A lot of nutritional studies have a common flaw: 1. Various health outcomes and indicators are caused by a food or combination of foods that you regularly eat. 2. A diet is studied where these foods get removed from your diet, to be replaced with whatever the diet under study is. 3. The participants in the diet study have poorer than average health outcomes and indicators (otherwise, why would they try changing their…
Because they're asked to, as part of the experiment?
>4. Regression to the mean happens, and all sorts of dietary changes appear to be good for your health
Why would the mean be better, since in (3) you mentioned that those people had "poorer than average health outcomes and indicators" to begin with? For regression to the mean, they should have started the diet after some extreme readings. But who said those are what prompted them to the diet? People don't start on a diet when they get extremely heavy, they start and stop diets all the time, while being continuously obese.
>Basically, if you're fatter than average, and it's caused by the foods you eat, you can fix that by making a list of everything you eat and not eating those foods. "Whatever you don't eat" is a more average diet than what you're currently eating, so you'll get more average results.
Well, if you eat the same calories from your new "Whatever you don't eat", then you'll get as fat.