This is what I always wonder about these studies. How do you differential people who do regular exercise with those who eat healthy? Clearly there is correlation between those.