Really wrote:
But, as already mentioned, small sample size and it is just 1 study. Not enough to make any definitive statements.
It's not just one study though. I've seen maybe three or four studies over the years that have looked at this and NONE of them have shown that running on trails versus roads reduced the likelihood of injuries. That's consistent with my own experience and observations. I'd get to 150 miles a week and most of it by far was on roads and streets and I never had injury troubles. Lots of guys in my club got to or beyond 100 and were mostly on roads and never seemed to have injury troubles either.
Steve Robbins at McGill has done a lot of research on what causes running injuries and he's found that instability is the biggest problem. Pavement is stable. But a longitudinal study only paints a general picture so there may well be individual athletes here and there who could have trouble with pavement and should stay off it. You can't invalidate someone's personal experience with a longitudinal study. But those particular people's experiences in the study will be offset by those of people who don't struggle with pavement.
If you're convinced that you need a trail you should run on one. Aesthetically they're often superior. But there's no need to obsess over running on them.