Commentary: Are callous unemotional traits all in the eyes? Examining eye contact in youth with conduct problems and callous unemotional traits – reflections on


  • Conflict of interest statement: No conflicts declared.

The presence of callous and unemotional (CU) traits (i.e., reduced guilt and empathy) are associated with a more severe form of conduct disorder (CD) than CD with low CU traits. As a consequence, youth with CD+CU are receiving increasing attention. Previous work had demonstrated that children with CU traits show impaired recognition of fearful expressions but that this impairment can be removed if the intensity of the stimulus is increased by asking the child to focus on the eye region of the fearful face. Such data suggested that problems in eye contact with significant others might be a significant feature associated with CU traits. In this important paper, Dadds and colleagues (2011) provide the first evidence that children with high CU traits show reduced eye contact with both their mothers and fathers during real world interactions. In addition, they report that levels of eye contact to fathers related to both level of fear recognition and mothers’ reports of child empathy levels.

On the basis of these data, Dadds and colleagues hypothesize, in line with Skuse (2003), that eye contact with attachment figures is critical for social-emotional development and that the reduction in eye contact with parents in youth with high CU traits leads to impairment in the development of empathy and conscience. This is an important theoretical statement in marked contrast to other positions stressing the primacy of emotion dysfunction (Blair, 2008). In short, is the reduced eye contact with attachment figures a cause of the emotion dysfunction in CU traits or does dysfunction in the systems necessary for emotional responding also cause the reduced eye contact? This question is likely to prove difficult to answer definitively. However, it is clear that damage to a neural system critically implicated in the emotion dysfunction, the amygdala (Blair, 2008), does cause reduced eye contact with others. This has been demonstrated in work with adult patients with amygdala lesions (Spezio, Huang, Castelli, & Adolphs, 2007). As such, dysfunction in the neural systems necessary for emotional dysfunction seen in children with elevated CU traits does cause a reduction in eye contact. But this, of course, does not disprove the possibility that reduced eye contact with others additionally has some degree of deleterious developmental impact. Longitudinal studies are necessary to investigate this possibility.

With respect to developmental impact, it is interesting to consider here the implications of these data for Skuse’s model. Skuse (2003) argued that eye contact with attachment figures was not only critical for emotional development but also for the development of social cognition, including Theory of Mind; the ability to represent the mental states of others (e.g., their beliefs and intentions). He argued that reduced eye contact with attachment figures might contribute to the development of autism. This would be particularly important because while the representation of others’ feelings may implicate the amygdala, the representation of the beliefs and intentions of others does not appear to do so. As such, a demonstration of a relationship between reduced eye contact in youth with elevated CU traits and impaired Theory of Mind would suggest a deleterious developmental impact of reduced eye contact that could not be attributed to shared neural architecture.

However, currently these novel data cause difficulties for Skuse’s position. In this paper, Dadds and colleagues demonstrate a significant reduction in eye contact with attachment figures in youth with elevated CU traits. Yet it has been repeatedly demonstrated in adults with psychopathy and youth with elevated CU traits (most recently by Jones, Happé, Gilbert, Burnett, & Viding, 2010) that this disorder is not associated with Theory of Mind impairment. Moreover, the cognitive dysfunction seen in autism is very different from that seen in youth with elevated CU traits (Blair, 2008). Dadds and colleagues do report a relationship between level of eye contact with fathers and ‘cognitive empathy’. This might imply that eye contact with fathers at least has some influence on the ease with which the individual represents the beliefs and intentions of others. However, the form of cognitive empathy assessed in this study related to the child’s understanding of others’feelings. As such, these data cannot be used to support the suggestion that reduced eye contact disrupts the development of Theory of Mind.

In many respects the most important implication of these data, particularly in lieu of Dadds and colleagues’ interpretation of them, is their implications for treatment. If Dadds and colleagues are correct that reduced eye contact between youth with high CU traits and attachment figures leads to the development of these CU traits, intervention should be focused on increasing eye contact in these youth with their attachment figures. In this regard, it is interesting to note that work by Dadds’ group has specifically shown that administration of oxytocin increases focus on the eye region of the face. As such, Dadds and colleagues’ hypothesis suggests an important potential therapeutic role for oxytocin in interventions with this population. This is particularly interesting as such a role is not suggested by other positions on elevated CU traits. For example, positions suggesting more general attention-based problems in youth with elevated CU traits would not generate this hypothesis as oxytocin has not been implicated in general attention. Moreover, positions stressing emotional and particularly amygdala dysfunction (Blair, 2008) would definitely not advocate oxytocin as part of a therapeutic intervention. This is because while administration of oxytocin does appear to increase focus on the eye region of the face it also appears to reduce amygdala responsiveness to fearful expressions. However, according to these positions, increasing rather than reducing the amygdala’s response to fearful expressions in youth with elevated CU traits is critical for their treatment, so the emotion dysfunction positions would definitely not support a role for oxytocin in the treatment of CU traits.

In conclusion, this is an important paper for three main reasons: (i) it critically extends our understanding regarding the social-emotional impairment in youth with elevated CU traits, demonstrating that these youth are marked by reduced eye contact with attachment figures; (ii) it places important constraints on hypotheses regarding the potential developmental impact of reduced eye contact; currently, it appears unlikely that reduced eye contact contributes to significant Theory of Mind impairment; and (iii) it suggests treatment implications for youth with elevated CU traits that are not generated by other positions that stress either a general impairment in attention or an emotion dysfunction basis to this disorder.

Correspondence to

James Blair, National Institute of Mental Health, National Institutes of Health, Department of Health and Human Services, 15K North Drive, Rm 115A, MSC 2670, Bethesda, MD 20892-2670, USA; Email: