اثر ارزیابی مجدد شناختی بر واکنش پذیری فیزیولوژیکی و حافظه هیجانی

Abstract We investigated the effect of cognitive reappraisal on emotional arousal, facial expressivity and subsequent memory. Men and women viewed emotionally negative pictures while they attempted to either increase or decrease negative emotions elicited by the pictures, or to simply view the pictures. Neutral pictures were also presented with instructions to simply view the pictures. Concurrent changes in emotional arousal and valence were assessed with skin conductance responses (SCRs) and facial corrugator electromyographic responses (EMG), respectively. Picture memory was assessed with an immediate recall test and a delayed recognition test. Relative to simply viewing pictures, voluntary reappraisal to increase negative emotion generated greater facial corrugator EMG and SCR responses, and reappraisal to decrease negative emotion generated decreased corrugator EMG responses. Men showed enhanced recognition for pictures presented during the increase and decrease conditions, whereas women showed comparable recognition performance across all regulation conditions. The modulation of subsequent recognition memory associated with decreasing emotion was inversely associated with changes in physiological responses. Our results suggest that sex is an important factor to consider in determining how reappraisal-induced physiological changes are associated with subsequent changes in memory. These findings contribute to our understanding of how reappraising emotion exerts both immediate and enduring influences on physiological responses and subsequent memory.

Introduction Humans have the unique ability to monitor and voluntarily modify their emotional states to achieve cognitive goals. The ability to appropriately regulate emotion is critically important to one's mental health and impairments in emotion regulation have been linked to a variety of psychological and somatic disorders (Bonnano et al., 2004 and Davidson et al., 2000). One particular form of emotion regulation, cognitive reappraisal, has been a focus of particular recent interest. Reappraisal refers to individuals’ cognitive efforts to increase or decrease the emotional impact of emotion-eliciting events by altering cognitive representations (Gross, 1998). For example, one's typical reaction to a picture of a person in distress might be decreased by imagining the person as being an actor. Conversely, one's reaction might be increased by imagining the person as being a close relative. Reappraisal has been theorized to be a less resource-demanding and more effective emotion regulation strategy than other emotion regulation strategies such as emotion suppression (i.e., suppression of overt emotional responses) because it can modulate early stages of emotion generation, before emotional reactions have fully unfolded (Gross, 1998). Recent investigations of emotion regulation have shown that conscious attempts to increase and decrease emotion result in modulation of physiological and neural responses to emotional events (Dan-Glauser & Gross, 2011; Demaree et al., 2004, Dillon and LaBar, 2005, Driscoll et al., 2009, Jackson et al., 2000, Kim and Hamann, 2007, Lee et al., 2009, Mauss et al., 2007, Moser et al., 2006, Ochsner et al., 2004 and Ray et al., 2010; Sheppes et al., 2009). For example, increased and decreased facial frown muscle activity were observed during reappraisal to increase and decrease negative emotions, respectively, in line with subjective affect report (Gross, 1998 and Ray et al., 2010). Increased skin conductance responses and heart rate were observed when regulation strategies increased emotional responses compared to decreased emotional responses (Driscoll et al., 2009). Recent functional neuroimaging studies showed that voluntary reappraisal to increase and decrease modulated neural activity in the amygdala (Kim and Hamann, 2007 and Ochsner et al., 2004), a neural structure that has been extensively implicated in the experience of emotionally arousing stimuli. Increased physiological arousal, and amygdala responses to emotionally meaningful stimuli have been widely implicated in enhanced episodic memory for the stimuli (Abercrombie et al., 2008, Bradley et al., 1992, Cahill and McGaugh, 1995 and Hamann et al., 1999). For example, research participants better remembered stimuli that felt more arousing than those felt less arousing (Bradley et al., 1992). When the identical images were presented with more arousing stories, participants’ episodic memory was enhanced compared to when the images were presented with less arousing stories (Cahill and McGaugh, 1995). Moreover, functional neuroimaging studies have found that increased activity in the amygdala in