“A calorie is a calorie” violates the second law of thermodynamics (2004)

The surrounding discussion is an interesting hypothetical. The problem occurs when people use unsettled science like "not all calories are created equal" to change their lifestyle (e.g. "all X diet"), then fail at making substantial change because they made it too complicated.

You can do very well believing "a calorie is a calorie" even if it is technically inaccurate, and you can do very badly believing "not all calories are created equal" even if it is technically true on some level.

Effective weight loss is simple (note I didn't say "easy"):

- Calculate your TDEE. TDEE is ((RMR + TEF + NEAT) * Physical Activity Level).

- Eat fewer calories than your TDEE in order to lose body fat or eat more to gain body fat.

- Most of the population are Sedentary or Moderately active. Many also overestimate their PAL by one level (i.e. you aren't "Vigorously" because you hit the gym twice this week for 20 minutes, you're Moderately).

- Exercise to improve overall health (mental, sleep, immune system, etc). Don't "double count" exercise calories, just set your PAL level and that's it.

- Don't eat less than 1500 / 1300 calories for men / women respectively and try to get some nutrients.

Ignore everything else, it is just adding complexity. Weight loss should be "Simple but hard." "Complex but easy" doesn't exist unfortunately, at least not in the medium to long term. Instead, you'll just burn out faster. So K.I.S.S., make it boring, make it a habit, eat whatever you want below your TDEE.

PS - Starting new exercise (or periods in women) can cause short term water weight increases in the order of 1-6 lbs. This is caused by inflammation. It is completely normal and will go away on its own. Just be patient.

Most of the comments so far are missing the point. The paper points out that the energy available from food, for doing work, creating fat, or any other process, is not, and can not be, equal to the energy content of the food: it is reduced by thermogenesis. Because the reaction pathways are different for different types of food, a calorie of protein does not provide the same usable amount of energy as a calorie of lipid, for example. A greater portion of the protein calorie goes to waste heat.

However, from their Figure 2, you can see that the effect of even large changes in dietary composition have a small effect on the energy available from food. You can drastically reduce your carbohydrate consumption, while maintaining your 2000 calorie intake, and that amounts to only about a 100 calorie reduction in available energy.

The conclusion: to a reasonable degree of accuracy it really is CICO. To lose weight, eat fewer calories.

Absolutely hyperbolic title. Sure, "A calorie cannot exceed a calorie" is more correct, but it's less punchy so nobody uses it. The point is that you can't get free energy from nowhere. It's very well illustrated by DNP: https://en.wikipedia.org/wiki/2,4-Dinitrophenol

I recently started tracking my daily calorie intake with Noom, and I was very surprised at how easy it is to get to 2000 calories if the diet doesn't consist of low calorie dense foods. I'm now focusing more on eating foods that are have a higher water content and foods that have a good fiber content. If I am still hungry after eating those foods, then I go for something else. I also don't refrain from sweets of simple carbs, but I do try to make mindful decisions about how often to consume them.

The problem isn't metabolic or thermodynamic or anything like that.

The problem is measurement.

We cannot actually measure calories in or out in a direct way. If I eat an apple, I know that X grams of apple will be about Y calories. Usually. For typical apples. I cannot actually measure the apple I ate. It's a destructive process. Both the measuring and the eating.

Similarly, I cannot measure the caloric expenditure of everything I expel. Either through breath, sweat, heat, waste, etc. It's difficult. We have approximations.

And it's hard to get that information because we're always breathing, drinking, eating, sweating, urinating, defecating, etc. So if were to weigh myself right now, fast for the entire day, then come back in 24 hours and weigh myself again, I'd still only have an approximation of how many calories I expended in that time.

When we talk about how things are "processed" what we are really doing is talking about whether or not they are processed. A calorie that gets expelled through waste is not really a calorie consumed. It just passed through you.

Look back to Olestra. It was marketed as a sort of 0 calorie oil because it wasn't processed by the body. It would pass right through you. Which kind of was the problem with it.

But there are other ways to cause your body to reject calories ingested. It's the mechanism behind bulimia. You can cause yourself to vomit or take laxatives to cause yourself to defecate. Whereas anorexia modifies the CI part of things.

End of the day, you find out roughly how many calories you burn a day and then eat roughly that many calories a day, you'll hover in your current weight range. Want to lose weight? Eat less than that. Gain? Eat more than that.

Did a double take when I read the authors. https://en.wikipedia.org/wiki/Richard_D._Feinman

The debate around "A calorie is a calorie" relates to the whole "calories in vs calories out" concept for weight loss/gain.

