A calorie is what?
Calories-in calories-out, diets, thermodynamics and all that - 1
The original promise of low-carbohydrate and ketogenic diets was that more weight is lost calorie-for-calorie than with other other weight reduction methods. There was resistance to the idea and there still is. The principle is now identified as calories-in calories-out (CICO). Adherents typically refer to the laws of thermodynamics. They mean the first law which they cite as “energy can neither be created nor destroyed.” I got into nutrition — I am trained as an enzyme and protein chemist — because it was obvious that thermodynamics was not understood and I thought I could explain things. The argument for greater weight loss on low-carbohydrate diet is actually experimental. It has been demonstrated in numerous ways. It doesn’t always happen but CICO is not a general rule — too many exceptions. In this post, I will try to explain in detail the errors in CICO and give a view of what thermodynamics is about. The Conclusion at the end of the post is reproduced here to make it clear where I am going or, if you only want the bottom line.
Conclusion: Calories-in Calories-out (CICO) as conceived as a mechanism of energy balance in nutritional studies is not correct. It is based on a limited understanding of the first law of thermodynamics and uses inaccurate definitions of calories. Chemical energy resides in the reaction not the reactants and products. It is a process not a thing. Thermodynamics more or less predicts that isocaloric diets of different compositions will have different efficiencies (ratio of useful energy to stored fat).
Calories in nutrition often refers to a specific value determined for complete oxidation as measured in the calorimeter. It is not directly applicable to other reactions.
Insofar as the outcomes of two diets have similar metabolic effects, it is a consequence of biological mechanisms — extensive feedback system — and is not predicted by any physical law. The Carbohydrate-Insulin-Model (CIM) is not in conflict with CICO. It refers to mechanism rather than energy metabolism. It is not really a model in that insulin is known to be anabolic and it is likely that CIM plays some role in the true mechanism.
The thermodynamics of the CICO argument is incorrect. That two isocaloric diets with different nutrient composition might have the same relative ability to store fat or provide useful energy is not predicted by thermodynamics. In fact, it would be very unlikely. Where it does occur, It is due to the characteristics of biological systems rather than any physical principle. The problems of obesity and more generally, metabolic disease remain as serious health threats and the goal is to explain how metabolic efficiency is controlled. I and my colleagues have published several papers on the relevant thermodynamic principles. I am not sure of their influence so I will try to explain the thermodynamics again.
Side comment: The first edition of Nutrition in Crisis was called The World Turned Upside Down (I made an analogy between the low carb revolution and the American Revolution). Unfortunately out of print, it included a cartoon by my daughter Robin Feinman illustrating the behavior of the critics who maintained that the data on low-carbohydrate diets are impossible because they would violate the laws of thermodynamics I felt it was like the cartoon characters who run out over a cliff and only fall when they suddenly realize that they are not on solid ground. Somehow the data are expected to go away once thermodynamics is invoked
While I was writing this, I came across a YouTube presentation and recent Substack by Mike Eades who took a different tack on things. He proposed, in agreement with a previous publication by Anssi Manninen and Francisco Arencibia-Albite, the idea of “mass balance” as a new paradigm. Anssi had sent me a MS a couple of years ago but it was several pages and mostly equations. I was disinclined to spend that much time and effort without knowing where they were going with this but never got a good summary.
Mike’s description was better but I am afraid it does not solve the problem. We do, however start from the same point emphasizing the idea that we don’t eat calories. We eat grams, that is, mass, not energy. The definition of physical units is part of the confusion. However the mass balance model does not solve the problem and is not quite correct.
Mike described the two competing theories of why we gain weight as CICO and the carbohydrate-insulin-model (CIM) which he explained as
“The Carbohydrate-Insulin Model posits that calories coming in are not just neutral energy, but instead provoke metabolic hormonal responses that, in turn, drive the urge to eat and store calories. The CIM proposes that people have varying degrees of carbohydrate tolerance, and that in those who are intolerant, carbs provoke a greater and more long-lasting insulin response. The continuously elevated insulin levels in those people end up driving insulin resistance and hyperinsulinemia. Since insulin is the metabolic hormone that both drives fat into the fat cells and keeps it there, those people who have too much insulin tend to increase their mass of fat tissue.”
This is surely part of the picture but CICO and CIM are not really alternatives because they are not about the same thing. CICO is a theory of energy balance. It is fundamentally wrong or at least inapplicable as a general theory. CIM, on the other hand, is a proposed mechanism which is separate from energy. Thermodynamics has no mechanisms. The laws of thermodynamics would be correct even if heat were a fluid, flowing from one body to another. In fact, historically, there was such a mythical substance, called the caloric, which was still invoked in the nineteenth century. I believe Carnot still believed in the caloric when he framed the second law of thermodynamics.