response to emotional pictures was positively associated with enhanced memory recall and recognition for the pictures (Cahill et al., 2001 and Hamann et al., 1999). Given the close relationship between emotional arousal and subsequent memory, and the modulatory influence of cognitive reappraisal on emotional arousal, one may predict that reappraisal-induced changes in arousal would also be associated with alteration of episodic memory of the reappraised events. Previous studies investigating cognitive consequences of emotion regulation have shown mixed results. Reappraisal efforts to increase emotional relevance produced enhanced episodic memory; whereas reappraisal to decrease emotional relevance produced either enhanced (Dillon et al., 2007) or unchanged (Richards and Gross, 2000 and Sheppes and Meiran, 2008) memory performance. Examples of memory enhancement despite decreased emotional arousal suggest that reappraisal may influence emotional memory via pathways external to its modulatory effect on arousal. This study was designed to better characterize the relationship between cognitive reappraisal and memory for the reappraised event. The retention interval between encoding and retrieval may play a role in modulation of emotional memory by cognitive reappraisal. Previous studies investigating the effect of emotion regulation on memory have typically tested memory performance shortly after the encoding regulation task (Bonnano et al., 2004, Dillon et al., 2007, Richards and Gross, 2000 and Sheppes and Meiran, 2008). However, the enhancing effect of emotion on memory often increases at longer delays, and this time-dependent effect has been hypothesized to result from the critical contribution of memory consolidation, a process that unfolds gradually over days and weeks (McGaugh, 2000). Therefore studies using longer retention intervals would be more appropriate in investigation of reappraisal-associated memory modulation. In line with this, a study of the modulation of emotional memory by arousal and reappraisal found significant emotional memory and modulation effects. That is, individuals who used reappraisal more often reported lower levels of emotional arousal for emotional words than those who used reappraisal less often (Nielson and Lorber, 2009). When a surprise recognition test was conducted one week after the initial encoding, the emotion-dependent memory enhancement for words was smaller for frequent reappraisers less frequent reappraisers. In addition to these factors, the type of episodic memory test administered may also have an important role in determining the modulation effect of reappraisal on memory. Studies of emotion regulation and memory have typically tested memory using either a free recall or a recognition task. Recall and recognition tasks are known to differ in task difficulty as well as in the relative contribution of two underlying memory processes. That is, recall is dependent on recollection, a process mediated by hippocampal and prefrontal regions whereas recognition can be performed on the basis of recollection or familiarity (Kim and Cabeza, 2009 and Yonelinas, 2002). Because reappraisal may differentially influence component processes contributing to episodic memory, and recall typically is considered more difficult than recognition, the effects of reappraisal may be observed differentially for recall vs. recognition tasks. In addition to retention interval and types of memory, sex may also play in important role in influencing how reappraisal affects emotional memory. Sex differences in the frequency of emotion regulation strategies have been reported (Gross and John, 2003 and Welborn et al., 2009). In addition, neuroimaging studies have reported that men and women engage prefrontal neural resources differentially when successfully regulating responses to negatively charged emotional stimuli (Domes et al., 2010, Mak et al., 2009 and McRae et al., 2008). Moreover, substantial evidence suggests that men and women engage different neural and cognitive mechanisms underlying emotional memory (Buchanan and Tranel, 2009, Cahill, 2006, Cahill et al., 2001, Canli et al., 2002, Ferree and Cahill, 2009, Kring and Gordon, 1998 and Seidlitz and Diener, 1998). These findings suggest that the neural resources available for successful memory encoding may differ between men and women, and this differential availability could interact with the effects of regulation on memory. Accordingly, in this study we anticipated that men and women might differ in their ability to regulate emotion, as measured by physiological responses. Because few laboratory studies have examined sex differences in emotion regulation, we did not have strong a priori predictions regarding which sex