Prefrontal-subcortical pathways mediating successful emotion regulation

Abstract

Although prefrontal cortex has been implicated in the cognitive regulation of emotion, the cortical-subcortical interactions that mediate this ability remain poorly understood. To address this issue, we identified a right ventrolateral prefrontal region (vlPFC) whose activity correlated with reduced negative emotional experience during cognitive reappraisal of aversive images. We then applied a pathway-mapping analysis on subcortical regions to locate mediators of the association between vlPFC activity and reappraisal success (i.e., reductions in reported emotion). Results identified two separable pathways that together explained approximately 50% of the reported variance in self-reported emotion: (1) a path through nucleus accumbens that predicted greater reappraisal success, and (2) a path through ventral amygdala that predicted reduced reappraisal success (i.e., more negative emotion). These results provide direct evidence that vlPFC is involved in both the generation and regulation of emotion through different subcortical pathways, suggesting a general role for this region in appraisal processes.

Figures

Figure 1. Mediation model and hypotheses

We selected right ventrolateral prefrontal cortex (vlPFC) as a predictor region, and performed analyses to search in subcortical regions of interest and throughout the brain for mediators of the relationship between reappraisal-induced vlPFC activity and reappraisal success. A) Path diagram with standard notation for path coefficients. B) Main hypotheses for mediation search analyses.

Figure 2. Results for the [ReappNeg – LookNeg] contrast

This contrast was the main comparison of interest for the mediation analyses. A) Significant reappraisal-induced activation. Positive effects are in hot colors (yellow-red), and negative effects are in cool colors (blue-purple). B) Significant correlations between reappraisal-induced activation and reappraisal success, limited to a search area comprised of activated regions in (A). Positive correlations are shown in red/yellow, and negative correlations are shown in blue. Positive correlations indicate a greater relative increase in activity for participants who report more successful reappraisal, and negative correlations indicate a greater relative decrease in activity for participants who report more successful reappraisal. Thresholds are shown in the color key on the figure. C) Correlation scatterplots for the average activity in left and right ventrolateral prefrontal cortex (vlPFC). L, left; R, right.

Figure 3. Mediation analysis results within regions of interest (ROIs)

A) ROIs in the nucleus accumbens (NAC, top) and amygdala (AMY, bottom). B) Path diagram showing the relationships between regions in the path model. The predictor region in ventrolateral prefrontal cortex (VLPFC) is shown at left, which predicts activity in each of the NAC (top) and amygdala (bottom) regions shown. These are the a paths for each mediating region. The lines are labeled with path coefficients, and standard errors are shown in parentheses. The mediator regions’ (NAC and amgdala) connection to reappraisal success (the outcome) are the b paths for each mediator. They are calculated controlling for VLPFC activity and for other mediators, as is standard in mediation models. ***, p < .001; ** p < .01; *, p < .05, two tailed. The direct path is the c’ path, and this is calculated controlling for both mediators. C) Partial regression scatterplots for the relationships between left (L) NAC and reappraisal success (left panel) and L amygdala and reappraisal success (center panel), controlling for VLPFC and the other mediator. The right panel shows an example of a bootstrapped mediation effect (path ab) for the left NAC. The range on the x-axis spanned by the lighter gray portion of the histogram is the 95% confidence interval for the effect.

Figure 4. Detail of effects in the amygdala (left) and nucleus accumbens (NAC)/ventral straitum (right)

A) Coronal slice (y = −5) showing reappraisal-induced activation in and contiguous with the amygdala region of interest correlated with reappraisal success. Blue: Negative correlations, showing a greater relative decrease in amygdala activity (LookNeg - ReappNeg) with more successful reappraisal. Red/yellow: positive correlations. B) Significant voxels in and contiguous with the amygdala showing a b effect in the mediation (the same relationship as in (A), but controlling for VLPFC activity). Results are two-tailed. C) The anatomical boundaries of amygdalar subregions. BL, basolateral complex; Ce, central nucleus; CM, cortomedian group; EC, entorhinal cortex; HY, hypothalamus; MN, mamillary nucleus. The strongest results appear to be in the entorhinal cortex bordering the basolateral and centromedian nuclear groups. D) Robust regression scatterplot showing relationship between reported reappraisal success (x-axis) and reappraisal-induced activation (relative to the mean) for the peak voxel in panel A. More successful reappraisal was associated with relative reduction in amygdala activity, min r = −.54, t = −2.82, Z = 2.61, p = .0043 one-tailed. E) Voxels in and contiguous with the NAC/ventral striatum region of interest correlated with reappraisal success, as in panel A. The display threshold is set at a false discovery rate of q < .05 for the whole brain. F) Significant voxels in and contiguous with the NAC/ventral striatum showing a b effect in the mediation. G) The anatomical boundaries of NAC and surrounding regions. Cau, caudate; CC, corpus callosum; LV, lateral ventrical; SCA, subcallosal area; SCC, subgenual cingulate cortex. H) Robust regression scatterplot showing relationship between reported reappraisal success (x-axis) and reappraisal-induced activation (relative to the mean). More successful reappraisal was associated with relative increases in NAC/ventral striatum activity, mean r = .76, t = 5.50, p = 7 × 10−6, two-tailed.

Figure 5. Mediation results in whole-brain search

A) Significant regions (p a, b, and a*b) mediating the right vlPFC-reappraisal success relationship throughout the brain. Positive mediators are shown in yellow, and negative mediators are shown in dark blue. The extent around significant effects is shown at p < .01 (orange/light blue) and p < .05 (pink/purple). B) Network clustering of the mediating regions showed the strongest evidence for two functional networks, a group of positive mediators (yellow) and negative mediators (blue). Lines show significant connectivity between regions that could not be explained by any other single mediating region. The path diagram shows the mediation model including activity in each of the two networks (averaged across voxels for each participant) as independent predictors. ***, p < .001; ** p < .01; *, p < .05, two tailed. C) Partial regression scatterplots for the b effect for each network, and histograms showing the bootstrap distribution for each mediation (a*b) effect.

Figure 6. Frontal regions mediated by nucleus accumbens/ventral striatum (NAC)/VS and amygdala

A) Diagram of the mediation model, in which both left NAC/VS and left amygdala regions were included as mediators of reappraisal success. The analysis searched over brain voxels for areas showing a significant mediation (a*b) effect. B) Regions mediated by amygdala. C) Regions mediated by NAC/VS. Yellow/dark blue: positive/negative effects at p < .001, 3 contiguous voxels; orange/light blue: positive/negative effects at p < .005, 10 contiguous voxels. BA, Brodmann’s Area; DMPFC, dorsomedial prefrontal cortex; FROP, frontal operculum; MFP, medial frontal pole; MOFC, medial orbitofrontal cortex; SGACC, subgenual anterior cingulate; SFS, superior frontal sulcus; Pre-SMA, pre-supplementary motor area; VLPFC, ventrolateral prefrontal cortex; VMPFC, ventromedial prefrontal cortex.