Fat?

Feb. 9th, 2005 08:56 am
dr_tectonic: (Default)
[personal profile] dr_tectonic
Ha! I knew it!

It is 9 am, and I have already been at work for an hour and a half, because today is the company health fair and I had a blood draw scheduled at 7:45. Oy vey. They had to stick me twice because I have uncooperative veins; now I know just to go straight for the expert phlebotomist.

Anyway, among the other health screenings, they have the nifty electrical inductance scales that calculate your bodyfat percentage. And now I finally have proof of something I have suspected all along: BMI (Body Mass Index) is a complete lie. (For me, anyway).

I weighed in at 232 pounds, fully dressed. I'm 5'10.5". Thus, my BMI is about 33 1/3. Even if we subtract off 7 pounds of clothes (my scale at home says I'm 225, naked) and round up to 71 inches tall, it's still 31.5. 30 is the cutoff for "obese". To have a BMI of 25, at the upper limit of the "normal weight" category, someone who's 5'10.5" is supposed to weigh between 175 and 180 pounds. (HA!)

BUT! The scale also measured my bodyfat at 27.7%. The attendant said she thought that was even probably several percent high (normal error is +/-3-5%) because I'd been fasting for 12 hours for the blood draw and I was likely a little dehydrated, so let's call it 25%. If I weigh 232 pounds and 25% of it is fat, that means that there are 58 pounds of fat in my body. If I lost ALL of them, I would weigh 174 pounds. So at 0% bodyfat, I would have a BMI of 25, just past the upper limit of "normal"! In other words, it's not physically possible for me to have a "normal" BMI -- not unless I lose a limb or something!

Still, I am kinda fat. How much should I really weigh? Well, if I dropped down to about 200 and the loss was all fat, I'd go from 58 pounds of bodyfat to 26 pounds of bodyfat. 26 out of 200 is 13% bodyfat, which is smack in the middle of the "healthy" range. 200. That seems like a totally reasonable number to me. So I think that will be my fitness goal. (Even dropping down to 215 would get me down below 20% bodyfat and into the "healthy" range.) Of course, the BMI for 5'11" and 200 pounds is about 28 -- in the middle of the "overweight" category.

If 14% bodyfat is the mean of the "healthy" range (for my age), and my weight at 14% bodyfat is 202 pounds, then the BMI for my "ideal" weight at 5'10.5" is 29.0. The mean BMI of the "normal range" is 21.75, which means that, assuming those mid-range numbers ought to line up, BMI is off by more than 7 points for my body frame. So really, my current BMI isn't 33 -- it's more like 26: a little bit overweight, not seriously obese.

I fucking KNEW it. Goddam population statistics.

Date: 2005-02-09 09:54 am (UTC)
From: [identity profile] ng-nighthawk.livejournal.com
How does one measure percentage bodyfat? I thought it took more than a scale. . .is it a special scale? I'd be interested in performing this experiment myself.

Date: 2005-02-09 11:11 am (UTC)
From: [identity profile] dr-tectonic.livejournal.com
Yeah, it's got electromagnetic inductance sensors. It measures the conductance of your body (fat conducts worse than muscle) and can estimate the % bodyfat from that.

My opinion on BMI

Date: 2005-02-09 09:54 am (UTC)
From: [identity profile] arcticturtle.livejournal.com
I fluctuate from 150-160 lb. At 160 and 5'8.5", my BMI comes in at 24.3, exactly 0.7 from being "overweight".

I'm the same weight I was in college. You remember me. If anybody tells me that I am on the very cusp of "overweight", I tell them they are out of their freakin' minds.

Re: My opinion on BMI

Date: 2005-02-09 10:00 am (UTC)
From: [identity profile] nehrlich.livejournal.com
I'm in the "overweight" range on BMI. And was more so when I was in much better shape at Stanford. BMI is really crude and doesn't take into account body type. For instance, muscle weighs more than fat, so when I was bulked out to 235 and had no neck from weightlifting three times a week at Stanford, I was borderline "obese" on BMI despite being in easily the best shape I've probably ever been in.

It's stupid. Trust your own judgment.

Re: My opinion on BMI

Date: 2005-02-09 11:21 am (UTC)
From: [identity profile] dr-tectonic.livejournal.com
Well, right, I'm just glad to finally have proof that it's stupid.

I'm not asking whether I'm fatter than I should be -- I know the answer to that question; it's "yes". The question that calculating BMI attempts to answer, though, is "how fat ought I be?" and it's that question that the BMI gives a totally bogus answer to. Now I have the bodyfat percentage answer, and it's actually useful information that has practical application to attempts to achieve greater physical fitness...

Re: My opinion on BMI

Date: 2005-02-09 12:39 pm (UTC)
From: [identity profile] arcticturtle.livejournal.com
doesn't take into account body type.
I can't imagine what body type it does take into account. Quinzee, maybe? It seems like a standard agreed upon by a board of fashion photographers for Vogue.