would exhibit better emotion regulation ability. Current evidence is equivocal with regard to whether men or women are better able to regulate emotions. For example, a neuroimaging study found that men were better able to down regulate activity in regions associated with emotional arousal, including the amygdala, and also exhibited less activation in regions associated with effortful processing, suggesting that men are more efficient at emotion regulation ( McRae et al., 2008). However, another neuroimaging study found that women recruited prefrontal regions associated with emotion regulation to a lesser extent than men during the regulation of emotion ( Domes et al., 2010). The goal of this study was to assess the effect of cognitive reappraisal on emotional reactions and emotional memory and to determine whether sex differences modulate these processes, particularly the effects of reappraisal on emotional memory. We obtained and analyzed both physiological emotional responses and emotional memory to examine the relationships between reappraisal and its influences on emotional memory. In the current study, participants viewed a series of unpleasant pictures and neutral pictures while performing a reappraisal task. The task for negative pictures included three conditions where participants were asked to either increase or decrease negative emotions elicited by the pictures, or to simply view the pictures. Neutral pictures were presented with instructions to simply view the pictures. We assessed emotional arousal via skin conductance responses (SCRs) and emotional valence via facial corrugator electromyographic responses (EMG) (Lang et al., 1993). Recording both SCRs and facial EMG concurrently enabled assessment of the effects of cognitive reappraisal on both arousal and valence. Memory for the emotional and neutral pictures was subsequently assessed with free recall and recognition tests. We predicted that reappraisal to increase emotion would increase physiological responses to negative pictures and reappraisal to decrease emotion would decrease physiological responses. We also predicted that men and women might differ in their ability to regulate emotion, as measured by SCRs and EMGs. Further, based on the close relationship between emotional arousal and enhanced subsequent memory, we predicted that these sex differences in emotion regulation would be associated with corresponding differences in subsequent memory for the stimuli for which emotional responses were regulated.

Results 3.1. Manipulation check: self-reported emotion regulation strategies Participants described the typical strategy they used in each regulation condition with the negative pictures. The majority of participants (76%) reported using only one strategy for each regulation condition. However, when participants reported using more than one strategy (24%) they were asked to order them to reflect the frequency of use. The relative frequency of the primary strategy (i.e., the strategy reported first) used by all participants was calculated separately for men and women. For the decrease condition, 11 men (64%) reported trying to actively pretend that the pictures were taken from unreal situations (i.e., movies) and were fake. Three men (18%) reported attempting to framing the situation to produce positive outcomes. Three men (18%) reported adopting a disinterested manner, as if they were watching TV or magazine or rather focus on peripheral details. For the increase condition, 15 men (88%) pretended that they or people personally important to them were involved in the scenes. A small percentage of men reported that they focused on the negative aspect of the scenes (6%) or exaggerated their emotional responses (6%). Women generally reported a similar distribution of strategy use. In the decrease condition, 9 (50%) women reported to actively pretend that the pictures were taken from unreal situations (i.e. movies) and were fake. Three (16%) women reported focusing on positive outcomes, and another three (22%) women reported adopting a disinterested manner as if they were watching TV. Other responses included suppressing their facial reactions (6%) and focusing on peripheral aspect of the pictures (6%). For the increase condition, 12 women (66%) reported pretending that they or people personally important to them were involved in the scenes. Three women (16%) reported focusing on negative aspects of the scenes. Other responses included exaggerating (12%) or maintaining (6%) their emotional responses. Chi-square tests on the frequency of each strategy used for enhancing and decreasing emotions showed no sex difference for either the decrease or increase conditions, χ2 (5, 35) = .02, ns, and χ2 (5, 35) = .79, ns, respectively. Finally