the future. In the first case, the husband of a friend of mine was declined health insurance because his BMI was too high. This is why I use the term "BMI-obese," since I am categorized as "obese" by the BMI table. (My previous entry on BMI talks a lot about the problems with BMI as a modern-day scale as well as the logical problem of using it as an individual measure.) Since he cannot get medical insurance, he is one more of the 40 million Americans on the uninsured lists due to a "pre-existing condition". It is likely, too that many people who are BMI-obese may have their insurance dropped if this fact is found out, or may have it drastically increased.

In the second case, let's assume that pre-existing conditions cannot be a cause for dropping (or not enrolling) a person to health insurance. Under these circumstances, the costs of the truly obese (as opposed to BMI-obese) will be borne by all of the payers. However, if rates of obesity continue to increase, then that cost burden will also become greater and greater. If there is not mechanism to award people who are healthy (as opposed to only penalizing people who are not healthy), then the financial problems of obesity will not go away.

What does this all have to do with processed food? Well, in addition to what the numbers and charts show in the graphic above, food purveyors want their customers to purchase their product. What manufactured food does is prey upon the human evolutionary desires of sugars and fats, and thus give us sugary drinks and fatty foods, which we (in turn) consume with all the evolutionarily pressured gusto we can manage before going back for more (and more and more). This cycle tends to lead to obesity in a population and (if left unchecked) obesity of a population, which (in turn) leads to increased health care costs as well as increased macro-economic costs.

That's enough navel-gazing for right now on this topic. However, it's likely to not go away from my mind any time soon...

Once Again: On BMI

It seems like I write something about BMI each year (2009, 2010, 2011), save (strangely) last year. The main thing that I keep pointing out is that BMI is:

1. An objective measurement that is
2. Used improperly, which leads to
3. People imbuing tons of social construction into BMI, because
4. Objective numbers (for some reason) are psychologically powerful (probably because they can't be refuted on their own).

Well, we should all know by now that BMI is definitely not a measurement of an individual's health, of fat, or or anything else other than weight/(height^2).

And what is the significance of weight/(height^2)? About as much as the significance of:

ln(Q) = ln(A)(0.846)+ (P)(0.001)+ (%Himalaya)(1.001)− 4.698

which provides the user with the annual average water discharge of India's Ganges River, based on the total upstream watershed area (A), the annual precipitation of the upstream watershed (P), and the % of the upstream watershed that is in the Himalaya mountains. It's actually a highly predictive formula, with an R-squared of 95.5%, which is - for such large-scale modeling - pretty damn good for using only three variables.

However, this formula is next to useless in determining what the flow of the Ganges river is in any one year, in any one season, or during any one day. Why? Because it doesn't actually measure any of those.

And BMI - much like the simple equation I derived for my master's work - does not measure health or fat simply because it doesn't actually measure those things. It is merely a statistical regression equation that is based on a specific population (19th Century Belgian men) and split into arbitrary categories that mapped (during the 1970s or 1980s) onto concepts of health and obesity.

The BBC World Service did an episode on the efficacy of the BMI and a possible new equation for calculating BMI. (The story did a good job of also showing why the new formula is about as useful for giving individual advice as the existing formula.)

Obesity, diabetes and an image problem

In a recent story from the People's Daily, more young Chinese are becoming diabetic. The story goes on to ascribe this trend to the changed lifestyle that many young urbanites have switched to:

Two years ago, Ke Li, was living a life typical to many office workers. He frequently worked until late at night, ate junk food and hardly did any exercise. The 6-foot-tall young man never thought he would be a candidate for diabetes.

"Who would pay attention to blood sugar at that age?" 27-year-old Ke told the Global Times. "I went to hospital because I had fatigue and always felt thirsty, even in summer, but I was soon told I had diabetes after a blood test."

Ke was not aware that his growing weight, which had hit 94 kilograms when he was diagnosed, was a warning.

Okay... so this isn't too surprising. After all, type 2 diabetes has been linked to obesity, and obesity has been linked to unhealthy lifestyles (not enough exercise, eating junk food, high stress, etc.). It would, therefore, make sense that, in a country with an urban population of several hundred million people exercising less and having more ready access to Western junk food, type 2 diabetes would become more prevalent.

