How do I know they got their statistics right? It seems like a very easy mistake to make to look at health of people and see runners tend to be more healthy, not taking into account that health issues might prevent people from running in the first place. (So running would just be an indicator for health, not a cause).