Metabolic Advantage, Obesity and Eric Jequier

Since I had the whole metabolic advantage thing on my mind writing my last post, and "Anonymous LC Author" mentioned Feinman and Jequier, I was reminded of something that's bothered me for a while now regarding citations of Jequier's work. 

A calorie is a calorie violates the second law of thermodynamics argues for the potential of a metabolic advantage in low carbohydrate diets.   Oddly enough, they ultimately make a first law argument for the MA, but that's a subject for another day.

It was this paper that introduced me to the work of Eric Jequier after which I stumbled upon the subject of my most popular post to date, Nutrient Fates after Absorption referencing this Jequier paper:  Nutrient effects: post-absorptive interactions.  But the Feinman & Fine paper cites a different Jequier paper to which the TEF values of macronutrients can be ubiquitously traced in LC circles.  That paper being:  Pathways to Obesity.  

Surely that paper is all about how we eat too many carbohydrates, right?  One might think so given that Jequier and his reported TEF differences are so often cited by those advocating for the supremacy of  low carb and/or those fingering carbohydrates as the cause of obesity.  But when one takes a closer look we see he feels quite differently:

The prevalence of obesity is increasing worldwide, which indicates that the primary cause of obesity lies in environmental and behavioural changes rather than in genetic modifications. Among the environmental influences, the percentage of fat energy of the everyday diet and the lack of physical activity are two important factors, which contribute to explain the rising prevalence of obesity. In this review, several lines of evidence are presented to illustrate why dietary fat does affect obesity development. There are four factors which support a link between dietary fat and obesity development:

1. The thermic effect of nutrients, expressed as percentage of their energy content, is 2 – 3% for lipids, 6 – 8% for carbohydrates and 25 – 30% for proteins. This means that the efficiency of nutrient utilization (calculated as 100%—the thermic effect of the nutrient) is higher for fat than for carbohydrate or protein.

2. Postingestive fuel selection favours the oxidation of dietary proteins and carbohydrates, whereas dietary fats are preferentially stored as triacylglycerol in adipose tissue. Alcohol, by inhibiting lipid oxidation, indirectly favours the storage of dietary fats.

3. High-fat diet promotes excessive energy intake by passive overconsumption; the fat-induced appetite control signals are too weak or too delayed to prevent excessive energy intake from a fatty meal.

4. The only proof that dietary fats contribute to weight gain is to test the long-term effect of ad libitum low-fat diets. Most studies on low-fat diets show that they induce a modest weight loss in obese individuals, but their long-term effect from a public health perspective is limited, probably due to a low compliance to the dietary advice

Now you or I may not agree with that, but I find it somewhat disingenuous to cite this paper where Jequier discusses obesity in an energy balance context.  Excerpt:

There are two mechanisms by which the utilization of ingested nutrients can affect energy balance:

1. the efficiency of nutrient utilization

2. the postingestive fuel selection or ‘nutrient partitioning’.

In (1), Jequier cites his now familiar TEF factors that work out to efficiencies in metabolism of 97-98% for fats, 92-94% for carbs and 70-75% for proteins.   So Feinman and Fine used the averages (97.5%, 93%,72.5%) and calculated the net caloric load of a 2000 cal diet of 55/30/15% for C/F/P and determined this yielded 1848 net calories.  They then do calculations by progressively reducing carbohydrate and replacing it with equal amounts of protein and carbohydrate, and after going on and on about the second law, proceed to do first law-based calculations of the effective caloric content of the diet.  Here's a screenshot of my calcs and F&F's plot:
  

Feinman & Fine go on to say that at 21% carb, a 100 cal deficit is achieved, and that 8% carb (hence the added line in my table), a 140 cal deficit is achieved.  Eight percent being the carb intake of the first phases of most carb restriction plans.  I'm beginning to think that I've uncovered a significant error in F&F's math, because my calories come out a bit differently, and my deficits to about half that of F&F's.  Indeed we don't even get to a 100 cal deficit by my math at 5% carb.  I've had email contact with Feinman before so I might shoot him a quick note and see if there's something I'm doing wrong here.  *See Math Addendum at the end of this post.

Zoe Harcombe really takes it to the extreme with her own calculations for a "high protein diet".  

"I repeated the calculation for a 10:30:60 high protein diet, as another example, and the yield drops to 1,641 calories."

This woman must be blinded by her ideology because not only would F&F's progressive reductions never get us to 10/30/60, but what low carbers eat like this anyway?  Those proportions may well be representative of a protein sparing modified fast, or, perhaps some version of KK (and I've seen a fast fat loss diet from Lyle McDonald that might come out around these percents, perhaps even lower in fat and carb by percent).  She's replacing carbohydrates 1-for-1 with protein!  In any case, Zoe's math (in agreement with mine) nets 1641 calories which would be a 185 cal deficit vs. the reference 55/30/15 diet.  WOW!  That sounds great doesn't it?  Ummm ... 60% protein out of 2000 cals would be 300g of protein while only consuming 67g fat.  If someone were even able to eat 300 g of protein (I've seen a ceiling of around 250g  EDIT: Theoretical protein intake appears to be over 400g), Zoe speaks of eating a lot of fat, not to shun fatty meats (although as a vegetarian who now occasionally eats fish, she doesn't practice what she preaches to others), etc.  IOW, in bashing WW in that article, she uses an extreme example that is neither viable nor what she advocates.  But that's really a subject for a different post, so back to the subject at hand .....

