Higher emotional granularity relates to greater inferior frontal cortex cortical thickness in healthy, older adults

Abstract

Individuals with high emotional granularity make fine-grained distinctions between their emotional experiences. To have greater emotional granularity, one must acquire rich conceptual knowledge of emotions and use this knowledge in a controlled and nuanced way. In the brain, the neural correlates of emotional granularity are not well understood. While the anterior temporal lobes, angular gyri, and connected systems represent conceptual knowledge of emotions, inhibitory networks with hubs in the inferior frontal cortex (i.e., posterior inferior frontal gyrus, lateral orbitofrontal cortex, and dorsal anterior insula) guide the selection of this knowledge during emotions. We investigated the structural neuroanatomical correlates of emotional granularity in 58 healthy, older adults (ages 62-84 years), who have had a lifetime to accrue and deploy their conceptual knowledge of emotions. Participants reported on their daily experience of 13 emotions for 8 weeks and underwent structural magnetic resonance imaging. We computed intraclass correlation coefficients across daily emotional experience surveys (45 surveys on average per participant) to quantify each participant's overall emotional granularity. Surface-based morphometry analyses revealed higher overall emotional granularity related to greater cortical thickness in inferior frontal cortex (p_FWE < 0.05) in bilateral clusters in the lateral orbitofrontal cortex and extending into the left dorsal anterior insula. Overall emotional granularity was not associated with cortical thickness in the anterior temporal lobes or angular gyri. These findings suggest individual differences in emotional granularity relate to variability in the structural neuroanatomy of the inferior frontal cortex, an area that supports the controlled selection of conceptual knowledge during emotional experiences.

Keywords: Affect labeling; Aging; Emotion granularity; Insula; Orbitofrontal cortex; Well-being.

Fig. 1

Emotional experience surveys for two example participants. Daily survey ratings for two participants are provided as examples to illustrate how individual differences in day-to-day positive and negative emotional experiences contribute to overall emotional granularity scores. A Participant with high overall emotional granularity reported highly differentiated, nonoverlapping emotional experiences. In this participant, endorsement of one emotion was less likely to be accompanied by endorsement of other emotions on that day. B Participant with low overall emotional granularity reported less differentiated emotional experiences. In this participant, endorsement of one emotion was often accompanied by endorsement of other emotions on that day

Fig. 2

Structural correlates of overall emotional granularity. A The neuroimaging analyses of overall emotional granularity focused on the following IFC ROIs: (1) posterior inferior frontal gyrus, which included pars opercularis and pars triangularis (areas 44/45; pink); (2) lateral orbitofrontal cortex, which included pars orbitalis (Area 47) and Area 12 (cyan); (3) lateral orbitofrontal cortex, which included Area 11 (blue); and (4) dorsal anterior insula (yellow). In control analyses, we also examined the anterior temporal lobes (light green) and angular gyri (dark green). B Higher overall emotional granularity was associated with greater cortical thickness in the left and right lateral orbitofrontal cortex when controlling for age, sex, education, group (control or intervention), time interval (the number of days between the survey date and the MRI), and overall emotional experience intensity. The left-lateral orbitofrontal cortex cluster extended into the dorsal anterior insula according to more parcellated surface-based atlases. The images were peak-level and cluster-level corrected (p_FWE < 0.05) for the ROI analysis, and peak-level thresholded at p < 0.005, uncorrected, for whole-brain analysis. The color maps (red) reflect the maximum T value in each analysis