Body Fat Calculation

"GaryG" <sorrynoemail.RemoveThis@NOSPAMX.com> wrote in message
news:clhJd.4401$Xz4.1165@fe03.lga...
> "John" <y_tu_momma_tambien.RemoveThis@hotmail.com> wrote in message
> news:%4hJd.82546$w62.69661@bgtnsc05-news.ops.worldnet.att.net...
>> What is the proper procedure and calculation for determining body fat
>> percentages? I had mine checked the other day by a "fitness trainer" and
> it
>> came back lower than what I believe it should be.
>>
>
> There are a bunch of different methods to estimate body fat percentage
> (see
> below). They're all estimates, some more accurate than others. Popular
> methods include:
>
> "Dunk tank", the gold standard.
>
> DEXA - low-power X-rays, which can supposedly show where you body fat is
> distributes.
>
> Bio-impedance - used in body fat scales, and some health clubs (works by
> running a low-power electrical current through your body). Subject to
> variation depending on body hydration levels.
>
> Calipers - often used by trainers, and considered reasonably accurate, if
> the operator is trained and careful with the measurements.
>
> Body Circumference measurements - aka, the "Navy" method. According to
> their
> research, this is supposed to be about as accurate as any other method,
> and
> it's easy to perform. (FWIW, I've built this method into my WeightWare
> program, and you can find an online tool here:
> http://www.he.net/%7Ezone/prothd2.html ).

Wow. Your site tells me I'm 26%. All other methods I've done tell me I'm
about 15%.

Good ole Navy...

--
JayB
 >> Stay informed about: Body Fat Calculation 

"JayB" <jerryb RemoveThis @qwest.net> wrote in message
news:VfjJd.1804$a03.1604@news02.roc.ny...
>
> "GaryG" <sorrynoemail RemoveThis @NOSPAMX.com> wrote in message
> news:clhJd.4401$Xz4.1165@fe03.lga...
> > "John" <y_tu_momma_tambien RemoveThis @hotmail.com> wrote in message
> > news:%4hJd.82546$w62.69661@bgtnsc05-news.ops.worldnet.att.net...
> >> What is the proper procedure and calculation for determining body fat
> >> percentages? I had mine checked the other day by a "fitness trainer"
and
> > it
> >> came back lower than what I believe it should be.
> >>
> >
> > There are a bunch of different methods to estimate body fat percentage
> > (see
> > below). They're all estimates, some more accurate than others. Popular
> > methods include:
> >
> > "Dunk tank", the gold standard.
> >
> > DEXA - low-power X-rays, which can supposedly show where you body fat is
> > distributes.
> >
> > Bio-impedance - used in body fat scales, and some health clubs (works by
> > running a low-power electrical current through your body). Subject to
> > variation depending on body hydration levels.
> >
> > Calipers - often used by trainers, and considered reasonably accurate,
if
> > the operator is trained and careful with the measurements.
> >
> > Body Circumference measurements - aka, the "Navy" method. According to
> > their
> > research, this is supposed to be about as accurate as any other method,
> > and
> > it's easy to perform. (FWIW, I've built this method into my WeightWare
> > program, and you can find an online tool here:
> > http://www.he.net/%7Ezone/prothd2.html ).
>
> Wow. Your site tells me I'm 26%. All other methods I've done tell me I'm
> about 15%.
>
> Good ole Navy...

FWIW - that's not my site. The Navy method has quite a bit of research
behind it, and their formula is accurate enough that it's used across the
Dept. of Defense, as part of their personnel evaluation process.

As with most of the other methods, you do need to take care with the
measurements. If yours is that far off, I suspect you've screwed up the
abdominal measurement (or, you're really fatter than you think ). What
method gave you 15%?

GG

>
> --
> JayB
>
>
 >> Stay informed about: Body Fat Calculation 

GaryG <sorrynoemail.DeleteThis@NOSPAMX.com> wrote:
>FWIW - that's not my site. The Navy method has quite a bit of research
>behind it, and their formula is accurate enough that it's used across the
>Dept. of Defense, as part of their personnel evaluation process.

Any method that's built on the concept of taking a few
measurements and plugging them into a polynomial will
be exactly on the line of estimation represented by the
polynomial.

Which will be a zero-width path through a wide cloud of
actual observations.

None of them is "accurate" to better than tens of percent
(e.g., 3 points on a 15-percent bfp reading is 20%
accuracy), because people are that different.