So what makes this something more interesting? Well, according to the story, Ke Li is 6 feet (~183 cm) tall, and weighed 94 kg (207 lbs), which gives him a BMI of 28.1 (if I use 183cm as the actual height). Assuming that the guy's height is as low as 177 cm (which he rounded to 180 cm when talking to the reporter, who then rounded up to 6 feet when writing the story), the guy's BMI is exactly 30.0. At the end of the story, Ke Li is reported to have lowered his weight to 78 kg... which would (using the 177 cm low estimate) put him at a BMI of 24.9: right at the cut-off point of the "normal weight" category. Why point this out? Well, the People's Daily is a state-run newspaper, and it seems kind of odd to me (a 6'3", 226lb person) that the main diabetic interviewee had a BMI that just barely (possibly) registered as "overweight" before just managing to sneak back into the "normal weight" category. Similarly, the image that accompanied the story (below) was of a young, female urbanite that clashes with a Western image of "obese".

 

This image was copied from the People's Daily news story, in case it is no longer shows up in the future. (Not that I think that my blog is that important or visible, just that some websites remove images from stories after a certain amount of time passes.)

For a country to which image portrayal is highly important, it wouldn't surprise me if there was some amount of editorializing with the numbers and choice of image. Depicting a woman who stretches the image of obese and presenting a man who barely wanders into the "obese" category of BMI might - independently - not strike me as too odd, but not when they occur together.

China - I would argue - doesn't like to be put in the spotlight with things that it finds embarrassing. Thus the angry rhetoric that accompanied the Nobel Peace Prize announcement. Thus, too, the huge embarrassment about the tainted milk scandal that just couldn't be handled quietly. Same with their condemnation of anything written with even a slight whiff of negativity about Tibet. I would argue that there is some of that coming through here. Although there are explanations as to the current Chinese obesity increase being tied to traditional Chinese culture (which was shaped by famine, and in which an obese person would be an obvious sign of wealth), the association with obesity in the traditional sense focused on the positive implications (i.e., more wealthy) than on the negative (i.e., increased risk of heart attack, stroke, or diabetes). In the light of this potential link of obesity with disease, it makes sense to me that the images (in text and in print) proffered by the state would show as rosy a face as possible, as opposed to the images one finds when doing a search for "obesity China".

A study linking obesity and public transportation

I've previously written on this blog about my conjectures that there is a connection between obesity and the availability of public transportation. The idea behind that conjecture was that public transportation makes the rider burn more calories (e.g., from walking to the station, walking between transfers, standing on the train, balancing through turns and starts/stops, etc.) than if that person were driving. However, the best piece of "evidence" that I came across was a correlation of obesity with the percentage of people who walk, bike, or use public transportation across countries. As I mentioned with that piece of evidence that the data used for the relationship doesn't work with categorical data, such as country, but that the evidence was interesting.

However, there are too many external factors that are wrapped up in such a relationship between countries, including social conditioning (the more obese people in your social network, the greater your chance of becoming obese), food consumption patterns (use of high fructose corn syrup, GMOs, etc), gasoline and car taxes (high gas and car taxes keep car ownership low), potential distance to be traveled (smaller socially perceived distance horizons imply a shorter travel distance), etc., etc. Therefore, a much better analysis would be a pre-test vs. post-test analysis of people in a city before (pre-test) and after (post-test) the installation of a public transportation network. Furthermore, the public transportation network needs to have wide-spread social buy-in. In other words, not buses, which get caught in the same traffic snarls as cars do, making their use less beneficial than driving (for those who have cars). In other words, a more conclusive study should try an apply an experimental design. Yet, how can one just do that?

Well, it works when you do a pre-test and post-test of a newly installed public transportation network in a major United States city: Charlotte, NC. The study in the American Journal of Preventative Medicine showed:

using light rail for commuting was associated with reductions in body mass index (BMI) over time. Specifically, LRT reduced BMI by an average of 1.18 kg/m2 compared to non-LRT users in the same area over a 12-18 month follow-up period. This is equivalent to a relative weight loss of 6.45 lbs for a person who is 5'5. LRT users were also 81% less likely to become obese over time.

