Self-reflection across time: cortical midline structures differentiate between present and past selves

Abstract

The processing of personal changes across time and the ability to differentiate between representations of present and past selves are crucial for developing a mature sense of identity. In this study, we explored the neural correlates of self-reflection across time using functional magnetic resonance imaging (fMRI). College undergraduates were asked to reflect on their own psychological characteristics and those of an intimate other, for both the present time period (i.e. at college) and a past time period (i.e. high school years) that involved significant personal changes. Cortical midline structures (CMS) were commonly recruited by the four reflective tasks (reflecting on the present self, past self, present other and past other), relative to a control condition (making valence judgments). More importantly, however, the degree of activity in CMS also varied significantly according to the target of reflection, with the ventral and dorsal medial prefrontal cortex and the posterior cingulate cortex being more recruited when reflecting on the present self than when reflecting on the past self or when reflecting on the other person. These findings suggest that CMS may contribute to differentiate between representations of present and past selves.

Fig. 1

Brain activity associated with reflecting on self and other in the present and past. (a) The ventral MPFC, dorsal MPFC and PCC were commonly activated by the four reflective tasks (present self, past self, present other and past other) relative to the control task, as revealed by a conjunction analysis. (b) Nevertheless, activity in the ventral MPFC, dorsal MPFC and PCC varied across reflective conditions, as revealed by a whole-brain voxel-wise ANOVA. Displayed at P < 0.001 (uncorrected) on the mean structural MRI of all participants.

Fig. 2

Activity in the ventral MPFC, dorsal MPFC and PCC. Functional ROIs were defined using 6 mm radius spheres around the local maxima of (a) ventral MPFC (MNI coordinates of peak voxel: −6, 54, −2), (b) dorsal MPFC (MNI coordinates of peak voxel: −2, 56, 26) and (c) PCC (MNI coordinates of peak voxel: −2, −58, 26) activation clusters yielded by the whole-brain voxel-wise ANOVA. Effect sizes correspond to average parameter estimates for each reflective condition (present self, past self, present other, past other) relative to the control condition (valence judgments). Error bars represent the standard error of the mean.

Fig. 3

Activity in brain regions that showed a significant interaction between target and time. Functional ROIs were defined using 6 mm radius spheres around the local maxima of (a) dorsal MPFC (MNI coordinates of peak voxel: 4, 46, 44), (b) dorsal MPFC (MNI coordinates of peak voxel: 6, 62, 16), (c) PCC (MNI coordinates of peak voxel: 2, −58, 24) and (d) left inferior frontal gyrus (MNI coordinates of peak voxel: −34, 26, 2). Effect sizes correspond to average parameter estimates for each reflective condition (present self, past self, present other, past other) relative to baseline. Error bars represent the standard error of the mean.