Date: 2005-02-09 10:24 am (UTC)
From: [identity profile] eto-theipi.livejournal.com
My bathroom scale has those inductance sensors, and I've noticed that my measured body fat varies several percent over the course of a day, without a similarly large change in weight. I'll also see measurements taken at the same time of day on subsequent days that suggest that I've gained or lost 3-5lbs of fat overnight, which seems implausible to me. Some of that is probably due to imprecision (the difference between 21.9% read as 21 and 23.1% read as 23), but it makes me wonder how good the numbers really are. But I still dutifully enter them in the spreadsheet and watch the trends.

The scales are cheap enough these days that you could easily get a larger set of readings to average among and, perhaps, get a more accurate picture.

BMI is nonsense, though. It presumes that everyone has the same frame and the same amount of muscle.

Date: 2005-02-09 11:12 am (UTC)
From: [identity profile] dr-tectonic.livejournal.com
Yeah, I gather that the error is all about how hydrated you are. They said plus or minus 3-5% on the little recording sheet.

Date: 2005-02-09 11:41 am (UTC)
From: [identity profile] eto-theipi.livejournal.com
+/- 3-5% of the total value? So if I'm 20% fat, it might read as high as 25 or as low as 15? Yeesh. That's a huge range.

Date: 2005-02-09 11:48 am (UTC)
From: [identity profile] dr-tectonic.livejournal.com
Yeah, that was the impression I got. Thus the need for your spreadsheet...

Date: 2005-02-09 10:45 am (UTC)
From: [identity profile] melted-snowball.livejournal.com
So I'll start out by saying I know absolutely nothing about this. (Though I've been annoyed about BMI numbers for a while, too.)

But...fat is stored in human cells, which have lots of water in them, even when they're fat cells.

So my vague understanding is that if you lost 58 pounds of fat, you'd probably lose a large amount of water by extension, and then be really, really thin.

In general, BMI numbers are weird, though. As I recall, it's basically weight divided by height squared. So if we assume [!] that people's mean "breadth" is linear in their height, then BMI is the mean weight of a horizontal cross-section of a person, times a constant of proportionality.

Once I started thinking about that, I basically stopped respecting the number, and went on with my life, worrying about eating good food, getting occasional exercise, and being able to fit into my pants. We don't own a scale.

Date: 2005-02-09 11:17 am (UTC)
From: [identity profile] dr-tectonic.livejournal.com
Well, I think the "58 pounds of fat" I calculated is probably more like "58 pounds of fatty tissue", not actual fat molecules. At least, that's the assumption I'm working on.

If I lost 58 pounds of them, that would be all of it, so I would indeed by very, very skinny. Ten of them, though, or 25 if I got super-athletic -- that's within the realm of reason.

Date: 2005-02-09 11:40 am (UTC)
From: [identity profile] melted-snowball.livejournal.com
Hmm. I think it's actually % fat, based on skimming this paper. (I'm assuming you've got decent e-journal subscriptions where you are...)

I read one paragraph toward the beginning as saying that they divide a person's mass into "fat mass" and "fat-free mass", and then estimate how much water is in the person, assume that their "fat-free mass" is 73% water, and do the obvious arithmetic.

But it's remarkable how hard it is to tease this down.

[And, the whole point of the paper is to indicate that all of these measures are statistically flawed. Though the entire paper is remarkably boring...]

Bleah. I'm bored just thinking about this.

Date: 2005-02-09 12:08 pm (UTC)
From: [identity profile] thedragonweaver.livejournal.com
Yah. BMI puts me at around 24, IIRC. You remember what I look like.

Of course, I'll TAKE that extra point or two in BMI because for the first time in my life, I've got *curves*!

(Whic, BTW, is my other reason for being annoyed at BMI. Same for men and women? Bull!)

Date: 2005-02-09 03:05 pm (UTC)
From: [identity profile] detailbear.livejournal.com
My numbers here.

As it was explained to me, the first calculation is to split into Total Body Fat and Fat-Free Mass.

Total Body Fat includes adipose tissue (the stuff you want to reduce) plus necessary fat around the organs and on the nerves. Olympic gymnasts get down to 5%. Below 3% results in body failure.

Fat Free Mass includes bones, muscles, cartilage, organs, skin and blood. Of that, you want to keep the water percentage of the blood to be above 60%. You're quite right about the lean calculation. If you got rid of all your fat, you'd be 175 lbs. Required fat to live is 9 lbs, for a total of 184. If you exercise to lose the weight, you'll gain muscle and the number would be higher.

And 2 trainers and my doctor have told me to ignore BMI. At 5'7", my target weight is 181 to 194 lbs. with under 20% body fat.

And the National Institute for Health lists restrictions on the accuracy of the BMI.