I've spent some time at the University of Western Australia studying for my Honours degree, and during that time my collegues and I had to measure and record the physical traits of many human subjects, as was the nature of the thesis.

It's pretty much accepted amongst Sports Science professors at universities and research centres across the world that measuring bodyfat as a percentage is not an accurate measurement. But thousands of 'experts' still use the % in their literature and seminars.

Generally, to attain a percentage, you need to take a collection of data from the subject and put that information into a formula, which will produce a percentage. There are literally HUNDREDS of different mathematical formulae used, and the results are so varied and inaccurate that you can pretty much choose a formula to give you the exact bodyfat % you wish to find. If I want to be only 3% fat, I'm sure I could find a formula which would turn my data into such.

The reason they aren't accurate, is because there is much more to a subject's level of bodyfat than skinfold measurements. You need to take into account: whether the skinfold has constant compressibility, whether that skin thickness is a constant proportion of the fold, whether or not fat constitutes a constant proportion of the entrapped adipose tissue, whether the ratio of subcetaneous to internal fat deposits remains fixed, whether inter-individual fat patterning is consistent.... in other words, things like height, weight, bone size, shape and density, water retention, skin compressibility, adipose tissue compressibility.... there are many factors which affect the measurement that comes up on the calipers, and many formulae don't take all of these factors into account. Also, we errors due to poor site locatoin, averaging 2mm for unexperienced individuals when the skinfold sites were not landmarked, and thus observed differences in bodyfat levels up to 11% on the same subject, using the same system and formula. There is so much room for human error that it renders the result completely unreliable. Even your choice of calipers can make a difference of up to 5%.

In one study, we used a system currently used by the main sporting research institution in Ukraine to determine the bodyfat of one of my collegues. Now this guy is pretty lean, he's a national-level 1500m runner.... but his bodyfat came out at
-3.3%.... go figure. Negative bodyfat?? We compared many different systems and formulae, and the percentages ranged from 27.5% to -3.3% bodyfat for the same subject. Doesn't sound too reliable.

That's why it frustrates me when I see people posting about their bodyfat percentage. The fact is, if you wanted to lose 10% bodyfat from your current level, you wouldn't need to diet or train or do any cardio. You could sit at home, eat ice-cream and watch TV and I could find a system and formula which would PROVE that your bodyfat is your ideal level.

The only accurate way to establish a subject's bodyfat levels and present them as a percentage, would be to put the subject in a blender, set them on fire and measure how much energy is given off. Not exactly practical, though.

It's widely accepted now that the skinfold measurements should be used directly for comparison without modification. The individual skinfold scores should be summed and the results can then be used for comparative purposes over time, or between individuals. Skinfold measurements should be taken from at least six sites*, although eight is better, (this is known as "S6SF" [or Sum of Six Skinfolds]) and the sum of these, should be used as a raw comparative figure, and not be tranformed or modified.... as this just introduces the potential for further error based on erroneous assumptions.

This system is known as "Densitometry".

I'm sure with me saying all this, most of you will continue to measure your bodyfat as a percentage, probably by standing on the bodyfat bathroom scales which supposedly determine your bodily composition through the soles of your feet.... it's just too convenient, and human beings like nice, round numbers. I just wanted to present some of my own findings on the issue, and explain why the bodyfat percentage is slowly moving aside and making way for S8SF and the like in the more credible athletic and health institues around the world.

*For those of you who are interested, the skinfold sites should be triceps, subscapular, supraspinale, abdominal, front thigh and medial calf. You can probably find more information, or how to find these exact sites in books in your library. Or feel free to e-mail me and I'll explain it further.

Regular scales are fine, but they can't differentiate between muscle loss/gain, fat loss/gain or water loss/gain. There's really no real method of measuring bodyfat. There's just too many variables.

It is amusing to see how some people do it, though. Some places even like the military still use the BMI (height:weight ratio). Of course, Dorian Yates weighed 277lbs at the 1996 Mr Olympia, and he was completely ripped. But for his height, he'd be classified as OBESE in the army. I had a guy come to me because he wanted to join the Australian Navy but he was too heavy. Their method was neck girth:weight girth ratio, and they sent him away until he lost some weight. He'd been doing endless sit-ups and crunches, but when he came to me I just told him to forget all that, and gave him a list of neck exercises. It's all completely absurd.

