I've been looking for some good, and readable, articles on this topic. Hard to find! Here's a link to a pretty good article that talks about why they're hard to find, among other things.
Collecting data on pedophilia is difficult. Collecting data on child sexual abuse is difficult too. In both cases, the pool of subjects is limited by who gets caught and who reports what, etc. The article I linked mentioned one thing in particular I found interesting. They asked a group of pedophiles if they'd been abused as kids and got a fairly high number of yeses. When a lie detector was used, the numbers changed, a lot. Not something I'd thought of. (Silly me!) Apparently when you ask a pedophile whether or not they were abused as a kid, they might say "yes" because that answers makes them a little more acceptable as a human being.