So if we're going to play math games with TEF efficiencies, why not do it with the percentages that low carbers commonly consume or advocate?  The first "game" we can play is to keep protein constant and replace the carb with fat.  This is the reverse of what LC advocates claim is the cause of the obesity epidemic:  that we were told to replace fat with carbs in our low fat diets to lose weight or not gain weight.  Here are the results.   And the metabolic advantage goes to?  Carbs.

For the next game, let's look at some typical LCHF diets, because I'm told over and over how I "don't do LC right" often eating a lower fat (not low fat!) higher protein version.  A "proper" LC diet is by definition a high fat/moderate protein diet, upwards of 60% fat, and there's the VLC/VHF contingency eating 75%-80% fat.  At 60% fat there's an advantage attributable to increased protein consumption vs. the reference diet.  But up the fat to 70% and it's either a wash or a slight disadvantage if one has just a few more carbs, and 80% VHF offers up a clear disadvantage.

In at least one study, though clearly not long enough in duration, swapping 15% fat calories for protein (carb  50%,  15% protein/35% fat  vs.  30% protein/20% fat) for two weeks, subjects ate weight maintaining caloric levels of each diet and did not lose weight.  I calculate about a 90 cal/day differential using Jequier's average TEF factors for ~2338 cal/day maintaining diet, so this would not rise to a level of statistical significance over 2 weeks, but for the next 12 weeks, the study participants (it wasn't a weight loss study!) were given extra food for the higher protein diet:

The subjects maintained a significant decrease in spontaneous caloric intake, relative to baseline (CRC2 compared with CRC3: 441 ± 63 kcal/d; P < 0.001), through the end of the study. This sustained decrease in spontaneous caloric intake resulted in a constant rate of weight loss during the 12-wk ad libitum high-protein diet, which amounted to 4.9 ± 0.5 kg lost by the end of the study (Figure 1). A decrease in body fat mass accounted for 3.7 ± 0.4 kg (76%) of the weight lost between CRC2 and CRC3, and the overall weight loss was fully explained by the cumulative reduction in caloric intake. 

Over 12 weeks,  I calculate about a 70 cal/day difference for the 1884 cal/day intake for these compositions -- essentially we're replacing 15% of the total intake of the lowest TEF macro (fat) with the highest TEF macro (protein).  Over 12 weeks, we're talking almost 6000 cal but it's doubtful this would translate to a difference in weight loss that would reach a level of statistical significance.  

In any case, given how calories were determined by Atwater, I suspect that at least part of the TEF is already incorporated into our average 4/4/9 factors.  I basically come down on the side of Buchholz and Schoeller that, in the end, a calorie is just a calorie.

Back to Eric Jequier, and point 2 from Pathways to Obesity, that being the role of postingestive fuel selection or ‘nutrient partitioning' on energy balance.  The discussion in this paper contains a sort of "Cliff Notes" version of his other paper.  Certainly a more layperson-friendly read!   I won't rehash that entire discussion here, but here's a summary quote:

In conclusion, both the efficiency of fat utilization and the postingestive fuel selection of fat indicate that this nutrient has a greater metabolic potential to induce weight gain than dietary carbohydrates or proteins.  The conversion of glucose into fatty acid, de novo lipogenesis, is a biochemical process which could invalidate the concept of individual nutrient balance if it was quantitatively an important pathway. Two methods of investigation have been used to assess de novo lipogenesis in humans: indirect calorimetry and stable isotope techniques. A significant net lipogenesis has been observed in humans only with experimental massive carbohydrate overfeeding, a condition that does not occur in everyday life.

Given that just about any discussion on MA I've seen on the net eventually leads someone to cite Feinman and Fine, and key to that paper is Jequier's Pathway to Obesity, it seems evermore obvious that we have a deliberate misrepresentation of metabolism by many of the LC "experts" and advocates.  But as Mark Sisson stated recently, it's all really about the buy-in.  If a snazzy gimmick gets more people to follow such diets, then it's all good, right?

*Math Addendum:

 2000 cals in a 55/30/15% C/F/P diet proportion yields 1100/600/300 cals

1100*0.93 + 600*0.975 + 300*0.725 = 1825.5 cals

 2000 cals in a 21/47/32% C/F/P diet proportion yields 420/940/640 cals   

420*0.93 + 940*0.975 + 640*0.725 = 1771.1 cals

 2000 cals in a 8/53.5/38.5% C/F/P diet proportion yields 160/1070/770 cals   

160*0.93 + 1070*0.975 + 770*0.725 = 1750.3 cals