First of all, the method used to choose a cut-off is somehow arbitrary. We followed a methodology described by Lin et al., using points of rarity in the observed age at-onset distribution (Lin et al., 2006), but the use of another methodology could have led to another result. An important point is that our cut-off at 22 years separates the sample into two groups of identical distribution, and correctly classify two third of each group, thus reducing the risk of misclassification in each group. Moreover, the exclusion of subjects with ASPD did not modify our results, thus arguing that the two groups do not differ solely on the presence or absence of an ASPD.