BTW, the dunk tank may be considered a standard, but it
only estimates one data point, body density, and tries
to estimate the quantity of a single body component,
fat, that has a known density; it has no way of knowing
the ratios of all the other components like bone, brain,
lung, muscle, organ meat, blood, so it can't tell what
the density of the lean body mass is, so even knowing the
density of fat doesn't give it a precise ratio of fat mass
to total body mass.

The "accuracy" of calipers is only in terms of how well
they can predict a dunk-tank reading, and that's becaus
the caliper equation was formulated using comparisons
between samples on calipers and dunk-tank data.

The only truly accurate estimate of bodyfat would come
from a fine-grained assay using some sort of tomography
(CAT or MRI or old-school: freeze it and slice it).

And in the end, BFP isn't precisely related to appearance,
which is the only absolute metric that anyone really cares
about.

What the measurements are good for is tracking change
and defining goals. The actual numbers don't matter,
whether you can see them trending consistently does.

Since calipers are the easiest way to directly measure
a subset of fat (subcutaneous fat) without confusing
it with muscle (as the tape measure does), and they're
cheaper than a bucket of whey powder, I recommend them.
You don't even need to bother with the equation; just
track the total millimeters; the equation does nothing
but add them up and convert the total to a number that,
as I said above, means little.

--Blair
"Mine was 7.7 this morning, in
case I didn't convince you."
 >> Stay informed about: Body Fat Calculation 

"Blair P. Houghton" <b.DeleteThis@p.h> wrote in message
news:XWAJd.22005$f47.3639@news.easynews.com...
> GaryG <sorrynoemail.DeleteThis@NOSPAMX.com> wrote:
> >FWIW - that's not my site. The Navy method has quite a bit of research
> >behind it, and their formula is accurate enough that it's used across the
> >Dept. of Defense, as part of their personnel evaluation process.
>
> Any method that's built on the concept of taking a few
> measurements and plugging them into a polynomial will
> be exactly on the line of estimation represented by the
> polynomial.
>
> Which will be a zero-width path through a wide cloud of
> actual observations.

Well, I've reviewed the research the Dept. of Defense used when they
developed their formulas, and it seems pretty compelling. It's available
from here: http://www.mwr.navy.mil/mwrprgms/fitness/devbodycomp.pdf

They don't claim 100% accuracy, but they did compare their formulas against
other methods, and validated them across genders and racial groups. Their
test subjects were military aged, so perhaps their formula might not apply
to adolescents, or the elderly. But, in my experience it produces
believable results for most folks (especially when compared with
bio-impedance body fat scales).

As you say, all current methods are estimates. But, the Navy method is: a)
easy to use, b) cheap, c) repeatable, and d) directly "related to
appearance" (esp., abdominal circumference). Even if it's not 100%
accurate, its repeatability, and direct relationship to circumference
measurements, makes it a useful alternative IMO.

> None of them is "accurate" to better than tens of percent
> (e.g., 3 points on a 15-percent bfp reading is 20%
> accuracy), because people are that different.
>
> BTW, the dunk tank may be considered a standard, but it
> only estimates one data point, body density, and tries
> to estimate the quantity of a single body component,
> fat, that has a known density; it has no way of knowing
> the ratios of all the other components like bone, brain,
> lung, muscle, organ meat, blood, so it can't tell what
> the density of the lean body mass is, so even knowing the
> density of fat doesn't give it a precise ratio of fat mass
> to total body mass.
>
> The "accuracy" of calipers is only in terms of how well
> they can predict a dunk-tank reading, and that's becaus
> the caliper equation was formulated using comparisons
> between samples on calipers and dunk-tank data.
>
> The only truly accurate estimate of bodyfat would come
> from a fine-grained assay using some sort of tomography
> (CAT or MRI or old-school: freeze it and slice it).
>
> And in the end, BFP isn't precisely related to appearance,
> which is the only absolute metric that anyone really cares
> about.
>
> What the measurements are good for is tracking change
> and defining goals. The actual numbers don't matter,
> whether you can see them trending consistently does.

Agreed...which is my main complaint about the bio-impedance method. It's
subject to large variations, even with measurements taken a few hours apart,
due to the influence of hydration levels.

>
> Since calipers are the easiest way to directly measure
> a subset of fat (subcutaneous fat) without confusing
> it with muscle (as the tape measure does), and they're
> cheaper than a bucket of whey powder, I recommend them.
> You don't even need to bother with the equation; just
> track the total millimeters; the equation does nothing
> but add them up and convert the total to a number that,
> as I said above, means little.

