Featured Published Manuscript of the Month: March, 2018


Chosen and summarized by: Karina Borja


Sethi, A., O'Nions, E., McCrory, E., Bird, G. &Viding, E.  (2018). An fMRI investigation of empathic processing in boys with conduct problems and varying levels of callous-unemotional traits. NeuroImage: Clinical, 18, 298-304.


Corresponding Author: Arjun Sethi, ([email protected] ), University College London.



The featured study investigated the neural basis of empathy in 66 children with conduct disorder (CP) and varying levels of callous-unemotional traits (CU), as compared to typically developing children (TD), aged 11-16, and matched for socio economic status, and intellectual ability. The CP group was further divided into a group of kids with CP and high levels and low levels of CU traits (CP/HCU, CP/LCU). Functional images were acquired during the presentation of fear and neutral- inducing situations under three affective- introspection conditions,  requiring: i) introspection about one’s own feelings ('Own');  ii) thinking about what another person is feeling in a fear-inducing situation ('Other'); and iii) introspection about one’s own feelings in response to hearing about another person in a fear-inducing situation ('Feel for others'). 




Some of the key findings were:


-A conjunction analyses across each condition (‘Own´ fear min. 'Own' neutral, ‘Other' fear min. 'Own' neutral, ‘Feel for other’ fear min. 'Feel for other' neutral) revealed what the authors described as a core affective introspective network, comprising clusters within the aCC/mPFC, the aI, as well as in the occipital lobe in the TD children.


-Such a network was not observed in the CP/CU groups.


-Group comparisons of the neural activity of all the regions identified within the core affective introspective network revealed no differences between the TD and the CP/CU groups across conditions.


-A significant group effect was observed within the occipital lobe, an effect driven by both CP/CU groups compared to the TD kids.


- A Bayesian approach supported the null hypothesis about a lack of differences in the core affective introspection network regions, and rejected it in the occipital lobe across conditions.


-Post-hoc explanatory analyses were then conducted to clarify whether the lack of  a core affective introspection network in the CP group was driven by reduced activation in any particular condition. The authors observed that:


·      TD showed greater responsivity to fear-inducing than to neutral stimuli across conditions within the affective introspective network regions.


·      CP/HCU kids showed a trending responsitivy to fear-inducing stimuli as compared to the neutral stimuli during the 'Own' emotions introspection within the core affective network ROIs.


·      Likewise, CP/HCU kids showed significant neural activity in the ACC/mPFC region during the 'Other' emotions but not in the aI or occipital lobe.


·      CP/HCU kids did not show significant neural activity in any of the core affective network regions or in the occipital lobe during the 'Feel for other' condition.


·      CP/LCU kids failed to show a significant responsitivy to fear-inducing stimuli as compared to the neutral stimuli during the 'Own' emotions introspection conditions within the core affective network regions.


·      CP/LCU did show significant neural activity during fear-inducing stimuli as compared to the neutral stimuli during both the 'Other' emotions and 'Feel for other' conditions in the core affective network regions but not in the occipital lobe.


Some key implications:


·      In TD kids, an 'affective introspective network' underlies the introspection of one's own and other's emotions. This network includes regions like the aCC/mPFC, aI, and the occipital cortex.


·      This network is not consistently activated in CP kids, but there is no robust evidence to confirm these observations, it may be the case that CP kids may active this network inconsistently.


·      Also, the lack of different neural activity between groups may be due to the task demands, which involved a paradigm that required deliberate introspection and relatively mild affective stimuli, a process that might have influence the spontaneous recruitment of the network (i.e. See Meffert et al., 2013).


·      Instead, context-specific deficits in CP/HCU kids during introspective conditions concerning others, suggest that these kids could lack the ability to understand other people's affective states.


·      CP/LCU kids seem to be less efficient in self-focused emotional states.