Fatties live longer!

This isn\’t new nor a surprise. But it does show the problem we\’ve got with the BMI measure:

Dr Katherine Flegal, of the National Centre for Health Statistics in the United States, found that people who are overweight had a six per cent lower risk of death than normal weight people.

The risk for those with a BMI (body mass index) of between 30 and 35 fell by five per cent. But those grossly obese with a BMI above 35 were 29 per cent more likely to die than slim people of the same age.

Let us assume, for moment, that the aim of all of this is indeed to enable us all to live long and enjoyable lives (as opposed to the Puritans getting to impose their idea of the good life on us).

Logically we should change the advice we give people. We should be aiming for a BMI of 30-35, not the 25 we are currently told to aim for.

That this has long been known and yet the advice hasn\’t been changed rather leads to the idea that it is all about the Puritans really….

23 thoughts on “Fatties live longer!”

  1. Does eating lettuce and drinking water make you live longer?

    No but boredom does slow down the passage of time.

    Given that the stick thin look is neither the optimal for getting laid, nor is it a good calorie inventory policy, it is unlikely to be the best option.

  2. This is a meta-analysis, and the results are not surprising.

    What the analysis cannot tell us is whether it’s being thin that makes you die, or being sick that makes you thin. A study that controlled for relatively recent weight loss found that for people whose BMI hasn’t much changed, a normal BMI gives the lowest mortality.

  3. What exactly does likely to die mean here? Everyone, no matter what their BMI, is not only likely but sure to die. I have not read the study but do they mean more likely to die while they are at a certain BMI?

  4. BMI was invented by a belgian so no wonder it’s useless.

    It completely fails to take into account lifestyle issues. Very fit athletes such as sprinters often have BMIs in the overweight range range as BMI doesn’t distinguish between body fat and muscule when it comes to weight.

    Why this is considered a relevant measurment is beyond me

  5. Why does anyone take this stuff seriously? As you pointed out over alcohol and the 5-a-day nonsense, the health police just make stuff up to keep themselves in work, hanging grim predictions on flimsy evidence that always needs more research/money (like climatologists).

    In this household, we always invoke Woody Allen’s film ‘Sleeper’ where he wakes up 200 years into the future to discover that chocolate fudge cake is now health food…

  6. BMI does badly with fit people. Many rugby players have a BMI that says they’re overweight. I do, for example, though my waist measurement is 2″ smaller than my inside leg. Maybe that has something to do with it.

  7. BMI also doesn’t distinguish people who have broad shoulders, big chests or wide hips from those who don’t. And the tables I’ve seen don’t discriminate according to age or sex, so it’s all pretty implausible.

    Twice in my life I’ve put on weight rapidly: each time it followed falling ill.

  8. Mine’s 22 which is apparently perfect and I can assure you I’m the least healthy fucker on the planet. I basically just eat fast food, drink like a fish and smoke like a(n old fashioned) chimney.

  9. Does anyone actually think BMI is anything other than the broadest of brushes? Perhaps a few idiots in government, and those Guardian-reading types who believe absolutely anything, but surely no-one else?

    It’s never intended to be anything other than a slightly better way of talking about the ‘right’ weight for your height and build. No doctor would look at a rugby player and worry about their BMI being ‘too high’ any more than they’d have previously worried about a muscular chap being ‘overweight’. It’s based on an average physique, and clearly adjustments have to be made when someone has a body significantly different from the norm. What you can say is that no-one has a BMI significantly different from the ‘correct’ one for their height without the reason for it being immediately obvious.

    When it comes to epidemiology, the problems with that field are a different issue – but for epidemiological purposes, BMI is a more useful stat than weight alone.

  10. The BMI is crap for every single purpose. It so wrong that it doesn’t even take into account physics. Elementary solid geometry says that the volume of an object is equal to the third power of its linear dimensions, so the weight should be proportional of the cube of the height.

    It is so dependent on the population that it can’t handle ethnicity. “Healthy” japanese BMIs are different to African ones and are different to Western European ones.

    Its only purpose is to find out what the average of a population is. But to do the reverse and find out where a particular person is against this average is just wrong. It’s like forcing people to fit on a bell curve when in reality the bell curve comes from looking at people.

    And it so arbitrary in that the magic “healthy” numbers have been picked out of thin air with no scientific basis. So it pretty much throws the whole measurement out the window.