Examples like "500 calories of twinkies and 500 calories of lettuce are equivalent" focus heavily on the "calories in", but forget to consider the "calories out" aspect. What I mean there is, when someone says that the body processes different macronutriets differently, a lot of that has to do with how it affects "calories out"

If you eat 500 calories of a food that your body will only process a small amount of, and expel the other 450 calories as waste, that 450 is part of the "calories out". That's only a net "calories in" of 50.

On the other hand, if you eat 500 calories of a food that your body chooses to process/store most of, that's 400 calories in, and only 100 calories out.

To clarify, I'm not talking as much about absorption, more-so about things like how a macronutrient affects insulin sensitivity.

Carbohydrates decrease insulin sensitivity more than protein and unsaturated fats would. Decreased insulin sensitivity means increased insulin resistance.

"Having insulin resistance means that the walls of the cells within the body that need to take in glucose for energy, have a much-lowered sensitivity to insulin. This makes it harder for the cells to receive the necessary glucose through the cell wall.

That glucose then remains in the bloodstream, resulting in high blood sugar levels, which the liver then has to manage. When this excess sugar reaches the liver, it is converted to fat. The fat cells leave the liver via the bloodstream and cause weight gain. If left unmanaged this can lead to obesity."

The calories out part of the equation isn't just the energy you expend per day through movement/metabolic processes/etc, it's also the calories that are expelled as waste because the body didn't store that macronutrient.

> If a "calorie is a calorie" were true, nobody would pay extra for high test gasoline.

The total output of an engine has to come from the stoichiometry of the fuel, regardless of its anti-knock properties. The total output is the work obtained plus heat, plus whatever energy remains in any incomplete combustion products that are exhausted. Inefficiency caused by knocking means we get less work, but that has to be accompanied by more heat. If we are only measuring heat, or only work, we are doing it wrong. One way is to simply reduce one to the other: make the engine do some work against friction which wastes all of it as heat, and put that whole thing in a closed system (the calorimeter). It cannot be that you get more or less heat out of it if you burn the same amount of gasoline, and obtain the same exhaust products.

The conservation of energy principle is inviolable. If we capture and measure all of the inputs and outputs of an organism (heat, work and solid/liquid/gaseous matter), then any discrepancy between the energy content of two means it is losing energy from its energy stores, or else accumulating it.

The main fallacies in fat control are not violations of thermodynamic principles per se but rather these:

(1) it's easy to simultaneously control the work/heat output and/or the input to obtain an arbitrary desired caloric deficit.

(2) a caloric deficit will always be covered by releasing and burning fat in preference to anything else.

A cup of gasoline has ~2,000 calories. Assuming you don't die from drinking it, this trivially and obviously would not have the same impact as eating 2 McDonald's McDoubles.

But it does have the same calories. So, it turns out, calories are kinda useless if your body can't break down what they're contained in. You'd poop it out, calories intact (or breathe it out, pee it out, or in the case of gasoline, probably just die).

Even pulling back into things which are wholly edible: ever heard of the meat sweats? They can occur because your body has to expend a lot of energy in breaking apart proteins commonly found in meats; more-so than while digesting other foods. In other words, if you eat 400 calories of meat, but it takes an investment of 30 calories to digest it, how does that compare to 400 calories of tomatoes?

What about lactose intolerance? The inability to digest, and gain nutrients, from milk products. One person drinks 200 calories of milk and gets lots of nutrients; another shits their pants and is now dehydrated.

It seems trivially obvious to me that anyone who says "a calorie is a calorie" is trying to sell you something. Though, I suppose, the people who say "my diet will work for you, buy my special food plan" are saying the opposite, but are also trying to sell you something.

Sure, the body loses some energy processing calories (TEF, thermic effect of food). But it's very efficient.

The larger flaw behind "CICO" reasoning is that it often ignores volume, hunger, and mood effects. 1200 calories of spinach will make you gain about as much as 1200 calories of sugar. But the former is basically impossible to eat: spinach comes to 23 calories per 100g, so you would have to eat ~5.017kg, or ~11lbs, to eat 1200 calories (sugar is 0.3kg or 0.6lb). When people say things like "500 calories of spinach = 500 calories of chocolate", well technically they're correct, but it's unfair because nobody is eating 500 calories of spinach.

