But yes - all these guesses are the scientific facts as far as we know them. You can complain about nutritional science being absolutely immature and probably highly corrupt, but it's still the best we have, and people who follow it aren't sheep anymore than I am for believing that there are atoms just because I was told so.
Many authors (Nassim Taleb, Michael Pollan) talk about using other measures as heuristics. Pollan says 'would your great-grandmother have recognized it as food?'.
Adopting that one rule beats most modern nutrition advice.
We don't have evidence physics is corrupt. That's why your belief about atoms is reasonable.
As far as recognising it as food goes, has there ever been scientific advice in favour of processed foods? This seems like an implicit strawman to me.
The fact that conclusions resulting from 'nutrition science' often seem to contradict, or are found to be completely wrong every few years should be an indication. Today drinking coffee is bad, tomorrow it reduces your risk of getting cancer. Yesterday nuts were bad and will make you fat, today thay aren't that bad after all because the calories they contain mostly leave your body undigested. At one point in time eating eggs more than twice a week was considered bad for you because of the cholesterol they contain, but according to the latest studies they are a 'superfood', and the cholesterol they contain appears to have no relation to blood cholesterol.
The problem with 'nutrition science' (I have to put it between quotes it because I consider it more like religion than science) is that health effects resulting from diet almost exclusively manifest themselves after many many years. This makes it very hard to reliably measure them, not only because it involves tracking large groups of people over a long period of time, who need to stay on a consistent diet, should not move around to much because otherwise other factors around them could affect their health etc.
Research into nutrition health benefits suffers from the same limitations many other epidemiology research suffers. As it turns out, health is correlated to so many other things it is almost impossible to reliably relate health effects to causes. The difference between epidemiologists and 'nutrition scientists' is that the former group directly acknowledges the limitations of their population studies, while 'facts' about nutrition are usually presented as truth, and nobody seems to care to challenge them.
Much of our current 'low fat, high carb' advice stems from the McGovern commission:
http://en.wikipedia.org/wiki/United_States_Senate_Select_Com...
This article provides a good overview:
http://www.slate.com/articles/health_and_science/medical_exa...
There are lots of other things. Nutritional advice on breakfast, for example, is not based on anything more substantial than correlational studies. It ignores the contradictory evidence of traditional eating patterns that avoided breakfast.
I could go on. We're told to eat a low cholesterol diet, for example, even though blood cholesterol is not directly affected by dietary cholesterol in most cases.
I don't have too many supporting links, hopefully someone else can fill the gap. But the cochrane review is a good source of reviews of the state of our evidence.
http://articles.elitefts.com/nutrition/logic-does-not-apply-...
Given breakfast's popularity in our society, one would expect that many people who don't have breakfast can't have breakfast - and this is highly correlated with socioeconomic status and poverty, and we've known for a long time that SES is a greater determinant of academic performance than just about any other factor.
In other words, are you sure your study isn't just talking about poor kids not having breakfast, or is it properly controlling for variables?
Interestingly, even research into eating habits (or at least the way the media picks up and reports on it) appears to be culturally biased, because where I live (Western Europe), there's an almost unrelenting stream of experts telling me how bad skipping breakfeast is. Supposedly it will make you fat, tired, weak and eventually even sick.
It's pretty annoying. I haven't been eating breakfast for almost 20 years without any perceived negative health effects, but every time I tell someone about it I get this lecture about how bad it is to not eat breakfast. This point of view is purely based on conjecture repeated from something people have heard before, never out of personal experience. When it comes to food & health, people really are like sheep, nobody really knows anything about it, but everyone has on opinion, and almost all these opinions are the same, based on 'common sense' and lore, not on scientific facts.
Only 10 years ago common sense said 5 glasses of milk a day were healthy, or bread and cheese for lunch. 1/5th of your diet should consist of dairy products. Today everybody thinks differently. Now counting calories is all the hype, but slowly research is starting to tell us that not all calories are the same in the sense that the kinds of food that contain them are digested differently.
tl;dr: I always get a little frustrated when people start telling me what is or isn't 'healthy' because most people are basically mindlessly repeating what someone else told them. Eating breakfast 'because it's healthy' is one of these things.