Compliance levels in studies aren't relevant to our individual choice of diet, though, because we can observe our own compliance for ourselves by recording our intake. Observing our own bodies' response to a diet takes a much bigger commitment, so it's nice to have some generalizations from science. For the public health debate, there's no reason to separate the two, but from an individual perspective it is unnatural not to.
It's one thing believe that the jury is still out on some of this; it's quite another to stick to preexisting beliefs on the grounds that virtually all recent studies have some flaw or other.
Interesting that you assume I have some particular opinion :-) I don't necessarily disagree with Taube, though I don't think it signifies much that he can persuasively make a point and disparage those who disagree with him when he's talking to lay people. I'm just tired of seeing studies discussed as if the results reflect what happens when people actually follow some particular diet, when that's not what researchers are really interested in. (I.e., researchers take compliance as part of the phenomenon being studied, but journalists and lay people reading and discussing the results assume that compliance is uniformly good or has been factored out somehow.)
Edit/PS: If you interpret studies as showing what happens when people actually eat a certain diet, then compliance _is_ a flaw -- a really big one! We don't know what people are actually eating, only that on average it's significantly more than they report. And there's no questions researchers would _like_ to know how actual intake relates to weight gain or loss. It's just rather inconvenient and expensive to find out, and it's arguably less important than the public health question of how instruction/exhortation relates to weight gain or loss.
Compliance problems in studies may be relevant to our individual choice of diet, since they are likely to predict our own personal compliance. In general (as Taubes points out), most people can lose significant weight by reducing their daily caloric intake to starvation levels. However, compliance is so poor that it's not a viable solution. And it wouldn't be correct to say that this problem is irrelevant to those who are looking to start a new diet.
I personally find the combination of the (potentially flawed) human studies and the (more scientifically rigorous) animal studies to be fairly persuasive. You can find fault with either of the two, but in sum, the two pieces of evidence seem quite solid.
"If you interpret studies as showing what happens when people actually eat a certain diet, then compliance _is_ a flaw -- a really big one!"
To me, the studies likely skew in that direction, while also providing stronger evidence when interpreted as showing what happens when people try to eat a certain diet.
"I don't necessarily disagree with Taube, though I don't think it signifies much that he can persuasively make a point and disparage those who disagree with him when he's talking to lay people."
Sure. On the other hand, he's a researcher who's been investigating this issue for around a decade, and has published a very rigorous (by most accounts) book on the subject. Given the stakes, I suspect that most people would eventually take to the streets in an attempt to improve the situation.
I think we do mostly agree, but I think compliance is a very different factor for each individual. Looking back through different fads and various swings in nutritional orthodoxy, many people have succeeded in building strong, lean bodies whether they believed in eggs, tofu, wheat grass, or protein shakes. That's a fact that simply disappears and is forgotten when we focus on aggregate results from mostly non-compliant subjects. I am pleased and heartened by the conclusion of the study Taube criticizes: "It appears that substantial differences in proportions of dietary macronutrients play only a modest role in weight loss success, and that success is possible on any of these diets provided there is adequate adherence. Getting individuals to adhere to whatever diet they choose to follow deserves more emphasis." That is excellent common sense.
It's one thing believe that the jury is still out on some of this; it's quite another to stick to preexisting beliefs on the grounds that virtually all recent studies have some flaw or other.
Interesting that you assume I have some particular opinion :-) I don't necessarily disagree with Taube, though I don't think it signifies much that he can persuasively make a point and disparage those who disagree with him when he's talking to lay people. I'm just tired of seeing studies discussed as if the results reflect what happens when people actually follow some particular diet, when that's not what researchers are really interested in. (I.e., researchers take compliance as part of the phenomenon being studied, but journalists and lay people reading and discussing the results assume that compliance is uniformly good or has been factored out somehow.)
Edit/PS: If you interpret studies as showing what happens when people actually eat a certain diet, then compliance _is_ a flaw -- a really big one! We don't know what people are actually eating, only that on average it's significantly more than they report. And there's no questions researchers would _like_ to know how actual intake relates to weight gain or loss. It's just rather inconvenient and expensive to find out, and it's arguably less important than the public health question of how instruction/exhortation relates to weight gain or loss.