Which statistical analysis to choose?

Hello all,

I'm working on my thesis and having issues with which statistical analysis to choose.
In my research i want to check whether doing physical activity (before and during work) increases work performance. And whether this relationship is mediated by vitality. I will have 2 groups: one intervention group (which will participate in physical activity classes every week) and one control group which will only answer the questionnaires but will not be participating in physical activity classes every week. We will have 3 measurement times: one pre-intervention measurement just before the intervention. One measurement after 3 weeks and one measurement after 6 weeks of intervention.

First I wanted to do a MANOVA: because it can have 2 or more outcome variables without losing power as well as comparing 2 groups. However my tutor said you can't do a mediation with a MANOVA, i forgot about that...

Now i'm really confused: maybe a normal regression mediation?
I really don't know anymore.

If anyone could help lots of thanks!