BMR Equation References?

EvgeniZyntx

Does anyone have a scientific journal reference to the Katch McArdle equation -- all I have is their book(s)?

I've seen in a couple of place that this equation - based on LBM alone is "better" than other equations. I'd like to see the actual variance data or an estimator evaluation that actually has their equation and compares it to others.

Thanks!

(and no, wikipedia is not a valid reference)

ElliInJapan

Is this it?

http://ajcn.nutrition.org/content/28/2/105.short

(Sorry, I just skimmed through)

myofibril

I'd like to know this as well.

ETA: thanks to the above for the link...

myofibril

Is this it?

http://ajcn.nutrition.org/content/28/2/105.short

(Sorry, I just skimmed through)

Unfortunately that's not it as far as I can see.

The search goes on because other than their 1996 book additional scientific journal entries seem thin on the ground.

myofibril

Oh, this looks interesting....

http://www.abdn.ac.uk/ibes/speakman/pdf_docs/229.pdf

It seems that for relatively athletic individuals using a calculation based on FFM may prove to be more accurate

Given the majority of the population is not really athletic though....

(NB: I should add this is not what that study specifically alludes to but my own interpretation given the difficulties of obtaining an accurate measure of BF% in a general setting)

EvgeniZyntx

Is this it?

http://ajcn.nutrition.org/content/28/2/105.short

(Sorry, I just skimmed through)

Thanks, but no - that is an evaluation of body comp methods like skin caliper vs underwater weight - not related to BMR in that paper.

EvgeniZyntx

Oh, this looks interesting....

http://www.abdn.ac.uk/ibes/speakman/pdf_docs/229.pdf

It seems that for relatively athletic individuals using a calculation based on FFM may prove to be more accurate

Given the majority of the population is not really athletic though....

(NB: I should add this is not what that study specifically alludes to but my own interpretation given the difficulties of obtaining an accurate measure of BF% in a general setting)

Yes, that is one of the articles that got me thinking - it specifically indicates that body fat (FM) is a factor in BMR which goes against the Katch-McArdle equation whcih has no FM variable.

The K.A. equation is only FFM based.

So while it is interesting it does not support the K.A. equation. In fact, it challenges it.

myofibril

Yes, that is one of the articles that got me thinking - it specifically indicates that body fat (FM) is a factor in BMR which goes against the Katch-McArdle equation whcih has no FM variable.

The K.A. equation is only FFM based.

So while it is interesting it does not support the K.A. equation. In fact, it challenges it.

Perhaps it is because the interpretation is that FFM is simply all their is left over after you have calculated FM (obviously) so it would amount to the same thing? I'm not sure to be honest.

I'd be interested to know if you find anything further.

ETA: Actually having thought on it further that does make sense. What is BMR supposed to constitute? The amount of calories needed to support the bodies main organs and systems. As such it should be based on FFM as opposed to FM...

EvgeniZyntx

Yes, that is one of the articles that got me thinking - it specifically indicates that body fat (FM) is a factor in BMR which goes against the Katch-McArdle equation whcih has no FM variable.

The K.A. equation is only FFM based.

So while it is interesting it does not support the K.A. equation. In fact, it challenges it.

Perhaps it is because the interpretation is that FFM is simply all their is left over after you have calculated FM (obviously) so it would amount to the same thing? I'm not sure to be honest.

I'd be interested to know if you find anything further.

Nope, you have the possible influence of T3, T4 and other hormones.
Let's see what others come up with.

ElliInJapan

Is this it?

http://ajcn.nutrition.org/content/28/2/105.short

(Sorry, I just skimmed through)

Thanks, but no - that is an evaluation of body comp methods like skin caliper vs underwater weight - not related to BMR in that paper.

Ahh, sorry, I didn't have time to look carefully but thought to post just in case.

heybales

But doesn't the original study use the BMR values obtained along with the FFM that was measured, which means the minor but still metabolically active FM was actually included in the formula by the mere fact it contributed to the BMR values obtained?

So while you are using a value for FFM only in the calc, an assumed ratio is there for FM contributing to that BMR.

So really, as FFM goes up, an overweight with fat person is outside that ratio, and would be underestimated because of carrying more fat than expected. An overweight (by tables I mean) with expected ratio would be right on, a bodybuilder with very little fat would be overestimated because of having less fat than expected.

I recall they used healthy weight individuals from a wide range of ages.
But details beyond that I don't recall ever seeing either. I always wondered what the average ratio was in the study participants.

