Emotional and cognitive processing of narratives and individual appraisal styles: recruitment of cognitive control networks vs. modulation of deactivations

Abstract

Research in psychotherapy has shown that the frequency of use of specific classes of words (such as terms with emotional valence) in descriptions of scenes of affective relevance is a possible indicator of psychological affective functioning. Using functional magnetic resonance imaging (MRI), we investigated the neural correlates of these linguistic markers in narrative texts depicting core aspects of emotional experience in human interaction, and their modulation by individual differences in the propensity to use these markers. Emotional words activated both lateral and medial aspects of the prefrontal cortex, as in previous studies of instructed emotion regulation and in consistence with recruitment of effortful control processes. However, individual differences in the spontaneous use of emotional terms in characterizing the stimulus material were prevalently associated with modulation of the signal in the perigenual cortex, in the retrosplenial cortex and precuneus, and the anterior insula/ventrolateral prefrontal cortex. Modulation of signal by the presence of these textual markers or individual differences mostly involved areas deactivated by the main task, thus further differentiating neural correlates of these appraisal styles from those associated with effortful control. These findings are discussed in the context of reports in the literature of modulations of deactivations, which suggest their importance in orienting attention and generation of response in the presence of emotional information. These findings suggest that deactivations may play a functional role in emotional appraisal and may contribute to characterizing different appraisal styles.

Keywords: appraisal; appraisal styles; emotion; reading; self-regulation.

Figure 1

Schematic depiction of a single trial. Each of the eight trials (one for each of the eight scenes) contained four versions of the textual description of the same scene, representing each of the four conditions of the design, defined by the frequency of emotional or abstract words occurring in the text.

Figure 2

Surface rendering projection of statistical maps associated with contrasts in the whole-brain analysis. The large yellow/orange areas illustrate the activation elicited by the main task of reading stories, irrespective of content. Modulation of signal by abstraction, emotion, and their interaction is superimposed in blue/yellow. For illustration purposes, statistical maps were thresholded at p = 0.005, uncorrected.

Figure 3

Individual differences in the use of emotional words. Top: modulation of deactivations by individual differences in the tendency to use emotional words in the medial aspect of the cortex (left), in the orbitofrontal cortex (centre), and at the level of the angular gyrus (right), shown as parametric maps of t-values overlaid on a template T1-weighted brain. Activations and deactivations of the main reading task are displayed in yellow/orange and light blue, respectively, thresholded for display purposes at p = 0.005, uncorrected. In violet, the modulation by individual differences, at threshold p = 0.01, uncorrected, of the emotional contrast. The left side of the transversal slices is on the left. Bottom: box-plots of signal differences between conditions, displayed separately in subjects with high and low use of emotional words (“HI EMO EXPR”, “LO EMO EXPR”) at the follow-up post-scan recounting of the stories (for illustration purposes, the first and last thirds of a tertile split on the emotional words use scores were used). Here and in the following boxplots, data were centered. Data are divided according to textual description type, in the TCM terminology: Rel (relaxing) contains few emotional or abstract words, Refl (reflecting) is rich in abstract words (“A” on the x-axis), Exp (experiencing) is rich in emotional words (“E” on the x-axis), and Conn (connecting) contains both emotional and abstract words (“A + E” on the x-axis). The light blue ovals highlight the textual version with emotional words, which differ in the groups with high and low use of emotional words at recounting the scenes. AI/VLPFC: anterior insula/ventrolateral prefrontal cortex.

Figure 4

Time-course plots in selected voxels (centered data). Top row: activated voxels in the left DLPFC and VLPFC. Bottom row: deactivated voxels selected from the regression of the individual differences in the use of emotional terms, showing relative activation during fixation. The black line is the regressor obtained by convolving the task with a canonical BOLD curve, which tracks activation due to presentation of the scene (in dark gray background) and the textual descriptions (light gray background). The white background corresponds to the fixation point. The red line is the median signal. Thick blue lines show variation intervals at the 25th and 75th percentile of the data. Light red lines show the same variation at the 2.5th and 97.5th percentiles.

Figure 5

Interaction between emotional and abstract story versions. Top: illustration of the spatial overlap between the interaction emotion and abstraction and the deactivations of the main task (light blue). Statistical maps of t-values overlaid on a template T1-weighted brain. As in Figure 3, yellow/orange denotes task activations, light blue deactivations relative to fixation, thresholded for display purposes at p = 0.005, uncorrected. The interaction is in yellow/blue or in violet (according to its direction). These areas may be compared with those of the regression on individual differences of Figure 3, especially for the medial aspect of the brain on the left, and the orbitofrontal cortex. Bottom: box-plots of signal differences between conditions. The light blue circles highlight the textual version where neither abstract nor emotional words were used, and which drives the interaction. For an explanation of the symbols of the box-plots, see the legend of Figure 3. AI/VLPFC: anterior insula/ventrolateral prefrontal cortex; ctx: cortex; Int: interaction.

Figure A1

Carol George, Malcolm West, Odette Pettem (1997). Adult Attachment Projective. All rights reserved.