I haven’t written about obesity in a while but this from Tyler Cowen prompts me

Dr. Sturm found no relationship between what type of food students said they ate, what they weighed, and the type of food within a mile and a half of their homes.

At the same time Scott Sumner provides with a rhetorical tool that might help me convey what is otherwise extremely difficult to convey. That is:

Never reason from a food choice

Just like price and markets, food choice seems like the most obvious aspect influencing nutritional outcomes. Its right there in front of us. However, once you realize you are dealing with an equilibrium system the most obvious candidate frequently becomes the least powerful one.

To start with the choice of what to eat and when is obviously extremely basic. All animals do it for sure. Maybe we can get into some debate over whether bacteria and protozoa do it, but we know for sure that all animals make choices about what to eat.

However, we don’t think that most animals actually make the same type of deliberate choices that we think of humans making. Some of them have such rudimentary brains that it seem strange to talk about the choosing at all or indeed being “aware” of what they have eaten. They live in the moment and operate on nearly pure instinct.

Yet, these animals balance their caloric intact. Together with reproduction that is the main thing that many animals do.

So, we know that there must be some mechanism that matches caloric intake and output. Where these two forces intersect the drive to eat is the result.

Now you observe an oddity. The body mass of human beings – and seemingly other animals but that is another story – is rising. Why?

Apparently a lot of people run to food choice and say, well people must be choosing the “wrong” foods. However, this is bad question. We are dealing with an equilibrium system. What exogenous force has caused food choice to settle at a different level.

We know food choice hasn’t simply gone off the rails because people do not increase body mass without bound. This is technically possible but almost never happens. Instead people increase a given amount of body mass and reach a new equilibrium.

This suggests something has happened to the feedback mechanism. But, what?

Maybe we say advertising or something else. But, then you study the mechanism closely and it is darn robust. For example, just to show extremes you can forcibly run a human being up to expending 8000 calories a day without worrying that the human will even slowly starve to death.

That means that it can take an at least 4 fold shock to its system and not miss its equilibrium values hardly at all. The misalignment associated with obesity is quite small compared to that.

This should make us think that its not a simple shock. Something more subtle has happened to the feedback mechanism itself.

In the modern world no one is going to let us cut open human beings willy-nilly and screw around with their endocrine and nervous system. If we could then we could solve this mystery in no time flat.

Given ethical limitations we are forced to make inferences on data that is far removed from the systems of interest. Nonetheless, we still know such systems exist and we should suspect that they have something to do with hormones and other peptides.

We know that when we go screwing around with people’s hormones their basic desires change. Food is a basic desire. We also know that body composition and food intake change hormones.

Thus our baseline should be that something weird is happening to the hormones.

And, that is really my message. Hormones ought to be the baseline here. Its of course possible that some other theory is correct but this should undoubtedly be the one we run to first.

As it turns out there is increasing evidence for the hormone theory and we think we may even know some of the ones that are involved ghrelin, peptide YY, etc. That’s great. But, its not the empirical evidence that should be swaying us here. Its is extremely preliminary.

What should be swaying us is basic reasoning about how dynamical systems operate.