Not sure I agree that calipers are easier to use than a tape measure. Some
of the caliper methods require taking measurements at sites that are not
easily self-measured, and for all caliper methods the recommendation is for
a "trained technician" to perform the measurements. Good quality calipers
are also quite a bit more expensive than a fabric tape measure.

What calipers do you use, and how much did they cost? Do you take your own
measurements, and how many sites do you test?

>
> --Blair
> "Mine was 7.7 this morning, in
> case I didn't convince you."
 >> Stay informed about: Body Fat Calculation 

> The Navy method has quite a bit of research behind it, and their
formula is accurate enough that it's used across the
> Dept. of Defense, as part of their personnel evaluation process
So are BMI charts which are woefully inadequate !!
 >> Stay informed about: Body Fat Calculation 

"gman99" <gman99.TakeThisOut@canada.com> wrote in message
news:1106743175.945982.31260@c13g2000cwb.googlegroups.com...
> > The Navy method has quite a bit of research behind it, and their
> formula is accurate enough that it's used across the
> > Dept. of Defense, as part of their personnel evaluation process
> So are BMI charts which are woefully inadequate !!
>

As part of their routine physical evaluations of personnel, the DoD
evaluation process uses a modified BMI standard and, when necessary, the
so-called Navy method for estimating body fat. Specifically, personnel must
fall below an age-adjusted upper limit for weight and height. If they are
too heavy for their age and height, they are then measured for percentage of
body fat (using the "Navy" body circumference method to calculate body fat
percentage). If their body fat is too high, they are considered out of
compliance with DoD standards - the impact of this includes
non-promotability, and assignment to a program to get them back into
compliance.

The DoD standards for weight (
http://www.armyg1.army.mil/hr/readiness/600-9_I01.pdf ) are age-adjusted,
and do not use the normal "Overweight" and "Obese" BMI cutoffs that we
normally hear about. For instance, the requirements for a 6' tall male are:

Age 17-20: less than 190 lbs (BMI=25.
21-27: less than 195 lbs (BMI=26.4)
28-39: less than 200 lbs (BMI=27.1)
40+: less than 203 lbs (BMI=27.5)

Assuming someone is over these standards, their body fat would be measured.
The upper limits for body fat are:

Age 17-21: 20%
Age 21-27: 22%
Age 28-39: 24%
Age 40+: 26%

IMO, these standards seem reasonable (generous, even). It would be nice if
insurance companies would adopt some sort of similar procedures, so that
healthy folks with high BMI but low BF are not penalized for their muscle
mass.

--
GG
http://www.WeightWare.com
Your Weight and Health Diary
 >> Stay informed about: Body Fat Calculation 

GaryG wrote:
> "gman99" <gman99.DeleteThis@canada.com> wrote in message
> news:1106743175.945982.31260@c13g2000cwb.googlegroups.com...
>>> The Navy method has quite a bit of research behind it, and their
>> formula is accurate enough that it's used across the
>>> Dept. of Defense, as part of their personnel evaluation process
>> So are BMI charts which are woefully inadequate !!
>>
>
> As part of their routine physical evaluations of personnel, the DoD
> evaluation process uses a modified BMI standard and, when necessary,
> the so-called Navy method for estimating body fat. Specifically,
> personnel must fall below an age-adjusted upper limit for weight and
> height. If they are too heavy for their age and height, they are
> then measured for percentage of body fat (using the "Navy" body
> circumference method to calculate body fat percentage). If their body
> fat is too high, they are considered out of compliance with DoD
> standards - the impact of this includes non-promotability, and
> assignment to a program to get them back into compliance.
>
> The DoD standards for weight (
> http://www.armyg1.army.mil/hr/readiness/600-9_I01.pdf ) are
> age-adjusted, and do not use the normal "Overweight" and "Obese" BMI
> cutoffs that we normally hear about. For instance, the requirements
> for a 6' tall male are:
>
> Age 17-20: less than 190 lbs (BMI=25.
> 21-27: less than 195 lbs (BMI=26.4)
> 28-39: less than 200 lbs (BMI=27.1)
> 40+: less than 203 lbs (BMI=27.5)
>
> Assuming someone is over these standards, their body fat would be
> measured. The upper limits for body fat are:
>
> Age 17-21: 20%
> Age 21-27: 22%
> Age 28-39: 24%
> Age 40+: 26%
>
> IMO, these standards seem reasonable (generous, even). It would be
> nice if insurance companies would adopt some sort of similar
> procedures, so that healthy folks with high BMI but low BF are not
> penalized for their muscle mass.

