Brain activity and prosocial behavior in a simulated life-threatening situation
Authors: Zanon M., Novembre G., Zangrando N., Chittaro L., Silani G.
Published in: NeuroImage, Vol. 98, September 2014, pp. 134-146.
Abstract: To study the neuronal basis of altruistic behavior, we investigated functional connectivity within brain networks of participants who exhibited either a self-benefit behavior or an altruistic one in a life-threatening situation simulated in a virtual environment. In particular, participants were asked to evacuate a virtual building on fire and, without being previously informed, they were faced with a decision on whether to stop and help a trapped virtual human, at the possible cost of losing their own life in the virtual experience. Group independent component analysis (gICA) applied on blood oxygen-level-dependent (BOLD) functional images revealed significant differences between the group of participants who showed selfish behavior and those who acted prosocially. Specifically, an increased functional connectivity in the salience network, comprising the anterior insula (AI) and the anterior mid cingulate cortex (aMCC), was observed in the selfish group compared to the prosocial one. Conversely, higher ICA weights in the medial prefrontal cortex and temporo-parietal junction (TPJ), were observed in the prosocial group. The findings show that an increased functional connectivity of the salience network, which suggests an enhanced sensitivity to the threatening situation and potential danger for the individual, resulted in more selfish choices, while the engagement of the medial prefrontal and temporo-parietal cortices subserved prosocial behavior, possibly due to their role in perspective-taking. The study provides the first online neurophysiological measurement of prosocial decision-making during threatening situations, opening new avenues to the investigation of neuronal substrates of complex social behaviors.
Copyright: This is the author's version of the publication. The original publication is available at