Latent-Variable Modeling of Brain Gray-Matter Volume and Psychopathy in Incarcerated Offenders

Advanced statistical modeling has become a prominent feature in psychological science and can be a useful approach for representing the neural architecture linked to psychopathology. Psychopathy, a disorder characterized by dysfunction in interpersonal-affective and impulsive-antisocial domains, is associated with widespread neural abnormalities. Several imaging studies suggest that underlying structural deficits in paralimbic regions are associated with psychopathy. Although these studies are useful, they make assumptions about the organization of the brain and its relevance to individuals displaying psychopathic features. Capitalizing on statistical modeling, in the present study (N = 254), we used latent-variable methods to examine the structure of gray-matter volume in male offenders, and assessed the latent relations between psychopathy and gray-matter factors reflecting paralimbic and nonparalimbic regions. Results revealed good fit for a 4-factor gray-matter paralimbic model and these first-order factors were accounted for by a superordinate paralimbic "system" factor. Moreover, a superordinate psychopathy factor significantly predicted the paralimbic, but not the nonparalimbic factor. The latent-variable paralimbic model, specifically linked with psychopathy, goes beyond understanding single brain regions within the system and provides evidence for psychopathy-related gray-matter volume reductions in the paralimbic system as a whole.