Thursday, January 15, 2015

Does high protein explain the low-carb "metabolic advantage"?

In 2012, David Ludwig's group published a paper that caused quite a stir in the diet-nutrition world (1).  They reported that under strict metabolic ward conditions, weight-reduced people have a higher calorie expenditure when eating a very low carbohydrate diet (10% CHO) than when eating a high-carbohydrate diet (60% CHO)*.

In other words, the group eating the low-carb diet burned more calories just sitting around, and the effect was substantial-- about 250 Calories per day.  This is basically the equivalent of an hour of moderate-intensity exercise per day, as Dr. Ludwig noted in interviews (2).  The observation is consistent with the claims of certain low-carbohydrate diet advocates that this dietary pattern confers a "metabolic advantage", allowing people to lose weight without cutting calorie intake-- although the study didn't actually show differences in body fatness.

In Dr. Ludwig's study, calorie intake was the same for all groups.  However, the study had an important catch that many people missed: the low-carbohydrate group ate 50 percent more protein than the other two groups (30% of calories vs. 20% of calories).  We know that protein can influence calorie expenditure, but can it account for such a large difference between groups?

A new study gets us part of the way to answering that question (3).  This is a follow-up to a study that was published by the same authors in 2013 (4).  For 8 weeks under strict metabolic ward conditions, researchers overfed 25 volunteers by 40 percent of their normal calorie needs.  They compared three groups at different levels of protein intake:

  1. Overfeeding with 5% protein.
  2. Overfeeding with 15% protein.
  3. Overfeeding with 25% protein.
Protein intake was increased at the expense of fat, so carbohydrate intake was similar in all groups.

After eight weeks, the calorie expenditure of the groups differed substantially.  It increased in all groups, as expected due to the overfeeding, but it increased much more in the high-protein group (group 3).  Within one day, total calorie expenditure in the high-protein group had increased by 130 Calories.  By the end of the 8-week period, the high-protein group was burning about 260 Calories more per day than the low-protein group, and about 180 Calories more than the medium-protein group!

This demonstrates that protein intake can have a large effect on calorie expenditure-- equaling the magnitude of moderate daily exercise.  Although it was a different type of study than what Dr. Ludwig published, when this study is considered along with the rest of the evidence, it suggests that there probably is a low-carb "metabolic advantage", but that this advantage is probably due to increased protein intake, not reduced carbohydrate intake.

The companion study also found, strangely, that all three groups gained the same amount of fat during overfeeding, despite the differences in calorie expenditure.  This is the same thing Dr. Ludwig observed, although his study was only four weeks long.  It's possible that 4-8 weeks isn't long enough for these differences in calorie expenditure to manifest in fat mass-- or perhaps something more mysterious is going on that will require further investigation**.  However, the higher protein groups did end up with more muscle mass than the low-protein group.

* They found that a moderate-CHO low-glycemic diet led to an intermediate calorie expenditure, although it wasn't statistically different from the other two conditions.

* For example, perhaps indirect calorimetry isn't the best tool for measuring calorie expenditure in this type of intervention.  The use of indirect calorimetry to measure calorie expenditure rests on certain assumptions, which to my knowledge have not been well validated in the context of overfeeding and weight-reduced people.  My friend and colleague Karl Kaiyala published a paper in 2011 discussing some of these issues (5).  It's possible that indirect calorimetry gives erroneous measurements in some contexts.

Thanks to Erik Arnesen for tweeting the study, and Pedro Bastos for sending me the full text.


Jeremy said...

It is possible the excess calories are spent on beneficial body activities. For example, could your brain work more on useful tasks? Could you stay warmer during cold weather? Could you be a better athlete?

D-Rock said...

I think your second point (**) on the indirect measurement of calorie expenditure is the crucial one. It seems the increased energy expenditure in the high protein groups should cause a decrease in the amount of weight or fat gained (as you and the authors alluded to). While it is possible that at 4 or 8 weeks this 'metabolic advantage' has not caused enough of an effect to generate a difference in fat or weight gained, the simpler explanation may be that the calorie expenditure methodology may not be good enough or may be 'fooled' in some way by the increased protein in the diet. This is a shame because the authors used the more reliable metabolic ward study, but then used a less reliable technique to measure metabolism/energy expenditure. No study is perfect.

Roland Denzel said...

Good stuff, Stephan!

Unknown said...

Would be interisting to see weightchanges for liver and kidney. Running these two organs all day full-throttle isn't exact the metabolic advantage the low-carb experts telling us.

J said...

Fascinating research. There is a cost to a high protein diet, though, isnt there? Compromised long term health; according to the plant based doctors. I think I prefer to stick to lower protein as a by calorie percentage of macronutrients (10-15%).

CPB said...

Atkins (New Diet Revolution, 2002) also mentioned the metabolic advantage of Harry Kronberg’s 1930 Cal low carb diet (~53g Carbs, my estimation) over his original 2127 Cal diet (~ 209g Carbs, also my estimation). It was considered nutritional heresy at the time (it was an uncontrolled, N=1 observation). It appears now that he might have been correct.

The metabolic advantage of low carb diets might be due to the additional energy (from protein, glucose and fat) needed to synthesize the ATP and GTP molecules that are necessary to synthesize glucose in the glucose-starved state.