Good point. I'm 47, 5' 10" and 195 lbs. However, BF is 13%, resting HR is
56 and I have a 33" waist. But when I went for life insurance awhile back,
I got dinged because of my weight. My BMI is 28.
--
-Larry
 >> Stay informed about: Body Fat Calculation 

>
> The only truly accurate estimate of bodyfat would come
> from a fine-grained assay using some sort of tomography
> (CAT or MRI or old-school: freeze it and slice it).
>
The only true method is autopsy. That the procedure requires you to be dead
is an unfortunate prerequisite. By which time you may have lost any
interest in your Body Fat%
 >> Stay informed about: Body Fat Calculation 

There has to be a way to distinguish yourself from a fatass.
Did you need a pre-insurance phyiscal for the life insurance?
I had to get one to increase my company provided insurance.
My doctor wrote "muscular build x2" in his notes. If that means
anything
and tranfers to the insurance forms I don't know (I 'failed' & got
denied for other reasons)
but I'd think there's something to identify you as
a healthy, regular execiser.
 >> Stay informed about: Body Fat Calculation 

jcase DeleteThis @attbi.com wrote:
> There has to be a way to distinguish yourself from a fatass.
> Did you need a pre-insurance phyiscal for the life insurance?
> I had to get one to increase my company provided insurance.
> My doctor wrote "muscular build x2" in his notes. If that means
> anything
> and tranfers to the insurance forms I don't know (I 'failed' & got
> denied for other reasons)
> but I'd think there's something to identify you as
> a healthy, regular execiser.

Since it was for 500k, there was a physical and access to medical records.
The physical was simply bloodwork, HR, BP and weight. If I had gone for
less coverage, there wouldn't have been that requirement. The other thing
that dinged me was that in my medical records, I had been diagnosed with ADD
about six years ago. Even though I don't take meds for it (other than
coffee ), that hit me hard too.

As I think about it, this was May of '03 and I was fatter then. Maybe
around 210? Don't remember exactly. So maybe I was a fatass after all!
--
-Larry
 >> Stay informed about: Body Fat Calculation 

GaryG <sorrynoemail RemoveThis @NOSPAMX.com> wrote:
>"Blair P. Houghton" <b RemoveThis @p.h> wrote in message
>news:XWAJd.22005$f47.3639@news.easynews.com...
>> GaryG <sorrynoemail RemoveThis @NOSPAMX.com> wrote:
>> >FWIW - that's not my site. The Navy method has quite a bit of research
>> >behind it, and their formula is accurate enough that it's used across the
>> >Dept. of Defense, as part of their personnel evaluation process.
>>
>> Any method that's built on the concept of taking a few
>> measurements and plugging them into a polynomial will
>> be exactly on the line of estimation represented by the
>> polynomial.
>>
>> Which will be a zero-width path through a wide cloud of
>> actual observations.
>
>Well, I've reviewed the research the Dept. of Defense used when they
>developed their formulas, and it seems pretty compelling. It's available
>from here: http://www.mwr.navy.mil/mwrprgms/fitness/devbodycomp.pdf

The Navy seems less compelled:

http://www.nhrc.navy.mil/programs/BodyFat/conclusions.html

Circumference measurement is a little closer, at 3-4%
standard error, but that still means that 3 sigma is 9-12%
and 6 sigma is 18-24%. And that's in clinical conditions.

There are going to be hordes of people who don't fit the
narrow line through the big cloud at those error levels,
and introduce measurement errors just by not being the
people who did the study.

When someone comes up with a fat estimation that brings that
6-sigma range down to 5% or so, we'll have something.

I mean, look at what they're measuring: circumference of four body
parts, and height.

When a guy like Bertil Fox (5'11, 230 lbs shredded) in
competiton shape would be considered by this method to
have a bodyfat percentage in the 30's or higher, you know
something illogical has happened.

I think they might get somewhere if they measured both the
skinfold on those bodyparts and the circumference; then
they'd have an actual measurement of the cross-sectional
area of the subcutaneous fat layer. Either method alone
misses the data the other method measures, orthogonally.

--Blair
"It's quite funny, from a mathematical
standpoint."


