Automatic adaptive emotion regulation is associated with lower emotion-related activation in the frontoparietal cortex and other cortical regions with multi-componential organization

Abstract

Although some researchers consider automatic adaptive emotion regulation to be an automatized strategy whereas others consider it to be implicit disengagement of deliberative process, to date, its neural correlates have been poorly investigated. In addition, the valence specificity of automatic adaptive emotion regulation and levels of activation relative to the neutral condition are controversial; the former is relevant to the attribution of resilient emotion regulation to positivity bias or emotional stability, and the latter to determining whether regulation is based on emotion-specific or emotion-non-specific processes. In this functional magnetic resonance imaging (fMRI) study, we presented positive and negative emotional pictures to healthy young participants and investigated the neural correlates of automatic adaptive emotion regulation in spontaneous emotional response. A significant negative trait effect (i.e., regression coefficient) on activation was identified both for positive and negative emotional responses in various cortical regions. A cluster analysis identified three clusters among these regions based on the valence specificity of the trait effect and level of activation relative to neutral stimuli. Cluster 1 included regions in the sensorimotor cortex characterized by negative emotion-specific decreases in activation relative to neutral stimuli in adaptive individuals. Cluster 2 included several cortical regions including the bilateral dorsal executive network, anterior cingulate, and inferior frontal gyrus, which were characterized by valence-independent decreases in activation in adaptive individuals. Cluster 3 included the bilateral ventrolateral and dorsomedial prefrontal cortices, right insula, and other posterior regions, which were characterized by increased activation for negative stimuli in non-adaptive individuals. These findings support the assumption that automatic adaptive emotion regulation involves the implicit disengagement of deliberative process and suggest the relevance of different cortical networks to the potential emotion- and valence-specificity of adaptive regulation.

Keywords: acceptance; emotion regulation; fMRI; implicit; mindfulness; prefrontal cortex; reappraisal; spontaneous.

FIGURE 1

Experimental design. Each block consisted of a serial presentation of three emotional pictures with the same valence, i.e., negative, positive, or neutral (control), from the International Affective Picture System. The triad of three blocks with different valences, with the order counterbalanced across participants, was repeated six times. The participants pressed the button after they had sufficiently appreciated each picture to ensure their wakefulness.

FIGURE 2

Negative effects of the emotion regulation trait on activation for (A) negative and (B) positive emotional responses. Significant effects (t-values) of the emotion regulation score (power to live questionnaire) in the second-level multiple regression model are represented by a blue-cyan scale, rendered on the surface and overlaid on the parasagittal section (x = –6) of an SPM12 standard structural brain image. Statistical significance was determined at a cluster-forming threshold of uncorrected p < 0.001 and then corrected to p < 0.05 (family wise error) for the cluster size. No positive trait effect was detected.

FIGURE 3

Results of cluster analysis of 34 identified regions with significant trait effects based on t-values for trait effect and average activation for emotional responses in two valences. (A) Dendrogram of the hierarchical cluster analysis; a threshold of 13.96 for between-cluster distances (dotted line) resulted in a three-cluster solution. (B) Two-dimensional plots of principal components. See Tables 1, 2 for anatomical labels. (C) Anatomical locations of the peaks, indicated by cluster-specific symbols. (D–F) Activation profiles of the representative regions of clusters 1–3 (i.e., the posterior left paracentral gyrus, the anterior left superior temporal gyrus, and opercular part of the right inferior frontal gyrus, respectively). Average emotional responses for two valences are shown separately for non-adaptive (n = 20) and adaptive (n = 20) groups, based on a median-split of low and high emotion regulation scores, respectively.