In an excellent article (that gives us a nice shout out), the Football Outsiders try to analyze why the Colts have so many injuries on defense in recent years (ESPN pay piece).
They don't come to much of a conclusion, but posit what so many have wondered before, if the size of the Colts relates to the number of injuries.
In the case of Indianapolis' defense, the designed clash of bodies usually involves its defender being smaller than the average defender at his position because vice chairman Bill Polian's mantra in populating Colts' Tampa 2 defense is "speed over size." Indeed, the undersized nature of the defense is an injury theory that has been floated by Colts fans themselves. The fact that other Tampa 2 teams like the Pittsburgh Steelers and Buccaneers show up in the top 10 lends further credence to the idea.
Now, are we ready to claim that we can predict injuries based on the average body mass index of a team's defense? Of course we aren't. After all, there's more than one way to tear a pectoral muscle; just ask D'Qwell Jackson. More generally, are upper-body injuries entirely what separates injury-prone defenses from relatively healthy ones? No. Dallas, for example, has had a low AGL in the past five seasons despite having the highest upper-body AGL percentage.
The findings aren't as conclusive as I would like. I worry that this hypothesis has been repeated so many times, that some will take it as fact, when every study I've ever seen says the same thing: we don't really know for sure. For a theory so wide spread, I would really expect to see more solid data supporting it. Until it comes, we have to hold open the possibility that much of the Indy AGL problem stems from the safety position. It could be that there's no defensive-wide issue as much as Indy has had fits keep Sanders and Bullitt healthy.