Imagine, Dear Readers, my joy on discovering that that BMI, or body mass index, popularised by Ancel Keys in 1972 when medical people first started to become interested in obesity, was in fact invented by the Belgian Mathematician Adolphe Quetelet in the early 1800s.
As far as I can tell, Quetelet was just really into measuring things, and thought that if he measured enough things about people he might find out something interesting.
And he published, in 1835, his book "Sur L'Homme, et le Développement de ses Facultés" with all his measurements in.
It was very well received, and very controversial, and got an English translation in 1842. Which google books has scanned and you can get it as a pdf:
Quetelet was not at all interested in obesity, almost as if it didn't matter and no one cared and it wasn't a problem. Not to pre-judge the question.
But he was interested in human development, so he took lots of measurements of people's weights and heights at various ages.
And he doesn't seem to think it much matters if you measure the poor or measure the rich, his measurements are of schoolchildren (rich!), orphans in a foundling hospital (poor but probably pretty well fed!), medical students (rich!) and customers at various hospitals (rich!) although he does note that 'the lower orders have been least numerous'.
At the point where you have got 3% more food than someone who is literally balancing on the edge of starvation, which is all it takes to get to 'scooter obesity' by the time you are thirty, I think you probably still count as a member of the lower orders. So I think we can be confident that almost all the people in this survey had the calories available to them to be as fat as they liked.
If mere availability of calories caused obesity, these should be some pretty fat people in this sample, right?
Now on top of this, Quetelet has measured people with their clothes on, and this is northern Europe in the Little Ice Age, so those clothes are going to be pretty heavy.
From different experiments, I think we may admit, as near the truth, that the average weight of the clothes at different ages is one-eighteenth of the total weight of the male body, and a twenty-fourth part of the total weight of the female.
I think that the measurements in the big table on page 77 of my pdf
are the unadjusted weights with their clothes on, but it's not entirely clear.
If they are that's going to add about 1 point of BMI.
Anyway a bit of copy and paste and a bit of python produces this graph (while doing this I noticed that the 19yr old girls are missing. I've just ignored that and taken the data as written):
It looks like BMI hits 22.2 at age 20 for both sexes.
Since he's measuring with clothes on that's actually pretty close to the hunter-gatherer 21 figure and to the Victorian prostitute numbers.
By 25 years old men are up to 23 and women still at 22.1, which makes sense as young men are still growing in strength but not height until 25.
At 30, men are 23.2 and women are still at 22.1
But then women's weight starts rising rapidly and men's only gently, which doesn't make much sense to me.
By 40 both men and women are 23.4
At that point both sexes start to gain weight, women faster than men
At 50 men are 24.1 and women are 24.8
And now women's weight starts to fall slowly, while men's keeps rising
At 60 men are 24.4 and women are 24.7
After 60, both men and women start to lose weight, women rather faster than men
At 70 men are 23.9 and women are 23.4
Finally at 80, men are 23.5 and women are 22.7
He doesn't talk much about the variance in these measurements, which is a shame. I'd like to see his original measurements, but I imagine they're lost.
But it doesn't really matter, they're averages, for every person with a high BMI there's going to be one with a low one, roughly speaking.
Remembering that we probably have to take a point off these numbers for clothing, I'd summarise this as 'Overweight rare, Obesity non-existent'.
With what little overweight there is concentrated amongst the middle-aged, and probably even then not enough to notice without measuring.
Not one of these ‘average rich belgians’ is even close to my full-grown BMI of 27, which stayed stable from when I was 25 until it all went wrong at 40. I’d have been literally off the chart. Add one for heavy clothing and I’d have been way off the chart.
And I might have been technically overweight, a fact I used to laugh about, but you wouldn’t have thought it to look at me.
All these people would have looked a bit scrawny next to me. Maybe the biggest guys in the whole sample might have been built the same way. That makes sense. I wasn’t quite the strongest man in my boat club, but it was close.
This noticeably doesn't square with Victorian photographs, where the middle-aged women do often look rather overweight to me.
Buh-buh-but muh classical paintings of ever-so-slightly curvy women!
Thanks for bringing together this data. It's pretty obvious to anyone with eyes that widespread obesity is a modern problem but always nice to have solid numbers as well.
Funny about the 19 y/o girls. How come he has numbers for 19 y/o boys? Wonder what the story is there, lol.
Is this whole thing before or after the Belgians invented French fries, that ultra-palatable, hyper-processed junk-food?