Evidence is hard to get due to variables in such studies. You have to observe people's daily routine for a long period of time, which gets to be expensive. You also cannot fool the subject, or the people monitoring progress, so there's no way to do such studies double-blind.

For these reasons studies on nutrition are also hard and mostly flawed. But do we need a study to tell us that sugar is poison?

Well, ok, we do, but we can also notice our grandparents weren't eating that much sugar, they weren't suffering that much from obesity or diabetes and that's enough for me.