[glenra]

One particularly interesting point Taubes makes is that there exist people (with various conditions) who are simultaneously emaciated and obese. Someone can be obese in the lower half of the body and emaciated in the top half or vice-versa. In this circumstance, your theory suggests they should simultaneously eat more AND eat less to fix it.

> If energy or matter is being destroyed by viruses

The viruses don't need to destroy or create energy, they just need modify the set point. (Which could be done by increasing hunger, decreasing satiety, or decreasing metabolism.) "net caloric balance" is true by definition yet is utterly useless as an explanation of why people gain or lose fat or have trouble doing so. What needs to be explained is why for some people their metabolic rate and hunger demands make them stable at a LOW set point and for others their metabolic rate and hunger demands give them a HIGH and/or INCREASING set point. There are lots and lots of environmental factors that COULD explain some of this. For instance, obesity in the US has increased as smoking has decreased; people who quit smoking tend to gain weight.

[jacques_chester]

> Someone can be obese in the lower half of the body and emaciated in the top half or vice-versa. In this circumstance, your theory suggests they should simultaneously eat more AND eat less to fix it.

I've met these people myself. They call themselves "women".

> The viruses don't need to destroy or create energy, they just need modify the set point.

"Set points" assumes that we are dealing with an automatic system. This is true only if people do not pay attention to what they are eating.

Given that the first and indeed only step in dieting is to observe and control your eating ...

[glenra]

> "Set points" assumes that we are dealing with an automatic system.

We are dealing with an automatic system. If we weren't, people would starve or become morbidly obese whenever they moved or changed jobs such that they were walking a different amount. Or when the restaurant down the street closed and they switched to a different one. Most people's weight stays remarkably consistent over the years despite significant changes in calorie consumption and energy expenditure. The reason that happens is (a) your resting metabolism speeds up if you eat more or slows down if you eat less, (b) your subjective sense of "being hungry" - your inclination to eat - adapts to your metabolic needs, (c) your inclination to be active adapts too.

People who are obese tend to be in the situation that their automatic system wants to keep them that way. Whether they "pay attention" or not, fighting that system is HARD. It's hard enough that we don't have reliable advice on how to do it. We have crappy advice that mostly doesn't work. People can lose some weight for a limited period of time, but they generally can't lose all the weight they want to (ever) or keep off the weight they do lose for as long as three years.

Most weight-loss studies don't follow people for long enough - 6-month studies tend to show just about every diet "working". But the studies that run long enough do see the weight come back. Almost everyone gets the weight back, from every diet.

Remember the "thin" version of Oprah Winfrey? Oprah is worth 2.8 billion dollars. She was hugely motivated to lose weight. She has a dedicated staff - a personal trainer, a private chef, access to every type of medical advice and equipment and treatment imaginable. If SHE can't keep weight off for more than 6 months at a time, what chance do ordinary shmucks have?

[jacques_chester]

We're actually arguing about system boundaries now. In my experience, all arguments about the causes of overweight and obesity wind up as boundary disputes.

> People who are obese tend to be that way because their automatic system wants to keep them that way.

As a nitpick, people frequently read homeostasis backwards. While the body's systems can be characterised as a control system with some central governor pulling levers to keep it within certain boundaries, that's not really what happens. This appearance is actually an emergent phenomenon of interacting feedback loops. X goes down, so Y goes up. Y goes up, Z goes up. Z goes up, X goes up. X goes up, Y goes down ... and so on. Equilibrium is maintained in the long run by the summative effect of counter-acting reactions and feedback loops.

Set point theory is basically a way of saying that different folks have different appetites. When they eat ad libitum, some people eat more, some less. Their weight converges to the asymptote of TDEE that balances their diet and the increased BMR from additional weight.

This not news. The point though is that consciousness and self-awareness allows us to examine our actions and modify them. But first you must internalise the locus of control. You either accept that you are the only person who can control your weight, or you don't. Oprah is a perfect example, because by hiring trainers and cooks and so on, she can externalise control. It stops working when she gets bored and fires them. Or when she reaches her "goal weight" and, not realising that it needs to be a permanent change in behaviour, fires them.

The final problem with set-point theory is that it doesn't explain why there has been a sudden burst in overweight and obesity in the past few years in rich countries only. Calories consumed per capita does; so does the steady deflation of the rate calories-per-dollar exchange. People's appetites haven't changed. People probably always wanted the salty, fatty, sugary things. It's just that now is the first time in history that it has been massively abundant and cheap in wealthy societies.

> But the studies that run long enough do see the weight come back. Almost everyone gets the weight back, from every diet.

It's a common truism to say that "the weight comes back", and yet actually, that claim is not really supported either. The literature on long-term compliance is sparse and I suspect the waters muddied because when you perform a medically supervised diet, again the locus of control is external. HAES advocates rely on a single study performed in 1959. A single study.