So I could play around with the linked spreadsheet here all day:
http://www.census.gov/population/metro/data/pop_pro.html
It allows you to graph population and population density by distance from City Hall for two different metros, or the same metro in 2000 versus 2010. Lots of interesting things become apparent if you look at the Pittsburgh Metro in such comparisons. One obvious thing is its density within a mile of City Hall has increased since 2000--that would be the Downtown residential development boom in action.
Another interesting thing I am finding is that the Pittsburgh population count peaks sharply at around 5 miles from City Hall, then starts a more or less steady decline (although with various interesting peaks and valleys along the way). Comparing Pittsburgh to a lot of other similar-sized metros, that's actually pretty unusual--usually they have a peak at least a little farther away, and sometimes it is more a plateau, but then they often decline more rapidly at around the 15 or 20 mile mark. Look for example at Denver, Columbus, Portland, San Antonio, or Kansas City. For really dramatic examples of metros with farther-out plateaus (albeit at a higher overall scale), look at Phoenix or Atlanta.
The one metro I have found so far that looks closest to me to Pittsburgh is Cincinnati. And that, combined with just some basic logic, suggests to me this is in large part a combination of when development occurred plus topography--Pittsburgh's overall development pattern is consistent with what we observed before, namely its peak population boom periods amounting to spreading out along the river valleys spiraling out from the center.
Again, I don't want to awaken any dormant debates, but these sorts of contrasts support being very careful about the assumption that things like transportation patterns used in other cities will automatically be appropriate for Pittsburgh.