Until they compare Metformin vs Acarbose in humans, this is the best rodent trial that has ever been performed to date (they used several different kinds and it was very well done). If you are really interested in this topic, I suggest you do the research on it because my brain can't remember all the details about everything I researched so it is best that I don't play the role of messenger.
Also, SGLT2 inhibitors like empagliflozin and canagliflozin are much better for general health than Metformin too.
I personally combine Acarbose + Empagliflozin after having done extensive research. I don't get any side effects besides maybe shitting a little more but Metformin did that to me too.
Metformin blunts the benefits of exercise, which makes it a hard no for me. I don't see the point in using it when there are better things.
If lowering A1C is the goal, I noticed better benefits from berberine than metformin, without the negative exercise effects.
This guy breaks everything down well here.