Hello!
I strongly doubt that there is any effective difference (no pun intended) between the 2 ways; it's likely that they result in the same code. (No, I haven't verified this.) If your code doesn't need to calculate the index for every iteration, I'd feel free to use either, as they'll both boil down to a simple subtraction. If, on the other hand, you need the index more than just occasionally, and you're iterating across the whole vector, you might consider a 3rd alternative: incrementing (or decrementing, as required) a size_t variable, and use that as an index.