A very nice idea for an experiment with good cohort of trained runners - the author should be commended on getting some good quality athletes to participate. However the protocol relies on precipitating a degree muscle damage. This is not necessarily an issue with regards to the cohort and there is evidence of damage in the study. The issue you have is the study design - the authors have used a cross-over design, which on first glance seems like a good way to approach the problem, but it does not account for the repeated bout effect (adaptation from previous bouts of similar activity) - in other words a reduction in the magnitude of damage indices, which has been shown to last for many weeks (and in some cases, months). I'm surprised there is not even some discussion of this limitation.
I understand the difficulty in recruiting good quality athletes, but it would be nice to see this experiment repeated with a between-subjects design.
