h2g2 The Hitchhiker's Guide to the Galaxy: Earth Edition

I *know* I'm unfit and unhealthy, so I'm not going to argue with a bullsh!t BMI.

I've got an actual leaving date of next Friday (which gives me two weeks' gardening leave). We've brought it forward because of my manager's holidays and because the rest of the team are moving offices as of Monday-week, and it seems simplest to leave first.

Five more days, probably only two or three that I actually go in for.

Very odd... Not sure how I feel about it..

Ben

Ben. How long had you been in the job?

I've just had a very odd online exchange with someone who was trying to sell me tv subscription stuff. Every time I asked a question, her response contained even more jargon I didn't understand. When I asked her to explain the jargon, I just got more incomprehensible nonsense so I gave up.

Don't suppose anyone here knows why my phone line should be involved if I don't want the phone or broadband services?

Just over 5 years permanent, 5 years 6 months all together.

Most of them suprisingly sh*te from a professional point of view.

BMI is just an expression of weight verus height. That's all. It's useful as part of clinical assessment, but it isn't clinical assessment. It's designed for normal adults, not trained bodybuilders.

It's a useful quick way of working out whether the adult in front of you has a normal weight for their height. That's all. If they have an abnormal weight for their height you then need to work out if the abnormality is fat, muscle, fluid, pregnancy, or something else.

I don't find it very difficult to tell muscle from fat when looking at someone! If I had any real doubts I could do more detailed tests.

I very rarely weight myself, but having done so this week, I found that my BMI currently fits within the 'normal' range (just), but I am well aware that there is fat that could very helpfully either not be there or be muscle instead. I'm getting there. Slowly. I dropped a belt notch between January and Easter. I know I'm not obese, but I'm not as fit or healthy as I could be, either.

I think the location of body fat is just as important as BMI. T's example of her rugby-playing daughter is a great example of being lean and still having a too high BMI. And Kelli is fit and gorgeous and certainly isn't at an unhealthy wieght. But there are also folks like me -- I'm not saying I shouldn't lose weight here -- who carry a very large percentage of thier excess weight in their hips and thighs as opposed to around their waist. So, my fat isn't as dangerous as it would be if it was located around my middle. And a standard BMI reading on me is going to say I'm uhealthier than I really am. I mean, in spite of my weight, my real age test turned out s being 6 years younger than my actual age.

http://www.cnn.com/2009/HEALTH/03/18/healthmag.obesity.lifespan/index.html

Toward the end of this article is information that being in the overweight range doesn't lead to early death at all.

Last time I did a BMI calc it suggested I was clinically obese. I am not. I've got a bit of a beer gut that won't shift, but I'm pretty sure my fitness is above average. I'm not particularly healthy, mind you. My diet over the past 24 hours has consisted of: 4 packets of crisps, 2 granola bars, a tray of chips, a sausage roll, 2 carrots, a slice of toast and err, mumblemumble pints of beer (I actually can't tell you how many, went to the beer festival last night).

That can't be good. On the other hand I've also walked 6 miles and jogged 5.

Oh, and a large quantity of

I came out as 8.1 years younger than I actually am : I would be doing my GCSEs by their calculation. Got nagged about my diet. I know their BMI calculator is faulty as they really got my height wrong, but impressively they think my BMI is only 2.3 under what the doctor's computer thinks it is-I had a routine check on Monday.

I'm busy at work: covering for people who are at the new place in Ashby, and going to be there all week next week.

KA, there is nothing wrong with .

Having see the pictures of Kelli, you should be proud of your size.

Photies of cooked and pickled striped beet, Hypatia:

http://picasaweb.google.com/lh/photo/ee9HRBfRVBi_pH4qNZHnJg?feat=directlink

http://picasaweb.google.com/lh/photo/GVxn7MepJX3xRGRpiNfYnw?feat=directlink

http://picasaweb.google.com/lh/photo/OPfkZkDvoKHYvXtB2ZQ-Ng?feat=directlink

It's a deep yellow and deep coral colour. Looks fab. Absolutely worth pickling separately from the red beetroot.

I'm off to see local heroes Ash tonight - it'll be the 4th time I've seen them live! First time was in Luxembourg, where I used my clout as the editor of an ex-pats journal to get a backstage interview. The at a small gig in Queens Uni with my son, then as support to Snow Patrol at Ward Park 2008.

> I just try to eat right and make sure I at least get a walk every day.
Sounds like a very healthy approach - one that we follow around here, too. My problem would really be to cling to the bottom limit of the BMI span - thin as a stick is the usual descriptio. Oh and I am in shape - so long as long and thin is a shape

Very pretty pictures Ben! I still don't like beetroot though

>>> Interesting article you linked to there HN. Still it doesn't explain why if a lot of people think that the BMI is crude mathematical junk we carry on using it. With modern body analysers you can get a much better hold on things such as %fat etc.