Also, while CIM is clearly part of the relevant mechanism, it is not a “model” except insofar as all scientific knowledge is a model. It is demonstrable that insulin is an anabolic hormone, the major effect is at least to keep fat in fat cells. You can show it in a test tube. The bottom line which I will elaborate here and in future posts is that energy metabolism is controlled by enzymes, that is, rates of reaction and, to a large extent, hormones control enzymes.
What went wrong.
“He conversed extremely well, if rather legalistically. I had fallen into his carefully laid traps so often that I had devised a standard method of getting out of them. It was not profound; I would simply ask him to define his terms. That annoyed him and set me free.”
— B.F. Skinner, Walden Two
When calories in are used in a CICO calculation, they are taken from the value derived from the calorimeter measurements. That’s where we get the values assigning 4 kcals to 1 gram of carbohydrate or protein and 9 kcals to 1 gram of fat. This is the basis for the fundamental error.
The energy in a chemical reaction is in the reaction, not in the reacting compounds. Energy is a process, not a thing. Mike brings this out in his presentation. The oxidation of glucose, for example, provides a good deal of energy but it comes from the difference in the chemical bonds, geometrical orientation (entropy) and electronic state between reactants and products. The total mass is still preserved. Einstein explained that if there is an energy change, this must involve some change in mass. Mike’s Substack and presentation points out, however, that this is infinitesimal and does not enter into what we are doing. (By comparison, in particle physics, where small numbers of particles and high energies are involved, mass is given in electron-volts which is a unit of energy).
So, when we make the association between energy and mass, we are using the results of a particular reaction, complete oxidation. In a calorimeter, food is placed in a small container that is immersed in a water bath. The device is maintained in an atmosphere of oxygen under pressure. The food is ignited and the heat generated changes the temperature of the water bath. From the change in temperature in the water bath, we can determine the energy of reaction. It all looks obvious but there are a couple of very important points:
The calories assigned to a food refers to a chemical reaction (not to the food):
Food + oxygen → carbon dioxide + water
Food + O2 → CO2 + H2O
(said: “food plus oxygen goes to CO‑two and water”).
If you ingest glucose and convert it all to carbon dioxide and water, if the physical conditions are approximately the conditions of the calorimeter, you will get 4 kcal of energy to do stuff with (physical movement, the chemical reactions, etc.). If you do anything else with the glucose, if you store the glucose as glycogen, or glycate your hemoglobin, or convert the glucose to fat, all bets are off. That’s the problem. That’s why CICO is not wrong so much as not measurable. Typically, diets are specified in calories. But…
What is the definition of a calorie?
When asked the question, most people respond with the definition of the physical calorie:
“The amount of energy required to warm one gram of air-free water from 14.5 to 15.5 °C at standard atmospheric pressure” or something similar.
That is not, however, the definition of the nutritional calorie (= physical kcal). The nutritional calorie as used in energy calculations is “the heat produced by complete oxidation of 1 gram of the designated food under standard conditions.”
When you try to do calories in, if you use the calorimeter values, that may simply be inappropriate. That’s why I emphasize that you don’t eat energy. You eat mass and your body carries out chemical reactions that produce energy.
Have you heard this before: “You are what you eat” is wrong. You are what your body does with what you eat.
Does mass balance model help?
The mass balance model recognizes that it is not energy that is consumed. To define the input, however, the method takes the specified calories and uses the reciprocal of the calorimeter value, that is, calories/gram and rather than comparing calories compares masses in different diets. Here’s a slide from Mike Eades’s YouTube. It is clear that if you design two diets based on total calories, the low-carb diet will actually provide less food by mass. The question is what does this mean? I’m afraid it doesn’t mean that the advantage to a low-carb diet is that you eat less. That’s what the CICO advocates say: low carb diets do better because you eat less. There is a fundamental error. Nothing is changed here. You can’t change the meaning of a parameter by taking the reciprocal. It means the same thing with a different perspective. Your car gets better mileage than mine but I get to use more gas.
Units of chemistry.
The reason this is not helpful is that metabolism, all chemistry, does not run on mechanics — amount of material. Mass is not fundamental in chemistry. Moles are. If you mix a pound of butter with a pound of egg yolks, you get two pounds of Hollandaise (questions of haute cuisine aside). If you react a pound of oxygen with a pound of hydrogen, you get 1 lb, 2 oz of water and 14 oz. of hydrogen left over. The ATP derivable from fat is in the energetics of the reaction. That follows from the moles and, not the mass. Fat is more efficient than carbohydrate in producing moles of ATP. If you process it in different ways you can convert the mass to more or less stored energy, That is the meaning of “we eat grams, not calories.” Again, the energy comes from the reaction.
