There are a few criteria one can use for choosing intervals.
- make the intervals as uniform as possible
- cover the range in a "reasonable" number of intervals
- appropriately reflect the data distribution
- make the boundaries unambiguous
Selection B has boundaries that overlap, so can be ruled out immediately. Selection D has ranges that go from 19 to 9, so does not have very uniform intervals. Likewise, selection A defines the first interval to be twice the width of the others, so it isn't a good choice, either.
IMO, the best of the offered choices is ...
... C.) 20–34, 35–49, 50–64, 65–79