Need help with pooling groups!

#1
So I am not completely sure whether I can pool my groups or not to compare them, or if it would be better to compare them pair-wise. The purpose of this comparison is to test the efficacy of my stress protocol. Here is my experiment design:

In my first data set I have three different types of subjects divided in three groups: 1, 2, 3
In the other part of my data set I have the same type of subjects from the groups 1, 2, 3 but they have been stressed. So their groups are called 1s, 2s, 3s.

Let me clarify further: for example, the three different groups differ only in age: 1 (20y), 2 (40y), 3 (60y). This is my first data set.
My second data set are the same group-conditions (20y, 40y, 60y) but stressed (different people, but same age). They only difference in the second data set of groups is that they been stressed.

Can I pool my non-stressed group 1-3 to one group and compare it to a pooled stressed group from the 3 stressed groups to assess the efficacy of my stress protocol? Using a Mann Whitney U -test, independent t-test (Welch etc.)?
Simply comparing each pair (i.e. 1 with 1s, 2 with 2s, 3 with 3s) would increase the risk of type 1 error, and is a big "nono"? Or am I completely wrong here?

Probably a dumb question. Your help would be very much appreciated!
 
Last edited: