I'll try to clarify. According to the studies I've seen, mortality is decreased for all runners compared to non-runners. However, the greatest benefit was for people who ran a moderate amount. People who ran a lot regressed somewhat vs. the moderate runners, but still not to the level of the non-running population.

There are also some serious methodological questions about the studies involved. The best analysis I've seen is here:

http://www.runnersworld.com/sweat-science/will-running-too-m...

So no, there's no solid evidence that any amount of running is harmful. Maybe if someone did a study in a place where pollution was extra bad (e.g. Beijing) they would find such an effect, but I'm not aware of any such efforts.