Why it’s wrong. What the first law really says.
Both mass balance model and CICO take 4 kcal of carbohydrate and 1 gram of carbohydrate as essentially interchangeable. As above, this is not correct. But people say: aren’t calories conserved? Doesn’t the first law even everything out?
The bad news is that thermodynamics is a difficult subject. You can’t shoot from the hip. Everything has to be defined precisely. So, I am going to give a more precise description of thermodynamics. I will explain how differences in efficiency (weight gained per calorie) between different isocaloric diets can arise. It is not the easiest thing to understand. What follows involves heavy-duty physical ideas. Even physicists are modest about their understanding. The math is simple but you have to think about the ideas.
The first law of thermodynamics is, in fact, a conservation of energy law. If you add up all the changes in energy in the system and the environment, they all adds up to zero. (Only tiny errors due to E = mc2).. That’s not the whole story, however. That is a popular or rough description. Here is a more precise description (slightly modified) as laid out in Nutrition in Crisis:
The first law says precisely that there exists a parameter called the internal energy, usually written as U so as not to confuse it with the electrical potential. The change (Δ) in the internal energy of a system is equal to the heat, q, added to the system minus the work, w, that the system does on the environment.
ΔU = q ‑ w
In other words, we don’t measure energy. We calculate if from the observed forms of the energy, heat and work. We could specify other forms of energy but we can usually assign these to either heat or work. “Work” is sometimes roughly defined as any energy that is not heat. What the equation really says, however, is that energy changes include energy that is useful and energy that is wasted. Living systems largely carry out chemical work. Energy is conserved but not all energy is the same.
To go to the next step, you need to understand the idea of a state variable. A state variable
is a variable where any change is independent of path. For example, the familiar temperature T and pressure P are thermodynamic state variables. If you blow up a balloon, thermodynamically it doesn’t matter whether you change the pressure quickly or slowly. The effect on the system is controlled by the difference between the pressure after the change minus the temperature before the change, that is, ΔP. Same for ΔT. The usual analogy of a state variable is the as‑the‑crow‑flies geographical distance, say, between New York and San Francisco. This is a state variable; it doesn't matter whether you fly directly or go through Memphis and Salt Lake City like the flights that I wind up on.
Now, here’s the idea. The total energy change in any transformation is a state variable and is always conserved. So oxidation of glucose provides the same total energy whether it is catalyzed by a digestive enzyme or you do it in a test-tube.
The heat, q, however, and the work, w, are not state variables. That’s why we usually write them in lower case or with some other distinguishing marks. As a general rule, in the classic thermodynamic scenario, we are trying to do work and most of the heat that is generated is wasted. We consider the ratio of work/heat as the efficiency of the process. It all harks back to the industrial revolution which played a big part in the development of thermodynamics. Efficiency of machines was a big deal. So, energy is conserved but how the energy changes are distributed depends on conditions like — cutting to the chase — which macronutrients take part in the reaction. Efficiency depends on how you do it: the calorimeter reaction has near zero efficiency. Nothing is produced and all the heat is wasted. Human beings can get as much as 50 % efficiency in metabolizing their food. Energy is the same in both cases. You may notice that the first law contains the germ of the second law which tells us that, for any (real) reaction, we always have some inefficiency.
Conclusion: Calories-in Calories-out (CICO) as conceived as a mechanism of energy balance in nutritional studies is not correct. It is based on a limited understanding of the first law of thermodynamics and uses inaccurate definitions of calories. Chemical energy resides in the reaction not the reactants and products. It is a process not a thing. Thermodynamics more or less predicts that isocaloric diets of different compositions will have different efficiencies (ratio of useful energy to stored fat).
Calories in nutrition often refers to a specific value determined for complete oxidation as measured in the calorimeter. It is not directly applicable to other reactions.
Insofar as the outcomes of two diets have similar metabolic effects, it is a consequence of biological mechanisms — extensive feedback system — and is not predicted by any physical law. The Carbohydrate-Insulin-Model (CIM) is not in conflict with CICO. It refers to mechanism rather than energy metabolism. It is not really a model in that insulin is known to be anabolic and it is likely that CIM plays some role in the true mechanism.
References
michaeleades.substack.com/p/the-arrow-111
Comparison of low-carbohydrate and low-fat diets. https://phcuk.org/evidence/rcts/