From the Independent (UK), in an article titled Scientists debunk decades-old theories on losing weight (emphasis my own):
If you cut out a chocolate bar each day you will lose only one-third of the weight that experts had thought. For decades, doctors have based their advice to those who want to lose weight on the assumption that cutting 500 calories a day will see the weight fall off at the rate of 1lb a week.
"This is wrong," Kevin Hill, of the National Institutes of Health in the United States, said. "It does not happen." The error has arisen because the calculation did not take account of changes in metabolism as weight falls. The body adjusts to reductions in energy intake (calories eaten) by slowing its energy output (calories expended). The result is that forgoing that daily chocolate bar containing 250 calories will lead to about 25lb of weight loss if it is sustained for three years, much less than the 78lb predicted by the old dieting assumption.
I don't go on restrictive diets, but on behalf of those who do, may I say: "Whoops!" doesn't bloody well cover it. Maybe they should've based their claims about weight loss on some kind of... oh, I don't know, empirical evidence?
Mainstream medicine has apparently been running with an untested theory. In theory eating 500 fewer calories per day should equal 1 pound of weight loss weekly. Here's how the Mayo Clinic puts it:
Because 3,500 calories equals about 1 pound (0.45 kilogram) of fat, you need to burn 3,500 calories more than you take in to lose 1 pound. So if you cut 500 calories from your typical diet each day, you'd lose about 1 pound a week (500 calories x 7 days = 3,500 calories).
It sounds simple, right? And that's the whole problem. Nothing in the human body is simple.
Researchers and doctors must have seen, over the years, that things didn't seem to be working out that way. But apparently, when the data put their hypothesis in doubt, they throw out the data. That is, if people aren't losing weight at the expected rate, then they must be lying about what they're eating, i.e. the data are flawed. Doctors didn't see that the equation wasn't true in actual human experience; they just saw fat liars. People do lie (prolifically!) about what they eat and drink, but someone surely should've gotten to the bottom of this much earlier, and realized the mistake. In the meantime, many thousands of overweight people who were diligently cutting calories have been disdained as fibbers by their doctors, on the basis of their less-than-predicted weight loss.
Medical folks should've known better for two reasons. First, the human body is extraordinarily complex, and "calories in, calories out" was always an absurdly, ridiculously simplistic idea (e.g. see this post and this one). Secondly, didn't they consider the hundreds of thousands of years of human evolution, during which starvation was a common danger? And surmise, therefore, that our bodies just might have ways of protecting energy reserves in the face of a declining food supply? Their simple equation, in which 500 calories x 7 = 1 pound fat, assumed that our bodies would not react to a sudden loss of 500 calories / day, but would simply burn through energy reserves to compensate. That didn't account -- obviously -- for the possibility that our bodies might find a way to conserve energy. There is very little respect for nature in modern medicine, else they might have realized what an asisine assumption that was.