Really? So would you like to explain the widespread trend of increasing adult onset diabetes, and do you consider having diabetes to be healthy or unhealthy?
First, I don't need to provide an explanation to demonstrate that other explanations are completely unfounded. If you don't have sufficient data to justify your statistical explanation or if your statistical reasoning is fundamentally flawed, then your explanation is unfounded whether or not I can provide an alternative one. Your statement is like an ancient Greek man saying to another: "Really? So if you don't believe in Zeus, what else could explain lightning?" If we don't know then we don't know. That someone comes up with one bogus explanation doesn't mean it's valid -- or even a good working hypothesis -- until someone else comes up with another bogus one. It is the magical mind that expects an explanation -- any explanation -- to any phenomenon. Science readily accepts that we just don't (yet) have an explanation for many things.
Second, there's a world of difference between saying something is related to nutrition in general, saying it relates to some global nutritional variable such as calorie intake, and saying something relates to a more specific nutritional variable such as intake of animal fat. While making each of these statements may be dubious due to the inherent difficulty of doing any sort of nutritional research (it must be observational and longitudinal, and there are too many confounding variables), at each of these three levels of nutritional explanation, there are so many more variables being added, so that to prove the third one you'll probably need many more samples than there are people on this planet. The dimensionality is just too big to tackle with statistics.
Imagine that the real explanation could be something like: amount of animal fat consumed between the ages 0 and 13, and unrelated to anything consumed later. Or it could also be that the same causes of diabetes also reduce the chance of, say, brain cancer, and so what we're seeing is actually a good thing. There are just too many variables, especially that we know food interacts in very complex ways with anything from hormones to gut flora.
What we can say with certainty is that life expectancy has been increasing steadily (not at a steady pace, of course) since the industrial revolution, and that it is very similar among all industrialized countries. The variations are hardly more than a couple of percentage points in spite of sometimes very different diets, different levels of pollution, different amounts of sunlight, different occupations, different leisure activities etc. So countries are so different, yet life expectancy is almost the same everywhere (in the industrial world, of course). Another thing we know for sure is that life expectancy in the developing world is much, much lower.
Of course, rather than clinical studies -- or "clinical", as nutritional studies are observational and wildly confounded -- one could try to study nutrition through biological metabolism, but we're probably decades away from having a really good grasp of human metabolism.
Second, there's a world of difference between saying something is related to nutrition in general, saying it relates to some global nutritional variable such as calorie intake, and saying something relates to a more specific nutritional variable such as intake of animal fat. While making each of these statements may be dubious due to the inherent difficulty of doing any sort of nutritional research (it must be observational and longitudinal, and there are too many confounding variables), at each of these three levels of nutritional explanation, there are so many more variables being added, so that to prove the third one you'll probably need many more samples than there are people on this planet. The dimensionality is just too big to tackle with statistics.
Imagine that the real explanation could be something like: amount of animal fat consumed between the ages 0 and 13, and unrelated to anything consumed later. Or it could also be that the same causes of diabetes also reduce the chance of, say, brain cancer, and so what we're seeing is actually a good thing. There are just too many variables, especially that we know food interacts in very complex ways with anything from hormones to gut flora.
What we can say with certainty is that life expectancy has been increasing steadily (not at a steady pace, of course) since the industrial revolution, and that it is very similar among all industrialized countries. The variations are hardly more than a couple of percentage points in spite of sometimes very different diets, different levels of pollution, different amounts of sunlight, different occupations, different leisure activities etc. So countries are so different, yet life expectancy is almost the same everywhere (in the industrial world, of course). Another thing we know for sure is that life expectancy in the developing world is much, much lower.
Of course, rather than clinical studies -- or "clinical", as nutritional studies are observational and wildly confounded -- one could try to study nutrition through biological metabolism, but we're probably decades away from having a really good grasp of human metabolism.