Thanks for reading the article, Phil. You raise a good question. At the risk of annoying both the geeky with an over-simplified explanation and the math-phobic with any mention of math, I'm going to try to answer it.

Suppose I wanted to summarize a whole lot of Height and Weight data. I could build a statistical model. Suppose you have a lot of body fat vs body volume data, and you build another statistical model. Suppose someone else build another model based on circumference measurements of arms, legs, waists, and chests.

Now that we have more than one model, the question becomes, which model shall we use? This is a little like asking, which is the better tool, a hammer or a screwdriver? It depends on the task at hand. Statistics can be descriptive (how well does this model fit these data?) or predictive (given these data, how well does the model predict an outcome)? And within these general categories, models can be better or worse at describing or predicting.

The BMI calculation reduces two measurements (height and weight) to a single number. With it, someone can describe the height and weight of one large population and compare it with another. One can compare with a single number the Masai with the Inuit, or the Irish with the German. Now, you might be asking, why would anyone want to do that? I'm not sure, but in the 1800s, the early days of Anthropology, when Belgian mathematician Quetelet proposed the BMI calculation, this sort of thing went on all the time.

To use a descriptive calculation like BMI to predict anything is to use the wrong tool. It's worse than using a rifle when a cannon is indicated. It's like using a rifle when a sewing machine is indicated.

Since we are actually trying to predict something really important, such as how long and how well an individual can be expected to live, it is sensible to use the best tool we have at hand. We want a predictive model which is relatively accurate (or it's pointless) and relatively cheap and easy to use (or it just won't happen).

To predict an outcome, use a predictive model, not a descriptive model. There are several predictive models from which to choose, such as a %body fat calculation, as Kelli suggested, or various circumference measurements, as another researcher alluded to. We can even pretend that BMI is predictive and throw that in the mix.

Now we can figure out which of these models is the best at predicting what we are really interested in - individual health. Am I healthy? How long am I likely to live? Will I be a robust or sickly ?

To predict this sort of answer, we need a lot of data over a long period of time, and we ought to perform some extremely tedious calculations that tell us how well each model fits data sets and predicts outcomes. Until computers were widely available, these calculations were rarely done, and a lot of bad habits regarding model selection became ingrained.

As it turns out, BMI performs poorly as a predictive model, which is unsurprising, since it wasn't meant to be used in this fashion. With no disrespect meant to our medical Salonistas, I suspect BMI continues to be used because it was the first model, and its use is habitual, quick, and easy. However the model is known to have a rather high degree of statistical bias, that is, it does not fit the data well. As a result people with low muscle mass, who really could use more exercise have falsely reassuring BMIs. Those who get regular exercise have depressingly high BMIs.

Sure, BMI may be only one tool in the toolbox of a trained clinician, but why use it at all when there are better tools? The best models fit the data well and predict outcomes well.

The best models here measure or estimate both %body fat and %muscle tissue. The most accurate model uses a dunk tank to determine %body fat and %muscle tissue. The drawback is the tank, an expensive piece of equipment that takes up valuable room.

For accuracy, a close second is the calipers test, which works by measuring a skin fold to estimate the %body fat. The %muscle tissue is also estimated. This test could quickly and easily be performed at the doctor's office (or even the gym). Advantages here include much lower equipment cost and size.

Armed with accurate data, patients could take appropriate steps to increase muscle mass and/or reduce body fat and thus improve their health. Armed with misleading information, they could get all despondent about their year-long exercise program only increasing their BMI and drown their sorrows in a cheesecake binge.


HN

Did you like Snow Patrol too?

*Wonders whether Ash followed by Snow Patrol was part of a 'white flakey stuff which sometimes falls out of the sky' theme night*

<BB<

I find myself wondering to what extent BMI is actually used in medical practice, rather than being used by the insurance companies as a way to avoid paying out on insurance claims; something which affects USians far more than us poor benighted denizens of countries who suffer under the oppression of socialist models of healthcare.

Or to flip it around, I've never experienced it as being used as anything more than part of the puzzle at best, or a piece of appropriate and innocuous short-hand at worst.

Thanks for that, HN. We need to start a Ditch the BMI movement. HN can be in charge of writing the campaign literature.

Ben, those jars of pickled beets are beautiful. I love oven roasted beets and pickled beets. For my fall garden I've decided to grow some of the chioggia and the golden globe. I've only ever grown the cylindrical ones that are so good for slicing.

[GDZ]