Gary Taubes and the First Law of Thermodynamics

It has been my observation that many people don't understand the application of the First Law of Thermodynamics (FLT) especially how it applies to the discussion of diet and nutrition, and the biological system that is a human being, or any other animal.

Gary Taubes is a science journalist who has stirred controversy in his writings about the FLT. We hope to explain what is going on here.

The FLT states that the change of energy in a system is the difference between energy going in and energy going out. This seems rather obvious. If we say that a person has gained weight, the FLT says that she has taken in more energy than she expended.

Let's label the statement "person has gained weight" as "A", and the statement "person took in more energy than was expended" as "B". The FLT says only that A is equivalent to B. They are the same. A means B, and B means A. The FLT allows us to substitute one for the other, because they are equivalent.

Here's an example of how the FLT applies, wherein Gary understands, but Don does not:

Gary:  "I got fatter over the holidays"
Don:  "Well, you ate more calories than you burned."
Gary:  "Duh! I just said that!"

Now let's look at how to misapply the FLT. If I say A, and you say that's because B, we get the following:

Gary: "I have gained weight."
Don: "That's because you took in more calories than you expended."

Aha! Don is now espousing the usual energy-balance hypothesis of weight gain. But the FLT says that we can substitute in the above. The same argument stated differently (yet equivalently) becomes:

Gary: "I have gained weight."
Don: "That's because you have gained weight."

This above illustrates why any invocation of the FLT in this context is a tautology; it's just a restatement of the proposition. It doesn't explain anything -- it asserts no causation. The FLT does not say A causes B, or B causes A, it simply says that these two statements are equivalent: A means B. When one asserts such a causation, now that is no longer the FLT, it is a hypothesis. Correlation (in this case, exact correlation) is not necessarily causation!

It may well be true that overeating causes weight gain, but there is another hypothesis available to us, what Gary Taubes has called "the alternative hypothesis": weight gain causes overeating. This reverses the causality of the standard hypothesis, and we have not violated the FLT either way (in fact, it may be the case that one, both, or neither hypothesis is true; that is why we need more rigorous studies.)

One of Taubes' often repeated examples of the alternative hypothesis is explaining growth in children. Children as they grow take in more calories than they burn and gain weight, but it's not caused by overeating. It's caused by growth hormones pushing raw materials into the body to construct tissues, which causes great hunger and provokes compensatory eating in order to maintain baseline metabolism. Growth drives hunger and eating, it's not the other way around.

Now, since we've seen the energy balance hypothesis many times explained with an analogy to money in the bank, let's consider the following. Suppose a very financially naive person has an investment that pays her $100 dollars a month in interest. She can do arithmetic, however, and budgets very carefully, spending that $100 dollars a month, so the account balance regularly floats back to zero. Now suppose that luckily, interest rates go up. Now she receives $105 dollars per month, but still budgets for $100. Slowly her bank account fattens up. At her yearly review, she asks her financial advisor why she has gained wealth. The advisor says, "Well, that's because [$60] more went into your account than came out!"  Exasperated, she says, "Duh! I just said that!" and fires him. A different advisor counsels, "All you have to do is spend more money, and the balance will go down to zero again." He doesn't get the job either, but finally a third explains how financial interest works and the question is answered satisfactorily.

If we want to find an explanation for weight gain, we must look elsewhere than the First Law of Thermodynamics for an answer.

Eat like Conan!

Bodybuilders are some of the leanest people around.  At world class levels competitors will certainly have less than 10% body fat.  So in addition to training, what kind of diet should help promote this level of leanness?  Let's look at what Arnold Schwarzenegger advised in his 1984 book, Arnold's Bodybuilding for Men.  I recently found a photocopy of p.197 in one of my old notebooks, which I have not looked at for 25 years. At the time of publication Arnold was just past his prime and had been arguably the best in the world during the 1970's.  Here's what he had to say about diet:

"... [some bodybuilders] eat diets consisting of 50 to 70% protein, something I believe to be totally unnecessary.

"It is hard for me to convince them that what they ought to be eating is a basic, balanced diet, just like the one they were taught about back in health education class in school. I know they want something more exotic, but I can't help the way things are. That kind of balanced diet is necessary to provide the body with all of the essential nutrients it requires for the difficult and demanding training that bodybuilding involves.

"Here is my formula for basic good eating:

1. Eat about 1 gram of protein for every 2 pounds of body weight.
2. Eat no less than 60 and no more than 100 grams of carbohydrate per day.
3. Limit your fat intake.
4. Take a limited amount of vitamin and mineral supplementation just for insurance.
5. If you want to gain or lose weight, vary your caloric intake -- and that variation should be mostly in carbohydrates, in the form of vegetables, potatoes and fruit.

"Earlier in my career, I believed that a bodybuilder needed to eat as much as 200 grams of protein a day in order to develop the maximum muscle mass. Since then, my research has shown me that body- builders do need more protein than the average per- son, but probably no more than around 100 grams, and certainly no more than 150. This gives enough protein for muscle-building, without adding any unnecessary calories to the diet. Non-bodybuilders, on the other hand, can easily get by on no more than 1 gram of protein for every kilo (2.2 pounds) of body weight."

Now, before looking closely at these recommendations, note the comment about school health education.  Arnold would have been learning about this stuff in late grade school or junior high, but that would have been in Austria around 1960, most likely using textbooks published in the 1950's.  The advice in those books I think would not likely have borne any resemblance to today's government food guides, but perhaps some industrious and resourceful reader will find out for us.