And yes, spinach is an extreme example. You can gain weight on healthy "low calorie" foods like chicken and apples. But you'll be stuffed. Most people, when trying to lose weight on processed foods, will constantly battle hunger and low energy. But when they eat unprocessed foods, they have much less hunger and feel better. Even when people say "I'm eating healthy but not losing weight", usually most of the foods they're eating are still high calorie (e.g. cheese and nuts).

Giles Yeo is a great source of popular information along the lines of this topic. I recommend his books, but here is a recent youtube from his talk at The Royal Institution https://www.youtube.com/watch?v=GQJ0Z0DRumg

the calorie is a very bad metric for human energy consumption and nutrition in general. First, it completely ignores other nutrients like minerals, vitamins, prebiotic fibers, etc.

Second, to measure the caloric value of something, you'll need to burn it and calculate the amount of heat generated. This is not how the human body digests food. Many types of substances can't be digested (like fibers) but generate a substantial amount of heat.

Bits and pieces of nuts and seeds can go undigested through to a large extent, depends on the size of the pieces that were swallowed, and so the body absorbs only parts of the fats, carbs and proteins in those types of foods (although nuts are very calorie dense)

We really need a new measure of energy that's suited for nutrition... The calorie is worthless.

Or more precisely:

"The particular interpretation of the adage 'a calorie is a calorie' that we have chosen here violates the second law of thermodynamics."

If, as shown in their energy diagram, all food converts to CO2 and H2O, then no foods can cause weight gain because all the carbon will be exhaled and water excreted.

I read this as an indication that numbers out of a calorimeter don't correspond precisely to the bio-available energy.

TL;DR: “ the first law holds even in irreversible processes – energy is still conserved – the second law says that something is lost, something is unrecoverable (..) The efficiency of a machine is dependent on how the machine works and, for a biochemical machine, the nature of the fuel and the processes enlisted by the organism. (..) In weight loss diets, of course, inefficiency is desirable and is tied to hormonal levels and enzyme activities. (..) The second law of thermodynamics says that variation of efficiency for different metabolic pathways is to be expected. Thus, ironically the dictum that a "calorie is a calorie" violates the second law of thermodynamics, as a matter of principle.”

About the above text: Tried to cut parts from the article to abstract the main points.

The purpose of "a calorie is a calorie" is not to be scientific truth; it is to serve as guidance for people who are trying to lose weight and are falling prey to particular mistakes.

It's like if you're at the gym and a coach is telling someone "sit your butt back, butt back, butt back!" "Sit your butt back" isn't a universal directive for that exercise, much less everything you need to know about it. It's a correction for what that person is currently doing wrong. Six months from now, they might be sitting too far back, and the coach will say, "Don't let your butt go back!"

Or the advice to, "Loosen up; people like it when you just say what's on your mind," versus, "You need to activate the filter between your brain and your mouth." These statements can both be true when directed at the right person in the right context.

The paper takes aim at "a calorie is a calorie" as an objective scientific truth, which makes it trivially wrong, but doesn't address its pragmatic truth as advice. "A calorie is a calorie" is "true" when it is a good corrective for the person it's directed at. It's "false" if the person needs to be paying more attention to the differences, like if you're an athlete trying to fine-tune your diet for performance, or if you somehow never got the message that people need vegetables in their diet.

It might be less true on HN, but in my experience with my family and acquaintances, "a calorie is a calorie" is a widely needed truth. Most people's approach to healthy eating is to sort foods into "good" and "bad" and try to minimize the harm from the "bad" foods, ignoring any potential for harm from the "good" foods. Some people give themselves a pass on the calories from olive oil and avocados because it's a "healthy fat." Other people focus on their intake of cheap processed corporate foods while not worrying about the generous quantities of expensive, local, artisanal bread and cheese and fancy meats they consume. There are even people in my generation who missed the whole last thirty years of carb controversy and mound up huge portions of spaghetti and garlic bread next to a 4oz cut of steak and call that a healthy meal because the only problematic food on the plate is in a tiny portion.

For whatever reason, whether it's because of how nutritional information is presented to the public by experts or because of exploitative marketing from the diet industry or because our brains are wired to see good and bad everywhere, that's how an unfortunately very large number of people approach eating for weight loss, and "a calorie is a calorie" is a good corrective for them, to remind them to pay attention to the quantity of everything they eat, not just the "bad" stuff.

I guess this is like saying "A Joule is a Joule" regarding different types of energy (gravitational potential, nuclear, etc). It may be true in an abstract sense, but the conversion efficiency makes a large impact in a practical sense.