No different than Harris and Mifflin, they have an assumed ratio FFM/FM too, and overweight with fat gets over-estimated BMR because they have less than expected FFM.

EvgeniZyntx

But doesn't the original study use the BMR values obtained along with the FFM that was measured, which means the minor but still metabolically active FM was actually included in the formula by the mere fact it contributed to the BMR values obtained?

Which original study? I'm looking for a Katch McArdle study (other than their book) that gives their methodology.

Sure the FM factor is probably included in that type of assumption or the range of FM was to tight that they didn't measure a variance with any confidence.

So while you are using a value for FFM only in the calc, an assumed ratio is there for FM contributing to that BMR.

So really, as FFM goes up, an overweight with fat person is outside that ratio, and would be underestimated because of carrying more fat than expected. An overweight (by tables I mean) with expected ratio would be right on, a bodybuilder with very little fat would be overestimated because of having less fat than expected.

I recall they used healthy weight individuals from a wide range of ages.
But details beyond that I don't recall ever seeing either. I always wondered what the average ratio was in the study participants.

No different than Harris and Mifflin, they have an assumed ratio FFM/FM too, and overweight with fat gets over-estimated BMR because they have less than expected FFM.

Except that a variety of people state that the KM equation is "better", what you are suggesting is that it might not be. I'd like to see the paper and take a look at this. Perhaps this is unpublished, and if it is - why are we basing so much trust on KM and not on Cunningham or the later papers?

heybales

Except that a variety of people state that the KM equation is "better", what you are suggesting is that it might not be. I'd like to see the paper and take a look at this. Perhaps this is unpublished, and if it is - why are we basing so much trust on KM and not on Cunningham or the later papers?

Now Cunningham has more or better available reading on it?

Because same basis for the formula and testing.

I like the part did they have a wide range of BF% besides age? If it was, was it significant, or disregarded because of desire to just use LBM for a specific purpose.

Kind of like some of the VO2max formulas using BMI show that other factors can effect the results, but they state they want a formula with just certain variables that would be readily available to a Dr or whoever, so they are never separated from the formula, just included. They figure the variance is good enough for the intended purpose, and I'd agree it is.

I saw the Katch study referenced briefly in an online book, might have even been one of those Google reprint type books, or even on Amazon, I know I've found things that way too. Get enough pages to view for the info desired.

I'll have to look at Cunningham a tad more, because if that is better available info, perhaps they discuss how closely or accurately related the BMR is from their RMR calculated.

Oh, my acceptance of it being better is compared to Harris and Mifflin that overestimate when carrying extra fat weight. It underestimates less than those overestimate. From that light reading of Katch. Harris and Mifflin talked about a lot more.

EvgeniZyntx

Cunningham is available here:

http://ajcn.nutrition.org/content/33/11/2372.full.pdf+html

Notice that he states BMR and not RMR but this is corrected here on pg 50.
http://books.google.com.au/books?id=arqNy5e1t60C&q=cunnignham#v=onepage&q=bmr&f=false

So reading that, the Katch McArdle equation is just a shift corrected linear estimation from Cunnigham's work.

In Cunnigham paper, it is worth noting:

"Since the body composition of the subjects are not reported [from the Harris Benedict data], estimates of LBM were calculated from the equations below, where M is body mass in kg and A is age in years, assuming all subjects to be normal. These
were derived by substitution and rearrangement of the prediction equations for total body water published by Moore et al. (7).
male LBM = (79.5 - 0.24 M - 0.15 A) x M + 73.2
female LBM = (69.8 - 0.26 M - 0.12 A) x M + 73.2"

It's surprising that KM is given such accolades - LBM variance is such that these estimates are bound to generate large variance. So the value of these equations is statistical in nature - much like BMI, applied to individual contains a high variability and risk.

I think I'll look at more recent papers.

EvgeniZyntx

In extremely obese women:

"The Mifflin-St Jeor equation was most accurate method for REE assessment in extremely obese women. "
http://pen.sagepub.com/content/31/3/217.abstract

and very interesting result from http://ajcn.nutrition.org/content/80/5/1379.full where 61 percent of the eREE was predicted by



for a much larger population. Notice that the slope is the same as KM or Cunningham equations.

