Everything I've read so far on the subject has documented a correlation between gut flora and certain diseases, but I've yet to see any causative relationship established either way.
It's an interesting dilemma, but it's probably unethical to test the hypothesis directly, i.e. infect naive humans with H. Pylori to see if they subsequently develop metabolic syndrome or type 2 diabetes.
I think you can test it indirectly by testing type 2 diabetics for H. Pylori antibodies and treating those who are positive for it. The National Health and Nutrition Examination Surveys found a correlation between H. Pylori antibodies and elevated hemoglobin A1c. That would be another population for treatment of a presumptive H.Pylori in a double blind trial. If that resolves type 2 diabetes or prediabetic hyperglycemia, that speaks for itself. If I had my druthers, I'd test those responding to H. pylori treatment with normal blood glucose levels for any common defects in their innate or acquired immune systems.
The journal article referenced in this article isn't indexed at PubMed yet, so I can't read it.
I found the link in comment# 1 by going to the named journal, Cell Host & Microbe. It's a method I had to use after the updated PubMed couldn't support my old computer. I usually check the author's surname on the journal's homepage and in their advance online publication or equivalent link.
However, I question whether the response of TLR5 knockout mice to infection is relevant to the question of which comes first, the gut flora or the inflammatory condition.
That's not unreasonable, but I'd like to rule out defects in the innate immune system, e.g. Toll-like receptors and diabetes: a therapeutic perspective.
Association between gastric Helicobacter pylori colonization and glycated hemoglobin levels.
Helicobacter pylori infection is associated with an increased rate of diabetes.
I'd stick with cell culture studies or animals as a last resort. There are plenty of methods of testing these questions that don't involve clinical studies with human subjects. (I dislike clinical human studies anyway; I find them extremely limited in the types of questions that they can answer, and often they are flawed in other ways.)
I was able to find the abstract directly at the journal website. The full abstract, at least, seems to recognize that there is a more complicated two-way interaction between host and flora than the article posted here would indicate. Still, this article makes claims that are not, as far as I can tell, supported by the evidence--for example, the claim that it may be important to control the early environment to prevent conditions such as metabolic syndrome leading to obesity. I think that, to a large extent, the metabolic conditions cause the change in microflora. Since an obese person's biochemistry is not the same as a thin person's biochemistry, I would expect to see a difference in species prevalence of gut microflora as a result.