Guilt is a self-conscious emotion associated with the negative appraisal of one's behavior. In recent years, several neuroimaging studies have investigated the neural correlates of guilt, but no meta-analyses have yet identified the most robust activation patterns. A systematic review of literature found 16 functional magnetic resonance imaging studies with whole-brain analyses meeting the inclusion criteria, for a total of 325 participants and 135 foci of activation. A meta-analysis was then conducted using activation likelihood estimation. Additionally, Meta-Analytic Connectivity Modeling (MACM) analysis was conducted to investigate the functional connectivity of significant clusters. The analysis revealed 12 significant clusters of brain activation (voxel-based FDR-corrected p < 0.05) located in the prefrontal, temporal and parietal regions, mainly in the left hemisphere. Only the left dorsal cingulate cluster survived stringent FWE correction (voxel-based p < 0.05). Secondary analyses (voxel-based FDR-corrected p < 0.05) on the 7 studies contrasting guilt with another emotional condition showed an association with clusters in the left precuneus, the anterior cingulate, the left medial frontal gyrus, the right superior frontal gyrus and the left superior temporal gyrus. MACM demonstrated that regions associated with guilt are highly interconnected. Our analysis identified a distributed neural network of left-lateralized regions associated with guilt. While voxel-based FDR-corrected results should be considered exploratory, the dorsal cingulate was robustly associated with guilt. We speculate that this network integrates cognitive and emotional processes involved in the experience of guilt, including self-representation, theory of mind, conflict monitoring and moral values. Limitations of our meta-analyses comprise the small sample size and the heterogeneity of included studies, and concerns about naturalistic validity.
Keywords: Guilt; Medial prefrontal cortex; Meta-analysis; Neuroimaging; fMRI.