The average person needs only about 160g of Carbs/day (mostly for brain function), yet the average American consumes almost twice that. If you consume less than 160g of Carbs/day on a low carb diet then you will deplete your liver glycogen and tissue stores (~210g of Carbs). After that, the liver (and kidneys) begin gluconeogenesis to supply the required glucose. All of the required glucose can be synthesized in this way. However, there is an added metabolic cost to gluconeogenesis. Gluconeogenesis requires additional ATP and GTP molecules ( i.e., energy, macronutrients and Calories) to synthesize the required glucose than simply converting it from glycogen or Slurpees.

Ref: Gluconeogenisis:

Phil Koop said...

"perhaps indirect calorimetry isn't the best tool for measuring calorie expenditure"

If this experiment had been intended to demonstrate this conjecture - that the bioavailability of calories from protein is underestimated relative to that of fat - would that not have been the conclusion? And since the experimental design would not be affected by this intention, why should we not make this conclusion?

You don't explicitly say so, but I assume that since all groups gained the same amount of fat but the high-protein group gained more muscle, they gained more total mass. The alternative - that they lost a compensating amount of other mass (bone? skin? connective tissue? internal organs?) - would be pretty scary. And if they gained more mass, how could they truly have both consumed the same number of calories yet expended more? The proposition that more mass was created from fewer calories seems unlikely compared to the proposition that caloric measurement was incorrect.

Unknown said...

I was under the impression that the metabolic advantage (via much larger thermic effect) of a higher protein diet was well supported, as well as its tendency to spare muscle tissue during hypocaloric dieting.

Anonymous said...

Hi Stephan.

Excellent post. Very nice and very interesting commentary. Progress is being made.

Dr. Ludwig is a really nice guy and I spoke with him. He also agrees with me about the conservation of energy principle being a very rampant problem on the Internet. He even wrote a paper in JAMA in June about overly -simplistic interpretations of it with regard to obesity.

The research you post is highly interesting :)

Despite the erroneous protests of some Internet gurus with a financial gain at stake and loss of face, nothing in a possible metabolic advantage "violates" the conservation principle- only these gurus' childish understand of what it says. It is only a "state equation". Whether a metabolic advantage exists or not is a biochemistry/physiology question. Its existence is not in any way in contradiction to the conservation of energy in any way. Anybody disputing it on the basis of the conservation laws is wrong.

As Richard Feynman noted, all of "Dennis The Mennace's toy blocks" can eventually be accounted for but it can be very complex. There are limitations in measuring and accounting for alllll of it in human bioenergetics etc. Thus far, it is reliable etc.

How the body partitions, uses, absorbs and handles energy is not at all addressed. Energy can be led to many different pathways as biomedical researcher Dr. Garret Fitzgerald noted to me. As Jeremy above asked, my reply is that some excess energy can be wasted as dissipated heat ( 123 watts) or pooped out ( 7 % of total energy or so ) .Those are some options.

I applaud your scientific attitude open to the possibility one exists. I personally think it just may. :)

Best wishes,
Razz ( sorry I have to use Hafthor account today due to difficulties)

Carl said...

CPB you said "The average person needs only about 160g of Carbs/day (mostly for brain function), yet the average American consumes almost twice that." I knew that there was a daily protein and fat requirement but I didn't know that there was a daily carb requirement"

Vincent Gabriel said...

In fairness to how Low Carb in practice has evolved, the primary mechanism for why it works so well, at least as I understand it, is the shift in the body’s fuel preference toward fat (and hence the use of stored fat for energy) that occurs when the ratio of dietary calories shifts from carbs to fat. (There are studies and discussion of this on Suppversity and Eating Academy, among other places.) The macronutrient comparison used in the study here would now be called “politically correct low carb” and I think most low carb advocates refer to the best current protocol as Low Carb High Fat (LCHF) to be clear that the fat proportion is raised.

Something I experienced when I lost, and kept off, a significant amount of body fat on LCHF is a dramatic reduction in hunger and a corresponding reduction in the amount of calories I took in until my body composition reached steady state, so any study that uses the same number of calories while changing the profile of macros is inherently a mis-representation of how one of the mechanisms behind LCHF works in practice.

That said, I am aware that some people benefit from a high protein stage in their transition to a higher-fat-burning steady state (Robb Wolf used to be an advocate of this and perhaps still is), but at least for me, shifting from very high-carb/low fat to just the opposite worked fantastically well and pretty much exactly the way its advocates describe.

Max said...

This doesn't add up. The highest protein group is consuming the same number of calories, weighs more presumably (more muscle and same fat mass) and has a higher metabolic rate?

majkinetor said...


Muscle is more metabolically active, so it does add up.

Unknown said...

Stephan, would you comment on this essay?

Max said...


Exactly, muscle is more metabolically active. Meaning that if the high-muscle group consumes the same number of calories as the other groups, they should weigh less. But they don't. It doesn't add up.

Unknown said...

My understanding, which is a rough remembrance of reading every post on leangains 5 years ago (particularly the alcohol post, iirc), is that this is largely a part of TEF - the thermic effect of food. Effectively, due to TEF, protein could be considered as effectively lower in calories (closer to 3 than 4). Less so with carbs, and least of all with fat. In a hypercaloric state, where protein would be converted to fat, there would be energy expended on the conversion as well as inefficiencies in the process.

The increased fat gain is peculiar, particularly since the high protein group gained more muscle - I wouldn't expect the group with greater calorie expenditure eating less fat and more protein to gain more weight (same fat + more muscle).

Anonymous said...

If you're doing LCHF right you're eating MODERATE protein, not "high protein". Also, high protein was only ever shown to cause kidney problems in lab rodents which already had pre-existing kidney disease...................................................