I've started reading Dr. Richard D. Feinman's book, The World Turned Upside Down: The Second Low-Carbohydrate Revolution. It may be the best book on the topic of low carb diets to date, because of its broad perspective, and its readability. It appears to have at its core the same thesis Zooko and I have been trying to promote on our website, that low carbohydrate diets are good for health in many ways, particularly in ways connected to metabolic syndrome. In the book, Dr. Feinman expresses a problem that comes with that discovery, a problem which has caused us much personal frustration:
"[T]he problem with convincing people of the benefits of a reduced carbohydrate strategy is that it appears to be good for everything, good for what ails you. You can sound like a hard-sell pitchman." (p 204)
Another passage that particularly struck me was in the introduction. It has given me vocabulary for a concept we've tried to express several times.
"[B]iology tends to run on hormones and enzymes, that is, control mechanisms, not on mass action (the principle that chemical processes are determined by how much reactants are put into them). The grand principle in biochemistry is that there is hardly anything that is not connected with feedback." (p 7, Emphasis ours)
This point was made in the context of the diet-heart hypothesis, which has as a premise that you can control the amount of cholesterol in your blood by how much of it you eat (which is untrue). It is an equally appropriate concept when talking about the calorie control method for weight loss. That method is generally a lost cause, because the important mechanisms in weight control are hormonal. How much you eat is a downstream effect of your hormonal state.
Protein does not affect ketosis via mass action
It also explains well the idea Zooko and I have tried to put forth about protein and gluconeogenesis (GNG, the process of making sugar out of protein). We showed in this post that the amount of material available for GNG doesn't appear to have any effect on how much GNG actually occurs. Insofar as eating a high amount of protein can lead to reduced ketosis, which would then require your body to get more energy from glucose, which could increase demand for GNG, this must occur through a hormonal/enzymatic cascade.
That would mean that managing ketosis is not a simple matter of calculating some threshold of protein, after which the rest "turns into sugar", any more than managing weight is a simple matter of calculating some threshold of calories, after which the rest "turns into fat".
Most people find that if they restrict carbohydrate intake, excess fat is lost. The loss of fat entails a caloric deficit, but that is an effect of the fat loss, not a cause. One could argue about what would happen if people on low carb diets consumed excess calories, but it's largely irrelevant, because people on low carb diets following their hunger rarely do that.
The same may be true of protein. As far as I can tell, most people on a very low carb diet are in ketosis without consciously constraining their protein consumption. It happens naturally.
While I have heard from some people who need to manually manage calories or protein to stay in their therapeutic zone, even while on a very low carb diet, it doesn't seem to be the common case. Even in those cases, I have often seen the problem resolve when a high-fat, plant-free, sweetener-free approach is taken. This suggests that there are further (hormonal/enzymatic) mechanisms that can interfere with the hunger feedback loop.