Yep, I believe the only real way is to monitor yourself in the mirror and let your eyes tell you the truth. Bodybuilding after all, is about what you look like.... so if you can create an illusion, you're pretty much set. Of course, this too can be a problem also, when your eyes tell the truth.... but your brain distorts the image. Mostly common in girls aged 14-19.

But that's a different matter altogether.

Forget the bodyfat scales. They're a waste of time and money. If you're not competent enough to take accurate S6SF (which not many of us are, because it's just too much trouble) then forget about recording your boyfat as a number. Tell it how it is.... or better still, let people judge for themselves and show a picture.

quote:

---

Originally posted by -133-:
I've had this dicussion at bodybuilder.org

I've spent some time at the University of Western Australia studying for my Honours degree, and during that time my collegues and I had to measure and record the physical traits of many human subjects, as was the nature of the thesis.

It's pretty much accepted amongst Sports Science professors at universities and research centres across the world that measuring bodyfat as a percentage is not an accurate measurement. But thousands of 'experts' still use the % in their literature and seminars.

Generally, to attain a percentage, you need to take a collection of data from the subject and put that information into a formula, which will produce a percentage. There are literally HUNDREDS of different mathematical formulae used, and the results are so varied and inaccurate that you can pretty much choose a formula to give you the exact bodyfat % you wish to find. If I want to be only 3% fat, I'm sure I could find a formula which would turn my data into such.

The reason they aren't accurate, is because there is much more to a subject's level of bodyfat than skinfold measurements. You need to take into account: whether the skinfold has constant compressibility, whether that skin thickness is a constant proportion of the fold, whether or not fat constitutes a constant proportion of the entrapped adipose tissue, whether the ratio of subcetaneous to internal fat deposits remains fixed, whether inter-individual fat patterning is consistent.... in other words, things like height, weight, bone size, shape and density, water retention, skin compressibility, adipose tissue compressibility.... there are many factors which affect the measurement that comes up on the calipers, and many formulae don't take all of these factors into account. Also, we errors due to poor site locatoin, averaging 2mm for unexperienced individuals when the skinfold sites were not landmarked, and thus observed differences in bodyfat levels up to 11% on the same subject, using the same system and formula. There is so much room for human error that it renders the result completely unreliable. Even your choice of calipers can make a difference of up to 5%.

In one study, we used a system currently used by the main sporting research institution in Ukraine to determine the bodyfat of one of my collegues. Now this guy is pretty lean, he's a national-level 1500m runner.... but his bodyfat came out at
-3.3%.... go figure. Negative bodyfat?? We compared many different systems and formulae, and the percentages ranged from 27.5% to -3.3% bodyfat for the same subject. Doesn't sound too reliable.

That's why it frustrates me when I see people posting about their bodyfat percentage. The fact is, if you wanted to lose 10% bodyfat from your current level, you wouldn't need to diet or train or do any cardio. You could sit at home, eat ice-cream and watch TV and I could find a system and formula which would PROVE that your bodyfat is your ideal level.

The only accurate way to establish a subject's bodyfat levels and present them as a percentage, would be to put the subject in a blender, set them on fire and measure how much energy is given off. Not exactly practical, though.

It's widely accepted now that the skinfold measurements should be used directly for comparison without modification. The individual skinfold scores should be summed and the results can then be used for comparative purposes over time, or between individuals. Skinfold measurements should be taken from at least six sites*, although eight is better, (this is known as "S6SF" [or Sum of Six Skinfolds]) and the sum of these, should be used as a raw comparative figure, and not be tranformed or modified.... as this just introduces the potential for further error based on erroneous assumptions.

This system is known as "Densitometry".

I'm sure with me saying all this, most of you will continue to measure your bodyfat as a percentage, probably by standing on the bodyfat bathroom scales which supposedly determine your bodily composition through the soles of your feet.... it's just too convenient, and human beings like nice, round numbers. I just wanted to present some of my own findings on the issue, and explain why the bodyfat percentage is slowly moving aside and making way for S8SF and the like in the more credible athletic and health institues around the world.

*For those of you who are interested, the skinfold sites should be triceps, subscapular, supraspinale, abdominal, front thigh and medial calf. You can probably find more information, or how to find these exact sites in books in your library. Or feel free to e-mail me and I'll explain it further.

---