The controls were local children (n=475) coming from the same schools as the victimized subjects.