Saturday, 7 June 2014

Bray et al shows that a calorie *is* a calorie (where weight change is concerned).

Continued from Everyone is Different, Part 3.

EDIT: I made an error in stating that all of the extra calories came from fat, in the fat overfeeding phase. Thanks to commenter CynicalEng for pointing that out. It doesn't change the conclusion at all.

At 01:17 on 6th June, during a Facebook discussion, Fred Hahn told me:-
"Nigel Kinbrum - read this please.
Bray, et al. Shows that a Calorie is Not a Calorie and that Dietary Carbohydrate Controls Fat Storage.
Perhaps you'll learn something from a real expert who teaches metabolism to medical students at the largest medical school in the country."

So I did.

At 02:22, I replied:-
"Thanks for that. I read Feinman's blog post about Bray et al some time ago.
There's a fundamental error in Feinman's analysis. As LeonRover pointed out in his comment
In Diets:- "Absolute carbohydrate intake was kept constant throughout the study."
Also, in COMMENT:- "The extra calories in our study were fed as fat, as in several other studies, and were stored as fat..."
Oh, whoops! That may be why it was rejected by the editor."

Here's Figure 6 from Bray's study.

Some Definitions:-

LBM = Lean Body Mass
FM = Fat Mass = Body Fat

Weight change = (LBM change + FM change)

Weight change varies from ~+3.5kg (@ +2,500kJ/d) to ~+9.1kg (@ +5,900kJ/d).

(Maximum weight increase)/(minimum weight increase) = 2.6
(Maximum kJ/day increase)/(minimum kJ/day increase) = 2.36

∴ A calorie *is* a calorie (where weight change is concerned) ± some inter-personal variation.
∴ Insufficient protein can result in LBM loss (this is bad).

As LBM has a lower Energy Density (~600kcals/lb) than FM (~3,500kcals/lb),  LBM loss can increase weight loss, when in a Caloric Deficit.

See The Energy Balance Equation, for a simple explanation, and The Dynamics of Human Body Weight Change, for an incredibly complicated one!

I was rather chuffed when Alan Aragon left the following comment at 04:34:-
"Nigel is correct. From Bray et al's text:
"The extra calories in our study were fed as fat, as in several other studies [33,34], and stored as fat with the lower percentage of excess calories appearing as fat in the high (25%) protein diet group. The higher fat intake in the low protein group probably reduced nutrient absorption (metabolizable energy) relative to the other groups and this would have brought the intake and expenditure closer together in this group.""

Feinman has deleted his blog post. However, his post I Told George Bray How to do it Right is still there. I believe that Dr. George A. Bray M.D. sort-of did it right.

Dr. George A. Bray used a "weight maintenance formula" in all three groups for the weight maintenance phase. He then changed the formula in all three groups to low-P, med-P and high-P formulas, for the fat overfeeding phase. Carbohydrate grams remained constant in all three groups for all phases, but additional fat grams were fewer in the high-P group than in the low-P group, for the fat overfeeding phase.

I would have used the low-P, med-P and high-P formulas for the weight maintenance phase and for the fat overfeeding phase, to equalise the additional fat grams in all three groups.

Continued on Everyone is different Part 4, Fallacies and another rant!


George said...

"The higher fat intake in the low protein group probably reduced nutrient absorption (metabolizable energy) relative to the other groups"
Leaving aside the meaning of "probably" in a science paper when data about probability are not available, that's quite interesting. A calorie not absorbed or metabolised is still a calorie, or not?
There's a bit about that in this, my new favourite science paper:

Nigel Kinbrum said...

Calories In = Calories entering mouth - Calories exiting anus
So, non-absorbed calories reduce Calories In.
Technically, the above equation should also include Calories lost in urine, breath, sweat & any other bodily substances containing energy (!)

Pork is my favourite meat.

George said...

Well, while we're going there, this bit from Barendse's review is relevant and was news to me:
In addition, fat has a unique ability to stimulate colonic contraction, not found for carbohydrate or protein, and these contractions are essential to the voiding of faeces (Wright et al. 1980). Indeed, remedies for constipation before the Second World War recommended, in addition to ‘roughage’, water and the avoidance of diuretics, the consumption of fat such as bacon fat, butter, or olive oil, especially if the faeces was small and dry (Hutchison 1936), and for spastic colon no fibre and large amounts of the above listed fat was recommended. This is not a recommendation that is in any current medical guideline. Yet several pieces of evidence show that it is valid. First, infants take in no fibre, so amount of fibre is not a variable. Animal fats in the formula result in stools that are softer, larger, and that are easier to pass than plant fats (Forsyth et al. 1999), and babies also grow better when the source of the fat is from animal rather than plant sources (López-López et al. 2001).

There's more.

Nigel Kinbrum said...

Ah, calcium soaps. I remember them. I've read studies on faecal fat excretion/egestion. That's why high-calcium milk products such as yoghurt reduce Calories In by increasing Calories exiting anus.

Babies should drink milk, preferably human, rather than "formula". Dr Art Ayers covered that in

CynicalEng said...

Objectively, Bray's paper demonstrated that in 40% calorie overfeeding diets of variable composition but constant carbohydrate the *weight* gain was not the same "Overeating produced significantly less weight gain in the low protein diet group". Daily body fat gain was very similar across the protein groups and accounted for ~2/3 of the extra calorie intake.

Your second graph above needs redrawing as the incremental carb/protein intake in the three groups was 97, 67 and 17 g/day over baseline for LP, NP and HP respectively. I can understand how Bray's text hid this from the reader by saying it was all added as fat. It wasn't. If you have extracted the data it would be interesting to see what that regression looks like, with an R^2 value.

CynicalEng said...

my bad - spreadsheet error - the change in P+C was -38, +47 and +135 respectively. More coffee needed.

Nigel Kinbrum said...

I'm still bemused as to why Bray didn't use the same diet for weight maintenance, for overfeeding in each group. He didn't do it quite as right as I first thought :-/

Weight change (LBM change + FM change) is still proportional to added kcals (± interpersonal variation), as it varies from ~+3.5kg (@ +2,500kcals) to ~+9.1kg (@ +5,900kcals).

I'd need to add 2 more graphs, as P change is different from C change (zero) is different from F change.

I'll just delete the existing graphs and insert Figure 6 from the study!

CynicalEng said...

The added cals was about the same across groups ? LP was middle of the three in the table Ein, but the difference was only P=0.13. Shame NP was all black men - adds more fog to the battlefield.

Nigel Kinbrum said...

The LP group contained the lightest subject and were, on average, the lightest.

The NP group contained the heaviest subject and were, on average, the heaviest. Black people tend to have more muscle mass than white people. There were 6 males & 3 females, according to

The HP group were of intermediate weight.

Randomisation is hard!