You aren't going to. Even assuming, for the sake of argument, that such a relationship does exist in the first place, the feasibility of producing a compelling

You aren't going to. Even assuming, for the sake of argument, that such a relationship does exist in the first place, the feasibility of producing a compelling study that compares the health of people who consume added sugar to people who don't would be basically impossible. It's difficult in the extreme to avoid added sugar. So much so that people who do manage to do so are simply not going to be comparable to the rest of the population. Not even remotely. What you could do is an observational study that looks at overall sugar consumption and looks for some sort of correlation with the magnitude. But countless studies like that have already been done, so one must assume that they aren't providing the kind of results that you're looking for.