Now to the 5-point list of recommendations.  The first thing to notice is that Arnold's advice on apportioning the three macronutrients (protein, carbohydrate, and fat) is quite specific on protein and carbohydrate, but extremely vague about fat.  What does "limit" mean?  Without a number this is meaningless, but we can get a reasonable estimate of how much this should be with some calculation (my interpretation of this point is that we shouldn't specifically add fats and oils to what is already in our foods.  For example, trim your steak, enjoy the internal fat, but don't slather it with butter;  don't drench your salad with oil; avoid fried foods).

So if I were to follow Arnold's advice, as a non-bodybuilder weighing  73 kg (160 lbs), I should get by on 70 grams of protein and 100 grams of carbohydrate.  Considering a reasonably active lifestyle, I would need at least 2000 calories a day.   We know that protein and carbohydrate contain 4 calories per gram, and fat has 9 calories per gram.  So by Arnold's recommendation I will consume 70 * 4 = 280 protein calories and 100 * 4 = 400 carbohydrate calories.  In order to get my 2000, I will need 1320 more calories from fat, or 66% of my intake!  (The other proportions work out to 20% carbs, and 14% protein).  He may not explicitly have said it, but he is recommending what would certainly today be called a high-fat diet!

In addition, point 5 is clear that carbohydrates will be the determining factor in how lean you get.  It remains to be shown if this diet is healthy in the long run and if it would get results for most people.  Considering the number of testimonials out there, and personal experience, I would definitely recommend eating like Conan!

Don't feed your inner worm!

One of the most profound insights I got from reading Good Calories, Bad Calories by Gary Taubes concerns the interpretation of cause and effect when considering caloric balance and weight gain. It is a difficult concept to grasp, and I had to think about it a long time before I understood it. I've participated in various comment sections in different blogs and have found that I have not done a good job explaining it, so rather than cluttering the comments in them, I'll expound on it here.

In a nutshell, the idea is this: if one gains weight, then by definition one has taken in more calories than one has expended -- one has "overeaten". That's what thermodynamics tells us. What it doesn't tell us is why one has overeaten. Most people believe -- and it seems obvious by observation -- that overeating causes weight gain. But here is the other possibility: weight gain causes overeating. This might seem silly and perplexing at first, but I propose a thought experiment below to explain the profound difference.

Let's imagine I have been affected with an intestinal parasite, like a tapeworm. This worm lives for itself, and steals some of the food I eat. It is capable of growing to an enormous size, capable of weighing as much as its host, and it has food preferences.

Now suppose this parasite steals ten percent of all the calories I ingest. Over time, I find that I am less satisfied with my portions, because the amounts I am used to eating now leave me hungry. So I have an extra snack or meal. If I had been living happily on an average of 2,000 calories a day, I find that I must now eat 2,200 calories, because the tapeworm has eaten and stored 200 of them to foster its own growth. After 150 days of this, my tapeworm now weighs a pound. The extra 200 calories a day is hardly noticed by me, because it is so easily obtained, and after ten years of living with my tapeworm, it has grown to weigh 20 lbs.

Now poor me, ten years older and 20 lbs. heavier, wonders how I gained this weight, in spite of watching my portions and exercising regularly! "Eat less!" "Exercise more!" say all of my friends, doctor, and the conventional wisdom.

So I cut my portions for a while, and guess what? I lose weight! Yes, my tapeworm slimmed down a bit, as have I, but now I am trying to live on reduced calories, and my worm still takes its (now reduced) tithe of my meals. My metabolism and cells are screaming for fuel by making me hungry. I feel tired and have no will to exercise. I still need my daily 2,000 to function well.

Eventually hunger gives in, and over time I must return to eating 2,200 calories a day, and my worm gets bigger than ever. And again I need even more energy to carry around the extra weight.

Now, knowing that my weight gain is due completely to the pirated and stored calories in my parasitic worm, and inaccessible to my metabolism, would any reasonable person blame my weight gain on my "overeating"? Is my overeating causing my weight gain?

No, the root cause is that the fuel I consume is not all getting to my metabolic engine. It is the weight gain of my parasite that is driving me to overeat. And if I don't overeat, I will become sedentary, or lose lean mass as my body attempts to compensate for the semi-starvation imposed upon it.

Now let's imagine two other conditions: my tapeworm loves, just loves sugars and starches! It will actually eat not 10% but 75% of the sugars I ingest, but it's not so fond of protein and fat. Furthermore, it is triggered to start eating by sensing high insulin levels in its host. Now what kind of diet would be the worst I could eat? What kind of diet would go a long way to fattening my worm, and leave me with the least share of the calories I eat? What kind of diet would ensure that I became really, really hungry not long after eating? I think you know the answer.

If you can now transfer these parasitic qualities to fat tissue itself, you begin to see the idea. Imagine that Metabolic Syndrome is an imbalance in the regulation of the fat tissue itself, causing it to hold on to stored calories more than release them. The end result is the same. The fat tissue has become like a parasite, robbing the rest of the body of the fuel it needs, and the body responds the way it always does, signaling the host to eat by inducing hunger. If that hunger is not met, it will slow down metabolism to compensate for the lack of energy. Fat storage is causing overeating, not the reverse!

If this is indeed the cause, or one possible cause of obesity, it explains several problems that the caloric-balance theory does not. For instance, why caloric restriction and exercise are so ineffective as a long term solution: it does not correct the underlying imbalance.

As far as I know, there has not been enough research done that can actually pin down the cause of the metabolic syndrome. Once you get it though, it may be incurable, and if I go out on a limb here, the best treatment may be a low carbohydrate diet.

I hope the above has made the understanding of the idea of causality in the conservation of energy a little easier -- and thank goodness we know how to treat tapeworm infections!