Then a nice review study (http://ajpendo.physiology.org/content/279/3/E539/T1.expansion.html) has this summary of the linear equations:

Author Descriptive Equation Reference
Owen et al. REE = 19.7 × FFM + 334 28
Mifflin et al. REE = 19.7 × FFM + 413 25
Luke & Schoeller REE = 20.0 × FFM + 585 22
Jensen et al. REE = 20.0 × FFM + 662 17
Ravussin et al. REE = 20.82 × FFM + 471 30
Ravussin et al. REE = 20.93 × FFM + 478.7 31
Elia REE = 21.11 × FFM + 450 6
McNeil et al. REE = 21.5 × FFM + 329 24
Heymsfield et al. REE = 21.6 × FFM + 302 14
Cunningham REE = 21.6 × FFM + 501.6 5
Ravussin & Bogardus REE = 21.8 × FFM + 392 29
Owen et al. REE = 22.3 × FFM + 290 27
Heshka et al. REE = 22.94 × FFM + 356.7 13
Owen REE = 23.6 × FFM + 186 26
Kashiwazaki et al. REE = 24.5 × FFM + 304 18

Mean ± SD REE = (21.5 ± 1.4) × FFM + (407 ± 128)

However their attempt to develop a tissue/organ model is a limited success:

"The derived whole body level and tissue/organ level REE-FFM models are general and unsuitable for individual REE prediction. Future studies are needed to extend these observations and to analyze gender- and age-related, hormonal, ethnic, and other sources of variation in REE-FFM relationships (8-10, 26,39). "

Of note, this article (http://ajpendo.physiology.org/content/278/2/E308.full) has quite a few references in the discussion if someone wants to dig deeper.

Sabine321

Found that on calculator.net
BMR = 10 * weight(kg) + 6.25 * height(cm) - 5 * age(y) + 5 (man)
BMR = 10 * weight(kg) + 6.25 * height(cm) - 5 * age(y) - 161 (woman)

The daily calorie needs is the BMR value multiplied by a factor with a value between 1.2 and 1.9, depending on the activity level.

Sarauk2sf

Tagging as all this nerd stuff is sexay!

heybales

Cunningham is available here:

http://ajcn.nutrition.org/content/33/11/2372.full.pdf+html

Notice that he states BMR and not RMR but this is corrected here on pg 50.
http://books.google.com.au/books?id=arqNy5e1t60C&q=cunnignham#v=onepage&q=bmr&f=false

So reading that, the Katch McArdle equation is just a shift corrected linear estimation from Cunnigham's work.

In Cunnigham paper, it is worth noting:

"Since the body composition of the subjects are not reported [from the Harris Benedict data], estimates of LBM were calculated from the equations below, where M is body mass in kg and A is age in years, assuming all subjects to be normal. These
were derived by substitution and rearrangement of the prediction equations for total body water published by Moore et al. (7).
male LBM = (79.5 - 0.24 M - 0.15 A) x M + 73.2
female LBM = (69.8 - 0.26 M - 0.12 A) x M + 73.2"

It's surprising that KM is given such accolades - LBM variance is such that these estimates are bound to generate large variance. So the value of these equations is statistical in nature - much like BMI, applied to individual contains a high variability and risk.

I think I'll look at more recent papers.

So actually another case of a study using a previous study, or studies, data, and trying to come up with better results. Except some vital info was missing, so lets estimate what it must have been.

Huh. Even the VO2max studies aren't that bad when they look at BF% as a variable to measure, they leave studies out that aren't accurate enough with that stat.

So indeed there is an assumed ratio of FM to FFM, because they ran calcs to create the ratio.

I'm disappointed. Actually like that synopsis of all the formula's, since some where only tested in specific cases and found they held true.

My Bodpod results even used an RMR formula based on Nelson. Which wasn't on the list.

Have you ever worn your Garmin (I think you said you had Garmin) over night to see what the HR does while sleeping?
I was really surprised I had avg lower HR sitting at work for 8 hrs than a night's sleep. Which didn't even hit the same HR until the last hr almost. I guess I was repairing well, and oxygen was needed.

EvgeniZyntx

Cunningham is available here:

http://ajcn.nutrition.org/content/33/11/2372.full.pdf+html

Notice that he states BMR and not RMR but this is corrected here on pg 50.
http://books.google.com.au/books?id=arqNy5e1t60C&q=cunnignham#v=onepage&q=bmr&f=false

So reading that, the Katch McArdle equation is just a shift corrected linear estimation from Cunnigham's work.

In Cunnigham paper, it is worth noting:

"Since the body composition of the subjects are not reported [from the Harris Benedict data], estimates of LBM were calculated from the equations below, where M is body mass in kg and A is age in years, assuming all subjects to be normal. These
were derived by substitution and rearrangement of the prediction equations for total body water published by Moore et al. (7).
male LBM = (79.5 - 0.24 M - 0.15 A) x M + 73.2
female LBM = (69.8 - 0.26 M - 0.12 A) x M + 73.2"

It's surprising that KM is given such accolades - LBM variance is such that these estimates are bound to generate large variance. So the value of these equations is statistical in nature - much like BMI, applied to individual contains a high variability and risk.