  11. SBML>

    Rather missing the point there. If the advice is that the ideal BMI is, say, around 25, then that doesn’t mean you take that advice and apply it with no variation whatsoever to every person. Instead, you are supposed to adjust it depending on how your physique varies from the average, such that it’s appropriate for you.

    BMI is not a particularly good measurement at all, hence the need for such adjustments, but no-one (worth listening to) says it should be used as such. If I can give the traditional (bad) motoring analogy, a national speed limit of 70 MPH is only debatably a good idea – but it’s ridiculous to pretend that anyone says that it means you must always drive at exactly 70 MPH.

  12. Elementary solid geometry is not sufficient in this case: weight is not typically proportional to the cube of height. Empirically, a power law of about two and a half fits the data best.

    Of course BMI is not the best way for an individual to determine whether they’re overweight. If you want to know, have a look at a picture of yourself in a not very loose T-shirt.

  13. PaulB>

    Indeed, I somehow missed that line. It’s a particularly daft claim unless one thinks humans are spherical. If you double the length of one dimension of a cuboid, you double the volume, not multiply it by eight.

    To put it the other way around, which is maybe more fun, how much greater in volume would a person of 6’3″ have to be than a person of 6’0″ to have the same BMI? Roughly, a sphere 10″ in diameter, which is plainly significantly too great. (A trim person of that height would have a trunk circumference of about 32″ at the widest point. Even all the height was added at that point, and even if the cross-section was circular, which it’s not, we’d still be well short of the required volume after multiplying by the 3″ of extra height.)

  14. When I was about 6’3′ with a trunk circumference 32″ I was obviously underweight because you could see every bloody rib.

  15. Dearieme>

    That was a guess. 34″ suit you better? It doesn’t make any difference to the point. Treating the chest measurement as the circumference of a circle was very, very generous already, but even without that there’s more than enough volume unaccounted for to add a few more inches onto the chest.

  16. The volume of two three dimensional bodies of the same shape, but of different hight varies in proportion to the cube of the hight, regardless of whether they are cubes, spheres, statues or the nameless shapes taken up by human skin, human organs human bones etc. If two people are of the same proportion but one is twice as high as the other then he will also be twice as wide and twice as thick- or they are not of the same proportion. Simplest way of proving this would be to use the calculus- divide these bodies into a collection cubes approximating to their actual shape, then simultaneously reduce the size and increase the number of cubes to give a better and better approximation.
    But is the ideal proportion for a six foot man the same as that for a three foot dwarf? The legs of the six foot man need to be eight times stronger than the dwarf, so presumably they need to be heavier. One assumes that other parts also need to increase disproportionately- the heart has a lot more pumping to do for example, but then, does the brain and hence head need to get bigger? Hence I doubt that the ideal weight for the six foot man is exactly eight times that of the three foot dwarf- but that should be a better approximation than the square.
    And it is certain that the average hight of people in Belgium and round the world has increased in the last two centuries, so the difference between use of the square and the cube matters if we are using this to estimate the general “fatness” of the population. It also seems likely that malnutrition was more common in Belgium in the 1830s than today, hence affecting the view of what is normal.
    I know for certain that the only time I ever got my weight down to the middle range for my height I looked like a skeleton, was weak as a kitten, and lacked the strength to run. God knows what it would have been like at the bottom end of normal.

  17. Got a look at the study and they used 18 as normal. With my height of 5′ 9″ that would be a weight of around 130 pounds. Given my general weight of 160, I could easily drop to 130 as the result of some type of fatal disease then die within the year. Seems like the study is saying obese people die more quickly and a group including people with terminal diseases die earlier. Hmmm.

  18. Pat>

    Aren’t you rather demonstrating that indeed it is not ‘elementary solid geometry’? Since, as you agree, human bodies do not vary in strict proportion to height, no figure is accurate – and the average exponent is not an integer, but closer to two than three.

    “And it is certain that the average hight of people in Belgium and round the world has increased in the last two centuries,”

    The population of Belgium is irrelevant – BMI is based on an ideal figure, not an average. The height-variation between two populations being compared obviously makes some difference – but as far as I know, the kind of height increases we’ve seen in the last century (which is typically the kind of period being compared) are nothing like large enough to change the bigger picture.

  19. I’ve never seen a fat junkie – all the ones we have around here are rake thin and look about 30 years old than they really are. They also smoke like chimneys.

  20. Pingback: On BMI, Smoking and Physical Fitness – Malcolm Bracken

Leave a Reply

Your email address will not be published. Required fields are marked *