Basically, I consider this protein thing
a. a non-sequitur
b. irrelevant to the issue of carbohydrate intake. You can eat 1-1.5 grams of protein per pound of lean body mass on a carbohydrate-based diet just as easily as on a low-carbohydrate diet. Quite in fact, you should.
Fine, if you want to argue that high-protein is better than low-protein, I'm with you. One researcher (Westerterp-Plantenga) has argued that the higher protein intake, rather than the low-carbohydrate intake itself, is the cause of the differences in the first place.
But don't pretend that it has anything to do with low- versus high-carbohydrate. Frankly, that I should have to make such a mickey mouse point to a couple of PhD's (or their lapdog, Anssi Manninen) is beyond me. But apparently, they can't understand that differences in protein intake have zero to do with differences in carbohydrate intake per se.
The next theoretical explanation for a metabolic advantage has to do with gluconeogenesis. This is just an unwieldy word for the production of new glucose from other stuff. The other stuff in this situation is amino acids, glycerol (the fatty acid backbone) and lactate.
And it's true that
a. this process requires energy
b. this process is up regulated on a ketogenic (very low-carbohydrate diet)
Unfortunately, the theorists advancing this idea didn't really quantify the effect that well in their paper (as I recall) in terms of how many extra calories per day it should amount to.
As well, it has to be weighed against the loss of thermic effect for replacing carbohydrate with fat (the effect is mild but contributes). One study I recall found that the higher-carbohydrate diet (compared to higher fat but not ketogenic) had about a 100 calories per day advantage (due to the differences in the thermic effect of carbohydrate versus fat) and you lose this when you stop eating all carbohydrates, any effect of gluconeogenesis has to be weighed against that.
Perhaps more importantly, one of the primary adaptations to ketosis (a state where blood ketone levels go above a certain concentration) is to decrease gluconeogenesis.
That is, over the first 2-3 weeks of being in ketosis, the body switches to using ketones for fuel instead of glucose, which decreases the need for gluconeogenesis. A metabolic advantage that becomes almost insignificant after 2-3 weeks seems hardly worth pinning the success of a diet on.
On this note, I would like to mention that, empirically (and realize that I've been getting feedback on ketogenic diets for nearly a decade now, man that makes me feel old), folks do seem to report somewhat more fat loss in the first 2 weeks on a ketogenic diet than you'd expect based on the deficit.
Of course, it could just be the extra water loss throwing off the calipers too. In any event, after those 2 weeks, the effect is gone.
Again, for the typical person, the average overweight individual who may be dieting for weeks or months (or longer) to achieve their goals, an effect that disappears after a couple of weeks seems hardly worth pinning the success of the diet on.
And now we come to the final data point, the recent studies suggesting greater weight and/or fat loss. There have been at least a half dozen (perhaps more, I lose count) over the past several years, usually finding slightly greater weight loss (the average difference is usually on a few kilograms) and some have noted greater fat loss (using DEXA or other accurate methods to measure body fat).
Now, I mentioned that the difference in weight loss could probably be attributed to water loss anyhow. But what about the fat loss?
Well, in the first place, many of them reported protein intake being higher in the low-carbohydrate group. See my comments above. We're not just talking about the carbohydrate content of the diet here when 4 different nutrients (protein, carbohydrate, fat and fiber) may all be changing. Drawing conclusions about only the carbohydrate content of the diet and ignoring the rest seems a bit myopic to me.
Beyond that, here's the bigger issue: without exception, all of the studies done have relied on self-reporting of food intake. And this is not a trivial issue. We've known for many years now that people on a mixed diet tend to underestimate their food intake by up to 50%. That is, someone eating a carbohydrate-based diet who says they are eating 1500 (6300 kilojoules) calories may really be eating 3000 calories (12,600 kilojoules).
But what about on low-carbohydrate diets?
Well, nobody has really looked to see whether people under- or over-report their food intake but we have other data. Studies done decades ago often reported spontaneous food intakes of 1600-1800 calories on low-carbohydrate diets. A recent study in diabetics found a 1000 calorie per day reduction in food intake with the shift to a low-carbohydrate diet.
Basically, people on high carbohydrate diets tend to underreport their food intake (they are eating more than they say) while people on low-carbohydrate diets tend to spontaneously eat less (for a number of reasons).
So when you have the low-carbohydrate group saying they ate 1600 calories and the mixed diet group saying they ate 1500 calories, yet the low-carbohydrate group lost more weight/fat, you tend to question it. The carbohydrate-based group could be eating 3000 calories, based on previous studies of underreporting.
