We are grateful for thoughtful critical points raised by Drs Sekar and Ganapathy. We agree that the prevalence of personality disorders in our sample is high. This sample was chosen for its elevated rates of ASPD – perhaps attributable to a downward drift in occupational functioning due to antisocial features. It is certainly vulnerable to selection bias for several reasons (among them being the fact that some individuals did not meet inclusion criteria), and the 52.4% personality disorder prevalence rate should not be mistaken for a typical community population base rate, or a general rate applying to other populations (claims which were not made in our paper). Also, the prevalence rates of other individual personality disorders (e.g. narcissistic personality disorder – around 3%) were consistent with general population estimates. ^{1,Reference Levy, Reynoso, Wasserman, Clarkin, O'Donohue, Fowler and Lilienfeld2} Additionally, our focus was the characteristics of this ASPD/SSPD group, rather than making any assertions about its prevalence in the general population.

On reading the correspondents' comment about SSPD comorbidity with other personality disorders, we too became intrigued with this possibility and have since conducted further analyses. These revealed, among the other personality disorders, significant SSPD comorbidity only with narcissistic personality disorder. In fact, all three of our sample's individuals with narcissistic personality disorder had comorbid ASPD and SSPD. This additional comorbidity among our ASPD/SSPD group is not surprising, given the problematically high overlap of narcissistic personality disorder with ASPD and other Cluster B disorders, ^{Reference Levy, Reynoso, Wasserman, Clarkin, O'Donohue, Fowler and Lilienfeld2,Reference Millon and Davis3} the conceptual link between the narcissistic and antisocial personalities (e.g. Kernberg's ‘malignant narcissism’), ^{Reference Millon and Davis3} and the positive correlations observed between narcissistic personality disorder and other antisocial personalities such as psychopathy. ^{Reference Hare4} We still believe ASPD/SSPD comorbidity to be meaningful, and not an artifact of the sample, as 50% of individuals with comorbid ASPD/SSPD were not characterised by any additional Axis II comorbidity.

We agree that the frontal cortex (specifically the prefrontal cortex) may be a common abnormality and that this needs further investigation. Clearly, additional research is needed on this comorbid group, in both clinical and nonclinical populations – including ‘unique’ community samples such as our own.