I remember reading a few years ago that runners shouldn't do yoga because it increases the likelihood of injury. It must have convinced me as I haven't done any yoga in my life. It's too bad I don't remember the exact source of this insight, which claimed to have some basis in research; I think I'd have fun picking apart the method.
The first thing I'd look for is how the study identified subjects. If there was some element of self-selection I would immediately discount the findings. Let's think about the type of runner who might start doing yoga and the likelihood of this person being on the brink of injury - my guess is that those who felt really good about their running probably wouldn't be looking for new methods to take care of their bodies. If you studied runners who were turning to yoga, you were probably studying runners who had an elevated injury risk compared to those who were not turning to yoga.
I look for this self-selection flaw anytime I see some kind of data about running. For example, it's fairly common for a new type of sneaker to receive negative reviews because they cause some kind of injury. But what types of runners are buying new sneakers? Again, my guess (and that's the best I can do in most cases, due to lack of information) is that someone with worn out sneakers is the type of person in the market for new sneakers. And what type of person has worn out sneakers? Someone who runs enough miles to wear out sneakers; even if you think running risk is equally distributed, the fact that these people run more miles puts them at a higher risk if experiencing injury relative to less frequent runners (and their perfectly good sneakers). I'd suggest if you are looking into a new type of sneaker, study the reviews but ignore the comments about any injuries; if it looks good, buy the sneakers and try them once a week for a couple of months until you are sure they won't cause new injuries.
What I've learned about the data in running is that most of the time you can find a distinguishing feature between two comparison groups that might explain the result. This doesn't mean you should dismiss all research findings out of hand but I think it does preclude automatically running with any trendy new insight. As I noted above regarding the best way to try new sneakers, the key is to move slowly and incrementally through change so that the body has time to voice any concerns; no matter what the research says, runners are best advised to listen to their bodies as they move through the constant, controlled, and iterative process of regular experimentation with new techniques, strategies, and equipment.
