ArticlesPunishment and psychopathy: a case-control functional MRI investigation of reinforcement learning in violent antisocial personality disordered men
Introduction
Most violent crimes are committed by a small number of men.1 They display a pattern of antisocial and aggressive behaviour that begins in childhood and remains stable throughout the lifespan. They meet diagnostic criteria for conduct disorder in childhood and for antisocial personality disorder in adulthood. Lifelong patterns of poor decision making, impulsivity, and risk-taking behaviours characterise such men, and persist despite repeated punishments enacted by parents, teachers, and the criminal justice system.
Within this population, a subgroup presents with antisocial personality disorder and psychopathy, defined by the psychopathy checklist-revised (PCL-R)2 as including callousness, lack of empathy, an interpersonal style involving grandiosity and manipulation of others, and persistent reactive and instrumental aggression. Such individuals show reduced tonic skin conductance and cortisol concentrations, and have difficulty recognising fear and sadness in the faces of others.3 By contrast, men who have antisocial personality disorder without psychopathy are characterised by emotional lability, mood and anxiety disorders, and reactive aggression.4 Although the two phenotypes emerge early in life, those who develop antisocial personality disorder with psychopathy begin offending at an earlier age, engage in a broader range and higher frequency of offending behaviours,4 and respond less well to treatment in childhood5 and adulthood6 than those without psychopathy. The two groups show distinct differences in brain structure7 and functional responses to empathy-eliciting scenarios8 and emotional stimuli when engaged in goal-directed behaviour.9 Diagnostic classification systems, however, do not distinguish between antisocial personality disorder with and without psychopathy. Rehabilitation programmes typically exclude offenders with psychopathy.
Many decisions, including whether or not to engage in an antisocial act, involve the ability to assess consequences. Selecting an action is dependent on reinforcement learning, whereby possible rewards are weighed against possible punishments on the basis of past experience.10 Men with antisocial personality disorder show impairments in adaptive decision making, characterised by abnormal processing of reinforcement information.11, 12 Their behaviour seems to be driven more by potential rewards than potential punishments (reward dominance). Impaired learning about the consequences of actions might result in overly optimistic and inaccurate predictions of favourable outcomes, resulting in frustration, which could predispose to reactive aggression. Impaired self-control lowers the threshold for aggression in response to such frustration.13 In behavioural studies offenders with antisocial personality disorder and psychopathy have shown notable impairment in using reinforcement information when choosing between punished and rewarded objects in passive avoidance,14 extinction,15 and reversal learning tasks.11, 16, 17 The failure to use punishment information to signal inappropriate behaviour is thought to be the primary deficit in psychopathy, and emerges early in childhood.18
Reversal learning tasks examine the ability to adjust behaviour to changes in reinforcement contingency. Studies of healthy adults and adults with brain lesions have identified a neural network recruited during reversal learning.19 The dorsomedial prefrontal cortex, ventrolateral prefrontal cortex, and dorsolateral prefrontal cortex serve, respectively, to allocate attentional resources, assess conflicting responses, and initiate response modulation. The ventromedial prefrontal cortex, posterior cingulate cortex, and dorsal caudate track changes in reinforcement information, reducing activation in response to punishment, thus signalling behavioural error, change in reinforcement contingency (when reward is expected but punishment received), and the need to adapt behaviour.20, 21
Little is known about the neural systems serving reinforcement learning anomalies in offenders with antisocial personality disorder with and without psychopathy. Offline versions of probabilistic response reversal tasks are self-paced, which means that intervals between the stimulus presentation, response, and feedback vary across trials.11 A variant of this task has been developed for the functional MRI (fMRI) scanning environment that is designed specifically to minimise behavioural differences between antisocial individuals and normal controls. Importantly, the scanner version of the task provides reinforcement more often and more consistently than the offline task. This change lessens task difficulty and the likelihood of confounding introduced by large differences in performance, such as differential contamination by error events.22 The scanner version of the task has been used to assess adolescents with conduct disorder and callous unemotional traits,21 which are the childhood antecedents of antisocial personality disorder with psychopathy.5 Compared with healthy teenagers, adolescents who have conduct disorder with callous unemotional traits showed intact recruitment of the dorsomedial prefrontal cortex, dorsolateral prefrontal cortex, and inferior frontal cortices when changing their behaviour in response to punishment signals, but did not appropriately reduce activation of the ventromedial prefrontal cortex. This finding was interpreted as intact recruitment of regions involved in response change immediately after punishment, but impairment in regions that alter long-term value associated with responses, thereby increasing the likelihood of future incorrect responding.
We used the scanning environment probabilistic response reversal task to do an fMRI study of reinforcement processing and decision making in violent adult offenders with antisocial personality disorder with and without psychopathy. Focusing on responses to punished reversal errors relative to rewarded correct responses, we tested the hypothesis that violent offenders with antisocial personality disorder and psychopathy would show increased activation within the ventromedial prefrontal cortex, caudate, and posterior cingulate cortex to punished reversal errors, as identification of anomalous responding to reinforcement information within such individuals could be useful as a diagnostic biomarker.
Section snippets
Participants and study design
Between January, 2007, and January, 2011, we enrolled 50 men, aged 20–50 years, with reading age higher than 10 years as defined by the Schonell instrument. Eligible participants had no history of major mental disorders (bipolar 1, bipolar 2, major depression, or psychotic disorders), as defined by the structured clinical interview for Diagnostic and Statistical Manual of Mental Disorder, fourth edition (DSM-IV), axis I disorders (SCID I), or self-reported neurological disorders, head injury
Results
The three groups did not differ significantly except for education and PCL-R score (table 1). The offenders with antisocial personality disorder and psychopathy were significantly younger at first conviction than those without psychopathy. They also had higher numbers of convictions for violent crime, but this difference was not significant. Significantly more offenders had lifetime diagnoses of alcohol and cocaine dependence than non-offenders (appendix). Of note, though, the proportions of
Discussion
We investigated the neural basis of reversal learning in violent offenders with antisocial personality disorder with and without psychopathy. Offenders with antisocial personality disorder and psychopathy displayed abnormal responding to punishment signals within the posterior cingulate and insula, illustrated by significantly greater BOLD activation to punished reversal errors than to rewarded correct responses in both regions. These differences were not related to the number of antisocial
References (38)
- et al.
Not just fear and sadness: meta-analytic evidence of pervasive emotion recognition deficits for facial and vocal expressions in psychopathy
Neurosci Biobehav Rev
(2012) - et al.
Neural correlates of a reversal learning task with an affectively neutral baseline: an event-related fMRI study
Neuroimage
(2005) - et al.
Neural correlates of response reversal: considering acquisition
Neuroimage
(2007) - et al.
Artifactual fMRI group and condition differences driven by performance confounds
Neuroimage
(2004) - et al.
Posterior cingulate cortex: adapting behavior to a changing world
Trends Cogn Sci
(2011) - et al.
Common and distinct networks underlying reward valence and processing stages: a meta-analysis of functional neuroimaging studies
Neurosci Biobehav Rev
(2011) The amygdala and ventromedial prefrontal cortex in morality and psychopathy
Trends Cogn Sci
(2007)- et al.
Males on the life-course-persistent and adolescence-limited antisocial pathways: follow-up at age 26 years
Dev Psychopathol
(2002) - et al.
Psychopathy and the DSM-IV criteria for antisocial personality disorder
J Abnorm Psychol
(1991) - et al.
Effects of comorbid psychopathy on criminal offending and emotion processing in male offenders with antisocial personality disorder
J Abnorm Psychol
(2006)