Having trouble generalizing when a class of complexity $ D $ is closed under polynomial reduction.

For example, consider the following examples:

- $ bigcup_ {c ge 1} mathsf {DTime} (2 ^ {cn ^ 5}) $
- $ bigcup_ {c ge 1} mathsf {DTime} (5 ^ {cn}) $
- $ bigcup_ {c $ ge 1} mathsf {DTime} (n ^ {c log ^ 5n}) $

$ P $ is contained in all these classes, so if we are given a problem $ A $ who belongs to one of these classes, and a reduction of another problem $ B $ at $ A $the polynomial reduction will not affect the overall execution time of the decision. Can it be generalized to a higher class than $ P $?