Comment on Guyenet vs. Taubes; or Why I Don't Give a Crap What the Kitavans Eat

This post started as a comment to Stephan Guyenet's excellent post on the carbohydrate hypothesis of obesity, got too long, and so I'm putting it here. Do read Stephan's post, and keep an open mind. It's got loads of interesting and cutting-edge science, and this sort of debate and information exchange is how science progresses. If you find yourself experiencing cognitive dissonance, remember that absolute belief is antithetical to science. We always must update our beliefs as new information emerges.

Short summary of Stephan's blog post: the hypothesis that carbohydrates in general are fattening is probably over-simplified and does not reflect the most recent scientific understanding of metabolic regulation. It also leads one into a variety of paradoxes, a la the "French Paradox" of the diet-heart hypothesis.

I think part of what we're seeing here is the rather poor taxonomy of nutrition. We discuss things in terms of macronutrients, but those macronutrients come with (or without) all kinds of other metabolically relevant substances. And even within a given macronutrient group there can be significant metabolic differences, e.g. for fatty acids of different chain lengths, or between glucose and fructose (though Dr. Feinman might have something to say about the latter).

I've posted here before on my favorite example of this, and it seems like a good time to revisit (working from memory and about 4 hours of sleep, so please correct me if necessary). The Aztecs had a corn-based diet. They did experience obesity, but despite documenting a wide variety of health issues in detail never described diabetes. The Egyptians ate a wheat-based diet, also experienced obesity (along with heart disease, cancer, and the whole host of other fun "diseases of civilization"), and did document diabetes. Two high-carb diets, both resulting in some level of obesity, but from what we can tell (thousands of years later), both having radically different metabolic endpoints.

Two take-home points here. First is that we likely need to consider a broader dietary context than that imposed by our artificial macronutrient classification scheme, i.e. wheat and corn both provide primarily carbohydrates as energy, but probably do not have the same metabolic effects, particularly when considering over the timescale of a human life. Second, obesity is a symptom. A given symptom may result from multiple underlying conditions. We need to focus the discussion on more specific pathologies than just "obesity".

In the US and many other Westernized countries, one can take a look around and do a "liver check". How many people do you see with a protruding pot belly as opposed to a general body-wide distribution of fat? Most people I see have the big belly, sometimes even being very lean elsewhere (particularly in the arms and legs); there are a few "Rubinesque" figures as well, but the pot bellies seem to be running away with the obesity stakes. The big belly is indicative of fatty liver. Considering how central the liver is in metabolic regulation, it should come as no surprise that an inflamed fatty liver could lead to a whole host of metabolic disturbances: obesity, abnormal lipid profile, elevated blood sugar, elevated insulin, etc. In other words, metabolic syndrome.

I would argue that the rapidly growing health problem is not simply obesity, but metabolic syndrome (remember obesity is only one symptom, and there are thin people with metabolic syndrome too). We want to understand both how we arrive at metabolic syndrome (so our children can avoid it), and also how to treat it for those who did not avoid it. It is clear "carbohydrates" across the board are not causal in the development of metabolic syndrome. Stephan provides several counter-examples; another is the Tarahumara, who like the old-school Pima subsist largely on corn, beans, and squash, but who have one of the lowest rates of Type 2 diabetes in the world.

But the cure is not necessarily the reverse of the cause when it comes to disease. Metabolic syndrome brings a whole host of issues, not the least of which is broken carbohydrate metabolism. So while carbs in general may not lead to metabolic syndrome, once you've arrived dumping carbohydrates on your broken carbohydrate metabolism is tantamount to doing jumping jacks on two broken legs. I believe the science (along with a massive stack of anecdotal evidence) is pretty clear here, in that the most successful treatment for metabolic syndrome is carbohydrate restriction.

