The first study does not look at training effect.

The second study, does not have a robust design. There needs to be a control group.