Divergent network substrates of individual differences in empathy and mentalizing

Introduction: The ability to understand other people's mind and feelings encompasses different abilities such as empathy, the ability to share affective states of others, and mentalizing, the ability to attribute mental states to others. Functional neuroimaging has identified distinct substrates of both empathy and mentalizing. While for example empathizing with the pain or suffering of others consistently involves dorsal anterior insula cortex (dAI), inferior frontal gyrus (IFG), and anterior midcingulate cortex (aMCC), mentalizing relates to activity in a network including dorsomedial prefrontal cortex (dmPFC), posterior temporo-parietal junction (pTPJ), and superior temporal gyrus/sulcus (STG/STS). Whether this divergence is also found at the level of brain structure is unknown. Here, we employ MRI covariance analysis to jointly assess structural network substrates of individual differences in empathy and mentalizing. Methods: We studied 154 healthy participants (94 women, mean±SD age=40.5±9.5 years). We aggregated measures derived from multiple tasks (Samson et al., 2010; Klimecki et al., 2013; Kanske et al., in press) to create constructs of empathy and mentalizing. T1-weighted MRI was obtained using a 3T Siemens Verio scanner. FreeSurfer was used of generate cortical surface models and to measure cortical thickness (Fischl and Dale, 2000). Analysis was performed using SurfStat (Worsley et al., 2009). Based on intersections of task-based functional activations in the same subjects (Kanske et al., in press) and meta-analytical results (Lamm et al., 2011; Mar, 2011), we defined seeds involved in empathy (dAI, aMCC, IFG) and mentalizing pTPJ, STG/STS). We furthermore studied the right supramarginal gyrus (rSMG), a region proximal to the pTPJ that has been suggested to play a role in affective perspective taking (Silani et al., 2013). To map structural covariance networks, we correlated thickness of each seed with thickness across the entire cortical mantle. We studied the interaction between seed covariance strength and individual differences in empathy and mentalizing. Findings were corrected for multiple comparisons using random field theory. Results: We observed that individual differences in empathy and mentalizing related to the covariance of non-overlapping networks. Specifically, individuals with high empathy score had increased right dAI covariance to lateral prefrontal regions, whereas importantly, covariance of TPJ, dmPFC, STG/STS were not modulated by individual differences in empathy (Figure 1). On the other hand, individuals scoring high on mentalizing capacity had increased pTPJ and STG/STS network covariance to mPFC and TPJ (Figure 2). And again, no modulation of dAI, aMCC and IFG was observed by individual differences in mentalizing scores. Interestingly and in line with previous work, rSMG was modulated by empathy, but not mentalizing, to regions similar to the covariance network found in right dAI (Figure 3)