So while yes, Virginia, the Kitavans eat a very high-carbohydrate diet and exhibit general metabolic health, for my personal dietary choices I don't really give a crap. The Kitavans have healthy carbohydrate metabolisms, but I don't (prior to going low carb I had a trophy beer gut, which in retrospect was my liver telling me "You're killing me slowly"). If you look down toward your feet and can see only your protruding liver, you might consider trading in the bagels for bacon (better yet, get yourself a blood glucose meter and check your post-bagel blood sugar - it might frighten you). It is important to remember that carbohydrate restriction is successful as a treatment for a disease, but it doesn't necessarily follow that all carbs are bad for everybody. We have several examples of cultures who thrive on diets of lean protein and whole food sources of carbohydrate, like starchy tubers and fruit. We also have examples of cultures thriving largely on protein and fat. Humans appear to have a remarkable ability to survive as omnivores eating whole foods, which in no small part explains why we are one of the most widely spread species on the planet. So if you have a healthy metabolism, you probably can choose from a wide variety of whole foods (and by "whole food" I mean something you could plausibly obtain from Nature without the aid of much more technology than fire and a sharp stick). Once your metabolism is broken, you will likely need to make some choices to avoid those things which, due to your disease state, have become effectively toxic. In other words, make your nutritional choices based on actual knowledge of metabolism and your own state of health rather than picking a buzz-phrase and applying it blindly.

And for God's sake, stop eating wheat ;-)

On Taubes and Toilets

One of our toilets has been acting balky lately. Last night I went to flush it and nothing happened. I started pondering on the possible causes of this, and had a brief vision of a bunch of Ph.D's standing around, stroking their chins and sagely examining the toilet through glasses perched on the ends of their noses. After a few knowing glances at each other, they pronounced: "From the First Law of Thermodynamics, we know the problem with your toilet is that, at some point in the past, less water came in than left!"

Maybe I should have skipped that last martini at dinner last night.

Anyway, my imaginary colleagues were only acting as scientists often do, pronouncing "truth" without getting to the root cause of the issue. Or perhaps my subconscious has been imprinted from too many conversations like this with my children:

Me: "How did you get so dirty?"
Child: "I was playing in dirt."

Back to the toilet. My imaginary scientist friends, while technically correct, were (as scientists often are) totally unhelpful. If I were to fix my toilet, I would need to know how it works, particularly the possible failure modes. In other words, I need to get to the root cause of why it didn't flush. Once I know what's actually broken, I can fix it. Invoking the First Law of Thermodynamics might make one sound smart, but it doesn't get my toilet flushing again. And believe me, in this case it was a vital importance to identify and repair the root cause, posthaste.

The toilet is actually an example of a self-regulating system, by which I mean that when it works correctly, I don't have to pay attention to it. If you're not familiar with the workings of the common toilet, check out this entry at HowStuffWorks.com. Basically, there's some clever gadgets in there that make sure things go smoothly. When you push the handle down, it pulls up the flapper, basically a rubber stopper with a hole in the bottom. There's air inside the stopper, which causes it to float open as long as the water level is above the stopper. Once the water is gone, the stopper closes. That's the output side. The input side is controlled by a float-activated valve. When the water level falls, so does the float, which opens the valve and lets water into the tank. As the tank fills, the float rises until it hits the switch and shuts the valve, turning off the water. This whole setup is basically tuned to ensure that you have enough water leaving the tank at the proper rate to get a good flush, while not having too much water enter the tank and thus flooding your bathroom. The toilet has an additional fail-safe to avoid the latter fate, in the form of an overflow pipe. If your float switch doesn't work, then the water goes down the overflow pipe instead of all over your floor.

My particular problem was too little water, not too much. Since I have confidence in the First Law of Thermodynamics (at least in approximately flat regions of spacetime, like my bathroom), I know that something caused the tank to not fill. The water didn't fill the tank at some point then magically vanish. One possibility was that my water main had broken. Checked the sink faucet, plenty of water there. Rather than stand around and be mystified by the inner workings of my toilet, I opened it up and took a look inside. No water alright, and it looked like the float was stuck. A quick poke and the toilet started filling. The moral of the story is that the laws of physics don't tell you how things work, but rather the constraints under which they work (e.g. the amount of water leaving the toilet in a flush is the same as the amount that entered when it filled). To solve my problem I needed to understand the mechanism by which the toilet regulated water flow and level, and how that regulation could go wrong. In other words, if you know what causes the toilet to work correctly, then you can infer what might cause it work incorrectly, and take appropriate action.