Public transportation seems to work, folks. Still, it would be interesting to look at how these rates change as the network expands, as well as including how quickly people in currently unserved areas change their BMI as they get served by an ever-growing transportation network. (My thought is that the rate of BMI decrease will be faster as the network size increases, although the rate of decrease will be tangentially limited.)

UPDATE: In a recent study by Trust for America's Health, Michigan is rated as being the 10th fattest state in the nation, among adults (but 41st fattest among children). I don't know what their methodology was, but it seems to match up with some other things I've seen elsewhere. Anyway, if people want to use the evidence of the Charlotte, NC study to link up with obesity rates in the state, then it might be a new and interesting way to organize efforts to actually get more movement on regional light rail in Southeast Michigan. Of course, with Michigan being the traditional car capital of the world, this could still be a non-starter...

Population weight, public health, and environmental implications

I watched the following video



which brought up several points of concern:

1. The total adult population of the Earth weighs roughly 287 million metric tons (316 million US tons).
• This is about 1/2 of the total biomass of the world's cattle
• This is about 1/5 of the total biomass of the world's ants
2. The world's average adult body mass is 62 kg (137 lbs).
3. The North American average adult body mass is 80.7 kg (178 lbs).
4. If the world's adult population all had the same average mass as the North American average, it would be like adding 935 million people to the current world's population.

The video goes on to explain why this is problematic for environmental and public health concerns (starting from 1:14). The original article - "The Weight of Nations: An estimation of adult human biomass" - can be found at BMC Public Health.

Abstract (Background):

The energy requirement of species at each trophic level in an ecological pyramid is a function of the number of organisms and their average mass. Regarding human populations, although considerable attention is given to estimating the number of people, much less is given to estimating average mass, despite evidence that average body mass is increasing. We estimate global human biomass, its distribution by region and the proportion of biomass due to overweight and obesity.

Abstract (Results & Conclusions):

In 2005, global adult human biomass was approximately 287 million tonnes, of which 15 million tonnes were due to overweight (BMI > 25), a mass equivalent to that of 242 million people of average body mass (5% of global human biomass). Biomass due to obesity was 3.5 million tonnes, the mass equivalent of 56 million people of average body mass (1.2% of human biomass). North America has 6% of the world population but 34% of biomass due to obesity. Asia has 61% of the world population but 13% of biomass due to obesity. One tonne of human biomass corresponds to approximately 12 adults in North America and 17 adults in Asia. If all countries had the BMI distribution of the USA, the increase in human biomass of 58 million tonnes would be equivalent in mass to an extra 935 million people of average body mass, and have energy requirements equivalent to that of 473 million adults.

Increasing population fatness could have the same implications for world food energy demands as an extra half a billion people living on the earth.

I (idly) wonder if anyone has used this sort of data analysis for actual trophic food web analysis, much like one might do with various fish species (e.g., here, here, here, here, here, here, here, and here; see also "trophic food web" on the University of Michigan's dissertations, theses, and research publications site).

Random health assessment: Resting heart rate

Just for shits and giggles, I decided to check my resting heart rate. I had been riding my bike as a daily commute, averaging 25kph to work and 22kph from work, and I wanted to see if there was a benefit to all this bike commuting.

According to topendsports, an average resting heart rate of someone 35-40 years old is 71-75 bpm.

My resting heart rate prior to re-starting my bike commute was about 70bpm (and I was 37 at the
time), which put me right around average, maybe slightly on the border with "above average." As a point of reference, my resting heart rate when I was a vasity swimmer in high school - at 16 years of age - was 47 bpm, which put me well within the athlete level.

Now, it's not surprising that resting heart rate will increase with age, but moving from an athlete level to average means that I knew what it was like, and 70 bpm seemed really fast. But now, my resting heart rate is roughly 55 bpm, which works out to being on the upper end of "athlete" for a man in my age category.

And that feels nice.

Maybe it is also time to check my BMI (with recognition of problems of height and muscle density) and my blood pressure?