I think I'll look at more recent papers.

So actually another case of a study using a previous study, or studies, data, and trying to come up with better results. Except some vital info was missing, so lets estimate what it must have been.

Huh. Even the VO2max studies aren't that bad when they look at BF% as a variable to measure, they leave studies out that aren't accurate enough with that stat.

So indeed there is an assumed ratio of FM to FFM, because they ran calcs to create the ratio.

I'm disappointed. Actually like that synopsis of all the formula's, since some where only tested in specific cases and found they held true.

My Bodpod results even used an RMR formula based on Nelson. Which wasn't on the list.

Have you ever worn your Garmin (I think you said you had Garmin) over night to see what the HR does while sleeping?
I was really surprised I had avg lower HR sitting at work for 8 hrs than a night's sleep. Which didn't even hit the same HR until the last hr almost. I guess I was repairing well, and oxygen was needed.

Have not worn the Garmin/Polar in bed - I think the Garmin might be more interesting as it has the HR over time. I'll try it some day.

But back to the BMR equations, I think we can draw a few conclusion from reading these studies.

1) At best, the free fat mass (lean body mass) accounts for 60-70% percent of BMR. The following is sort of obvious but is always worth repeating:

-- maintaining FFM while losing weight is essential in keeping BMR up - losing just 4 kilo of LBM is the equivalent of losing at least 120 to 160 calories you can eat daily (change in BMR*activity factor, all other variables constant). Exercise, resistance train.

-- activity is essential - simply by being more active TDEE can be increased by 20% to 50%. As a weight loss strategy for soemone that is obese this will make intial loses (at high BMRs, low activity levels) much easier. Move your butt, people. And this does not mean go from the couch to doing 2 hrs cardio per day overnight. Simple going from full-on sedentary to active will show a 300-700 increase in calories burned per day. One can't out-exercise a bad diet, but one can boost the weight loss process. Exercise.

-- hormonal, race, medical, organ-weight, etc factors are non-negligeable but should not be either abused by the "special snowflake" argument nor by the "if you can't follow IPORM you've got an ED/metabolic damage" crowds. There is a lot more there and I should probably get my thoughts together and start a thread on that alone.

2) The variability in the studies and the statistical abuse of data means that the best estimation can be off by 20% to 30%. In a very large majority of people this error in BMR will result in an estimate with <10% for 95-99% of the population, with some caveats:

- if FFM is correctly calculated, it often isn't
- if activity levels are correctly evaluated, they often are not

That means that TDEE calculations will be have usually less than 12% to 18% error for most.
Therefore, using them as an absolute value does not make sense but the TDEE calculations can be taken as an excellent first estimation and adjusted up or down after several weeks of experience - the usual advice of "track everything, evaluate loss, readjust calories consumed up or down to achieve goals."

3) Another reason why weight loss should be gradual - if rapid weight loss increases the risk of loss of muscle (given than muscle build up/protection is a slower process than calorie deficit driven lipolysis), low calorie diets or high deficits will be more likely to have a direct (ie FFM driven) negative effect on TDEE/BMR.

All of the above seems sort of obvious.

One of the things I found new, to me, is that it is likely that the single most important factor in metabolic down-regulation is not hormonal but newtonian - loss of weight and loss of FFM result in lower TDEE needs (and likely there is a factor in there for "metabolic-dead*" FFM like mineral bone (non-marrow) vs muscle ratio).

I knew that hormonal factor resulted in a lower metabolic need per mass as weight increased (fat burns less) and even a lower metabolic need per mass of LBM (a sort of metabolic "lethargy effect" of fat mass on LBM) but it seems to be significantly lower than I thought for the general population. Of course adrenal/thyroid changes due to illness can toss that out the window.

It's been an interesting exercise - it places me in the middle of the road between the IPORM-absolutists and "We are all different" crowds. I'm afraid I'm going to get run over by simplifications.


ETA: Katch-McArdle isn't necessarily better - just captures an essential factor that was missing. It isn't even a research published equation. The variability of these equations makes it a reasonable tool as an estimator just like the many others but no more indivually valid. One can round to the nearest hundred calorie without a worry and work off of that.


*it isn't "dead" -- my own early research on bone growth in porous hip implants was on activating the osteoblast/clast balance to increase the rate of bone modeling - but that's a vast other subject.