If you've read Gary Taubes most recent work, Why We Get Fat, you probably have realized by now that my toilet story is a bit of a setup. Why We Get Fat (WWGF) is generally described as "Good Calories, Bad Calories" lite, but it is a bit more than that. Taubes has focused on obesity in particular, and honed his arguments and presentation, and brought in some more recent research as well. It's an excellent and fast-paced read, and I highly recommend it.

The key hypothesis of WWGF in terms of obesity is the same as GCBC: that obesity is the result of a failure in the regulation of metabolism, specifically carbohydrate metabolism. A broad set of critics attack both GCBC and WWGF on various detailed points, while missing the big picture. For instance, there's much ado about the specifics of Taubes' hypothesis on how this failure in carbohydrate metabolism arises. Taubes posits that overconsumption of carbohydrates basically leads to insulin resistance (though notes that the situation may not be so simple), while others point to various evidence that it may be specific carbohydrates (fructose), or vegetable oil, lack of physical activity, etc. These make for nice academic discussions, but if you're one of the millions of people with a broken metabolism, none of this is very helpful. Much as was the case with me and my toilet, if you're going to fix your metabolic machine, you need to have some idea of how it works and what might be go wrong. WWGF is a great place to start educating yourself, provided you don't fall in the trap of running around screaming "I can't see the forest because I've got blood in my eyes from running into all of these damned trees!"

The publication of WWGF has also revived the strident preaching from the members of the Holy Church of the First Law of Thermodynamics. Now, I'll give you a pass if you read GCBC and perhaps came away thinking that Taubes implied that low carbohydrate diets somehow got around energy conservation. GCBC was a dense book, and Taubes (who was a degree in physics) no doubt thought that the First Law was just generally held to be true and that nobody would question his belief in it, and so didn't focus on it much. Taubes clearly learned the hard way that you can't take these things for granted. WWGF has two chapters on this topic, and makes it very clear that 1) Yes, Virginia, the First Law of Thermodynamics is alive and kicking, but 2) that the First Law adds no information as to the cause of obesity, or what you might do to fix it. If you read WWGF and still think Taubes is claiming that thermodynamics doesn't apply to biological organisms, then you either didn't really read the book, weren't paying any attention, or have the logical facilities of a monkey on crack.

The real lesson of WWGF is the same as my toilet story: just knowing the constraints on the workings of your body (e.g. conservation of energy) is not the same as knowing how the pieces actually fit together, the cause-effect relationships that make the whole machine go. You can't fix something without having some idea of how it works, whether it is a toilet or the human body. Like my toilet, your metabolism (and that of all living organisms) is self-regulating. Humans seem to be control-freaks in general, and we think that every aspect of life needs constant attention, much like driving a car (I wish people paid as much attention to driving as they do to other less consequential things, like whether or not their children poop enough times a day). But when my toilet works right, I don't have to sit in the bathroom and monitor it, waiting to shut off the water if there's an overflow, or fiddling with the float valve switch. It just does it's thing. Metabolism is the same way. Energy regulation is the key aspect of life, from bacteria to humans, and most life doesn't have the capacity to fret about how many calories it ate or how much it exercised. If your metabolism is operating correctly, by definition it is impossible to eat too much. When you have too much energy stored, the body has ways of eliciting biochemistry and behavior (which is just complicated biochemistry) to bring things back into balance: appetite is decreased, thermogenesis is increased, you have the urge to move around, etc.

If you're obese, you don't have a character or mental defect any more than my balky toilet does. You have a physical problem in metabolic regulation. Invoking the First Law of Thermodynamics and berating obese people as having a behavioral issue does not address the root cause of obesity, any more than a similar approach would have worked in fixing my toilet (I'm having visions of registered dieticians bitching out my commode for lack of self-control). WWGF is a great place to start "opening the box" and empowering yourself to start giving the "experts" the finger, stop feeling like a failure because your experience doesn't agree with their beliefs, and get down to actually solving the problem.