But until such studies are complete, and the FDA is content with the efficacy and safety of these drugs in prediabetes, it's entirely up to each patient's physician as to whether to prescribe one or another of these drugs for patients who don't yet have type 2 diabetes. In theory, they should help in patients with prediabetes, and several would help diminish insulin resistance. Indeed, the DPP shows that metformin does indeed help

Whether the good things these drugs might do (such as decrease insulin resistance and delay the onset of type 2 diabetes) outweighs the bad things that might occur (such as gastrointestinal side effects and the rare risk of lactic acidosis with metformin, and the drain on the patient's pocketbook as insurance will probably not cover off-label use) is an individual decision based on the individual's medical history and tolerance of risk.

I think any patient contemplating using medications to prevent type 2 diabetes should sign up for a study that's recruiting such patients, or have a long talk with their physician about whether such medication would be appropriate for them.