It is interesting to consider how far the science of a neural substrate for lack of emotion in driving callous behaviour in boys has progressed. Almost 20 years ago, sufficient controversy surrounded holding a conference on the genetics of criminal behaviour that the meeting was moved from the USA to the UK. Now, there is widespread agreement that—in this context—biology cannot be ignored. Fortunately, though, the progress is not just in tolerance for the field, but also in sophistication of the field. It is now relatively uncommon to see references to the biology or genetics of anti-social behaviour. Rather, the emphasis has switched to the biology of specific forms of emotional disruption that give rise to specific forms of anti-social behaviour. There is no biology of criminal behaviour; given sufficient gain and insignificant cost, we might all offend. (If you were told that by stealing the office pencil you would earn a million dollars, would not you be tempted?) But there is biology with respect to specific forms of dysfunction that increases the propensity for offending behaviour.
One form of emotional dysfunction associated with an increased risk for anti-social behaviour is the presence of callous-unemotional traits. Such individuals show markedly reduced guilt, empathy and anxiety. They are more likely to show goal-directed, instrumental, anti-social behaviour; i.e. to use anti-social strategies in order to achieve their goals. Moreover, their offending history typically starts at an early age and continues throughout life. The basic hypothesis that most researchers in the field agree on is that the form of the emotion dysfunction disrupts typical forms of moral socialization. The individual is less able to learn that some actions should be avoided because of their negative consequences to others and self. They are thus more likely to engage in anti-social behaviour given sufficient motivation. The existent neuropsychological and neuro-imaging literature predominantly implicates the amygdala, orbital frontal cortex and superior temporal cortex in their pathophysiology (Blair, 2007).
Of course, work on biology of conduct problems associated with callous-unemotional traits is in its infancy. There have been only two published fMRI studies in adolescents. Moreover, the paper by Stefano De Brito and colleagues in this edition of Brain is the first structural MRI study of boys with this trait. De Brito and colleagues use optimized voxel-based morphometry to examine both grey matter volume and concentration in a relatively large sample of children with conduct problems associated with callous-unemotional traits (n = 23) in comparison with typically developing (n = 25) adolescents. They examined four regions of interest informed by the previous literature: orbital frontal cortex, anterior cingulate cortex, anterior insula cortex and the amygdala. Their results were clear. This population of children with conduct problems associated with callous-unemotional traits showed increased grey matter concentration in the medial orbital frontal cortex and anterior cingulate cortex as well as increased grey matter volume and concentration in the temporal lobes bilaterally.
There are two particular procedural strengths with this study. The first is the use of voxel-based morphometry. Many of the previous studies that have examined anti-social populations using structural MRI have involved manual tracing or semi-automated region of interest guided measurement of brain structures. Such studies have two important disadvantages. First, there is always the increased risk for observer bias in techniques that are not automated. Second, and more fundamentally, it is difficult for independent laboratories to replicate work that is not automated. Definitions of the boundaries of some structures, for example, vary across laboratories. Without precise descriptions of the exact anatomical definitions used, something not typically noted by authors, it is hard to replicate the structural MRI procedures. As such it is then not easy to know whether divergent findings among laboratories reflect inconsistencies in the population or structural MRI procedures. A major strong point of this paper by De Brito and colleagues is the eminent possibility for clear replication that it provides.
The second, and arguably equally important strength of this article, is the relatively restricted age range of the sample (10–13 years). There is now considerable information on structural changes throughout adolescence. As De Brito and colleagues note, grey matter volume decreases from age 4–21 years in a region-specific manner in several cortical areas. Previous work with wider age ranges and unclear matching by subject for age (the groups might be matched by mean age but not necessarily age distribution) may have led previously to confounded results.
How have the findings advanced the field? Importantly, the current study confirms the dysfunction within medial orbital frontal cortex. This region is critical for the representation of reinforcement outcome information (Schoenbaum and Roesch, 2005), a capacity that is crucial to decision making including reaction to moral dilemmas (Blair, 2007). Structural abnormalities have been noted in adult samples within this region and the fMRI literature has also indicated dysfunction, but this is the first study to document structural abnormalities in this region in children.
But potentially more important may be the anterior cingulate cortex findings. Structural abnormalities in anterior cingulate cortex have not been reported in any of the previous structural MRI studies of anti-social populations. However, fMRI studies with psychopathic adults (who show callous-unemotional traits) have implicated anterior cingulate cortex dysfunction during aversive conditioning (Birbaumer et al., 2005) and emotional memory (Kiehl et al., 2001). Moreover, a recent model of psychopathy/callous-unemotional traits suggests that all of cingulate cortex is compromised (Kiehl, 2006). It could be proposed that structural abnormalities within anterior cingulate cortex might explain deficits in the processing of aversive stimuli and empathy that characterize adults with psychopathy and children with conduct problems associated with callous-unemotional traits. However, this suggestion is unlikely. Anterior cingulate cortex lesions do not appear to give rise to the neuropsychological profile seen in this population [e.g. selective deficits in aversive conditioning and response to fearful emotional expressions; see Blair (2007)]. But the current results certainly prompt increased attention to potential rostral and dorsal anterior cingulate cortex deficits in this population. Indeed, while the region has been implicated in psychopathy/callous-unemotional traits, an articulated reason as to why it might be implicated has not yet been provided. Accounts of anterior cingulate cortex functioning involving error monitoring and response conflict have been referred to but in fact observations, including fMRI studies, have shown dorsomedial sensitivity to these phenomena in adolescents with conduct problems associated with callous-unemotional traits (Finger et al., 2008). Of course, there has been recent argument on the basis of animal work that these functions are mediated by regions superior to dorsal anterior cingulate cortex, and that the anterior cingulate cortex itself is involved in the updating and representation of action values (Rushworth et al., 2007). Results from the paper by De Brito and colleagues prompt attention towards analysis of these putative functions.
One striking feature of their results is that children with conduct problems associated with callous-unemotional traits show increase in grey matter concentration relative to comparison individuals in orbital frontal cortex and anterior cingulate cortex. The authors propose an interesting hypothesis regarding their observed increases in structure focusing on a potential delay in cortical maturation. Importantly, this is a hypothesis that can be evaluated through longitudinal studies and if confirmed may eventually suggest novel treatment strategies. But it is notable that previous literature on both anti-social adult and adolescent populations (though not with adolescents whose callous-unemotional status has been ascertained) has typically indicated decreased concentration/volumes. The inconsistency may reflect the younger age of the population studied here—a finding that would be in line with the hypothesis De Brito and colleagues suggest. Certainly, the current results prompt further investigation.
The study by De Brito and colleagues contains important findings which, if confirmed, suggest future directions for the field. Moreover, critically, the paper helps to the development of the field and the understanding of this population. It is always important to remember that there are specific forms of emotional condition, such as callous unemotional traits, that give rise to an increased risk for anti-social behaviour. By recognizing these conditions, we may end stigmatization and begin appropriate treatment.