P.S. a = annular area estimate = pi*(r1^2 - r2^2) = pi*(r1-r2)(r1+r2)
s = skinfold measurement
c = circumference measurement
r1-r2 = s/2
r1+r2 = c/2pi + (c/2pi - s/2) = c/pi - s/2
therefore:
a = pi * s/2 * (c/pi - s/2) = s/2 * (c - pi*s/2)

in case anyone in the Navy is watching.
 >> Stay informed about: Body Fat Calculation 

"Blair P. Houghton" <b.DeleteThis@p.h> wrote in message
news:Bn%Jd.148752$f47.27060@news.easynews.com...
> GaryG <sorrynoemail.DeleteThis@NOSPAMX.com> wrote:
> >"Blair P. Houghton" <b.DeleteThis@p.h> wrote in message
> >news:XWAJd.22005$f47.3639@news.easynews.com...
> >> GaryG <sorrynoemail.DeleteThis@NOSPAMX.com> wrote:
> >> >FWIW - that's not my site. The Navy method has quite a bit of
research
> >> >behind it, and their formula is accurate enough that it's used across
the
> >> >Dept. of Defense, as part of their personnel evaluation process.
> >>
> >> Any method that's built on the concept of taking a few
> >> measurements and plugging them into a polynomial will
> >> be exactly on the line of estimation represented by the
> >> polynomial.
> >>
> >> Which will be a zero-width path through a wide cloud of
> >> actual observations.
> >
> >Well, I've reviewed the research the Dept. of Defense used when they
> >developed their formulas, and it seems pretty compelling. It's available
> >from here: http://www.mwr.navy.mil/mwrprgms/fitness/devbodycomp.pdf
>
> The Navy seems less compelled:
>
> http://www.nhrc.navy.mil/programs/BodyFat/conclusions.html

Not sure I understand...that web page concludes that: "Given that the
standard error of estimate for the Navy equation is lower than that for any
of the skinfold equations tested, there is a greater likelihood that the
Navy equation will provide the better body fat estimate."

>
> Circumference measurement is a little closer, at 3-4%
> standard error, but that still means that 3 sigma is 9-12%
> and 6 sigma is 18-24%. And that's in clinical conditions.
>
> There are going to be hordes of people who don't fit the
> narrow line through the big cloud at those error levels,
> and introduce measurement errors just by not being the
> people who did the study.
>
> When someone comes up with a fat estimation that brings that
> 6-sigma range down to 5% or so, we'll have something.

Do the skinfold estimates come this close? Based on the Navy research, I
under the impression that they don't.

>
> I mean, look at what they're measuring: circumference of four body
> parts, and height.
>
> When a guy like Bertil Fox (5'11, 230 lbs shredded) in
> competiton shape would be considered by this method to
> have a bodyfat percentage in the 30's or higher, you know
> something illogical has happened.

I find that hard to believe. Do you have his measurements?

Regardless, that's someone who is clearly a statistical "outlier" (in this
case, a far, far outlier). For the majority of folks, the body
circumference measurements are clearly "close enough for government work",
as well as easy to employ.

>
> I think they might get somewhere if they measured both the
> skinfold on those bodyparts and the circumference; then
> they'd have an actual measurement of the cross-sectional
> area of the subcutaneous fat layer. Either method alone
> misses the data the other method measures, orthogonally.
>
> --Blair
> "It's quite funny, from a mathematical
> standpoint."
>
>
> P.S. a = annular area estimate = pi*(r1^2 - r2^2) = pi*(r1-r2)(r1+r2)
> s = skinfold measurement
> c = circumference measurement
> r1-r2 = s/2
> r1+r2 = c/2pi + (c/2pi - s/2) = c/pi - s/2
> therefore:
> a = pi * s/2 * (c/pi - s/2) = s/2 * (c - pi*s/2)
>
> in case anyone in the Navy is watching.
 >> Stay informed about: Body Fat Calculation 

npne <npne DeleteThis @nowhere.net.cz> wrote:
>
>>
>> The only truly accurate estimate of bodyfat would come
>> from a fine-grained assay using some sort of tomography
>> (CAT or MRI or old-school: freeze it and slice it).
>>
>The only true method is autopsy. That the procedure requires you to be dead
>is an unfortunate prerequisite. By which time you may have lost any
>interest in your Body Fat%

I'm reducing mine to zero. By definition the percentage
error will be zero then as well.

--Blair
"Doesn't help my old, fat self, but then
he's dead to me, so maybe you're right."
 >> Stay informed about: Body Fat Calculation