> This seems intellectually interesting to me but it's frustrating because I feel it takes what is a relatively simple issue and makes it needlessly complex. Yes, maybe there is some amount of optimization you can do with your diet, but the simple fact is that there is a not a single study in the history of science which has been able to demonstrate eating at a caloric deficit and gaining weight. Show me an obese person not eating significantly too much.

Taubes would say that an obese person eats too much precisely because he is fat, not the reverse. The arrow of causality has been flipped. His body is demanding increased caloric intake (via insulin, leptin, and ghrelin) to maintain fat stores. Those fat stores can't be liberated when insulin levels remain high. If this hypothetical obese person keeps eating carbs and has become insulin resistant, those fat stores aren't going anywhere without a fight. A calorie restricted diet will help some, as less food produces less insulin, but it's far easier to restrict insulinogenic foods. This allows the fat stores to be utilized, and weight loss is the natural result.

It sounds counterintuitive at first, but this theory states that the human body knows how to maintain caloric intake. That's what hunger is for. But it can only do this if we allow the endocrine system to function normally.

Before dismissing the idea, listen to Taubes's talk: http://youtu.be/ywRV3GH5io0