Quite in fact, a recent study, by Brehm (who had done an early metabolic advantage study) directly measured a couple aspects of metabolic rate for high and low-carbohydrate diets. Finding no difference in anything (if anything, the high-carbohydrate group was slightly superior, as the thermic effect of food in response to a meal was higher).
The researchers concluded that the difference in weight/fat loss is probably due to under-reporting of food intake in the carbohydrate-based group.
Along with this, there are several key studies (which the metabolic advantage people like to ignore) where calories were rigidly controlled.
In one, a group of patients in a hospital was placed on a variety of experimental diets for 2 weeks. Protein was kept static and carbohydrate was varied from 0 to 70% of total calories, while fat varied in the opposite direction. Activity was controlled since they were bedridden. Calories were controlled with liquid diets. They found no difference in the number of calories needed to maintain bodyweight.
And this is really my big issue with the whole idea: if low-carbohydrate diets generate a metabolic advantage, it should be measurable with current technology. If it's not measurable, it either doesn't exist is far too small to worry about. And all of the theoretical calculations for what should occur don't change that. Especially when we have much more likely mechanisms for the effect.
The more likely explanation in my mind is that any 'metabolic advantage' inherent to low-carbohydrate diets come from the fact that they tend to blunt hunger (and this is especially true in people who are overweight and hyperinsulinemic, people with insulin resistance) and make people eat less.
And even that isn't guaranteed, people who don't have their hunger blunted, or who fall into the "I can eat whatever I want as long as it's low carb" camp and end up overeating calories don't lose weight or fat at all.
The bottom line in my mind: even if low-carbohydrate diets turn out to have a small metabolic advantage (I remain open to the idea but skeptical based on the data to date), it still comes down to caloric intake.
Q. Some claim that that your body will go into 'starvation mode' if you eat too few calories, preventing you from losing weight and that trying to lose weight by eating fewer calories doesn't work. What do you think?
A. Well there is no doubt that the body slows metabolic rate when you reduce calories or lose weight/fat. There are at least two mechanisms for this.
One is simply the loss in body mass. A smaller body burns fewer calories at rest and during activity. There's not much you can do about that except maybe wear a weighted vest to offset the weight loss, this would help you burn more calories during activity.
However, there's an additional effect sometimes referred to as the adaptive component of metabolic rate. Roughly, that means that your metabolic rate has dropped more than predicted by the change in weight.
So if the change in body mass predicts a drop in metabolic rate of 100 calories and the measured drop is 150 calories, the extra 50 is the adaptive component. The mechanisms behind the drop are complex involving changes in leptin, thyroid, insulin and nervous system output (this system is discussed to some degree in all of my books except my first one).
In general, it's true that metabolic rate tends to drop more with more excessive caloric deficits (and this is true whether the effect is from eating less or exercising more); as well, people vary in how hard or fast their bodies shut down. Women's bodies tend to shut down harder and faster.
But here's the thing: in no study I've ever seen has the drop in metabolic rate been sufficient to completely offset the caloric deficit. That is, say that cutting your calories by 50% per day leads to a reduction in the metabolic rate of 10%. Starvation mode you say. Well, yes. But you still have a 40% daily deficit.
In one of the all-time classic studies (the Minnesota semi-starvation study), men were put on 50% of their maintenance calories for 6 months. It measured the largest reduction in metabolic rate I've ever seen, something like 40% below baseline. Yet at no point did the men stop losing fat until they hit 5% body fat at the end of the study.
Other studies, where people are put on strictly controlled diets have never, to my knowledge, failed to acknowledge weight or fat loss.
This goes back to the under-reporting intake issue mentioned above. I suspect that the people who say, "I'm eating 800 calories per day and not losing weight; it must be a starvation response" are actually eating far more than that and misreporting or underestimating it. Because no controlled study that I'm aware of has ever found such an occurrence.
So I think the starvation response (a drop in metabolic rate) is certainly real but somewhat overblown. At the same time, I have often seen things like re-feeds or even taking a week off a diet do some interesting things when people are stalled. One big problem is that, quite often, weekly weight or fat loss is simply obscured by the error margin in our measurements.
Losing between 0.5 and 1 pound of fat per week won't show up on the scale or calipers unless someone is very lean, and changes in water weight, etc. can easily obscure that. Women are far more sensitive to this. Their weight can swing drastically across a month's span depending on their menstrual cycle.
Thing is this, at the end of the day, to lose weight or fat, you have to create a caloric deficit, there's no magical way to make it happen without affecting energy balance. You either have to reduce food intake, increase activity, or a combination of both.