I just ran across another mention of the most recent arthritis study, so I'm posting a link here. The abstract is public, the full text is subscription only and I don't have library access to Am. J. Preventive Med. If anyone else does, I'd be interested to see - from the abstract, it appears that the one thing they didn't control for is total cardio.

It's interesting that there was actually considerably less arthritis among the "long distance runners" than among the control group (the abstract doesn't define long distance, either), but the numbers were not statistically significant.

Again, I don't doubt that there are individuals whose joint condition prevents them from running, even with the best body mechanics. But whether running caused that is a whole 'nother question; and even if it had, it begs the question of what their body mechanics were like when they did run. If someone had, say, worked in the basket factory all their life and had five or six carpal tunnel surgeries, they probably wouldn't be able to bicycle, either.