we tested whether men and women differed in social desirability that was known to be associated with the tendency to comply with experimental demand and could therefore influence self-report responses (Crowne and Marlowe, 1960). An independent-sample t test was run on the MCSD scores. We found that men (M ± SD = 4.50 ± 2.17) and women (M ± SD = 3.8 ± 1.97) did not differ in social desirability (t = 0.975, ns). We also examined whether MCSD scores were correlated with reappraisal-induced changes in physiological responses and memory. No significant correlations were found (rs < 2.2, ps > 0.1). Overall, as instructed, all participants successfully used types of reappraisal strategies as their main regulation strategies. Three participants reported primarily using strategies (i.e. exaggeration or suppression of reactions) other than reappraisal and, therefore, were removed from further statistical analyses. One participant who failed to provide strategy was also removed. Thus, we report the results from the remaining 32 participants (15 women). The means and standard deviations for physiological responses and memory performance across picture types and regulation conditions are shown in Table 1. Table 1. Physiological and memory measures by emotion and regulation condition. Valence Neutral Negative Watch Decrease Watch Increase Mean S.D. Mean S.D. Mean S.D. Mean S.D. SCR (μS) Men N = 17 .164 .138 .193 .171 .160 .120 .293 .182 Women N = 15 .126 .130 .170 .150 .179 .227 .229 .147 EMG (μV) Men N = 17 2.94 .944 3.00 .952 3.03 .951 3.23 1.05 Women N = 15 3.56 1.54 3.57 1.55 3.66 1.49 3.95 1.47 Recall (%) Men N = 17 14.62 6.87 30.63 12.05 29.90 11.14 35.04 11.12 Women N = 15 14.81 5.05 34.72 9.79 39.16 10.18 39.44 9.02 Recognition (%) Men N = 17 41.79 18.83 70.26 13.38 64.13 19.09 70.75 12.30 Women N = 15 42.41 13.45 66.48 20.12 70.09 17.47 68.70 15.42 Note : S.D. Standard deviation; μS, Microsiemens; μV, Microvolts. Table options 3.2. Emotion regulation and facial corrugator EMG activity A 3 (regulation: Decrease, Watch, Increase) x 2 (sex) ANOVA was conducted on differential corrugator EMG scores for negative pictures, calculated as described above. A main effect of regulation was found, F(2, 60) = 15.21, p < .0001. Follow-up analyses revealed that corrugator EMG activity was greater during the Increase condition (M = .33, SEM = .07), t (31) = 3.98, p < .0001, and marginally smaller during the Decrease condition (M = .03, SEM = .03), t(32) = 1.76, p < .09, compared to the Watch condition (M = .09, SEM = .04) ( Fig. 2). No main effect of sex or interaction effect was found, F(1, 30) = .06, and F(2,62) = .85, respectively, ns. The means for corrugator EMG responses in each condition separately for men and ... Fig. 2. The means for corrugator EMG responses in each condition separately for men and women. The Y axis shows mean differential EMG activity, calculated by subtracting the mean EMG activity for the watch condition for neutral pictures from the corresponding value for each regulation condition for negative pictures. Figure options 3.3. Emotion regulation and physiological arousal A 3 (regulation: Decrease, Watch, Increase) x 2 (sex) ANOVA was conducted on differential SCRs for negative pictures. A main effect of regulation was found, F(2, 60) = 6.01, p < .005. Follow-up analyses revealed that the differential SCRs were greater during the Increase condition (M = .11, SEM = .03) than the Watch condition (M = .023, SEM = .02), t(31) = 2.82, p < .01, and the decrease condition (M = .036, SEM = .02), t(31) = 3.85, p < .001 ( Fig. 3). The overall difference in differential SCRs between the Decrease and Watch conditions was not significant, however, t(31) = .42, ns. No main effect of sex nor interaction effect was found, F(1, 30) = .19, ns, and F(2,60) = .36, ns, respectively. The means for SCRs in each condition separately for men and women. The Y axis ... Fig. 3. The means for SCRs in each condition separately for men and women. The Y axis shows differential activity, calculated by subtracting the mean EMG activity for the watch condition for neutral pictures from the corresponding value for each regulation condition for negative pictures. Error bars indicate the standard error of the mean. Figure options 3.4. Emotion regulation and memory correlates 3.4.1. Free recall A 3(regulation: Decrease, Watch, Increase) x 2 (sex) ANOVA was conducted to test the effect of regulation and sex on free recall of negative pictures. No main effects of regulation, F (2, 60) = 2.18, or interaction, F(2,60) = .87, was observed. The main effect of sex was marginally significant, F (1, 30) = 3.97, p < .056. Overall, women tended to recall more negative pictures than did men ( Table 1). 3.4.2. Two-week delayed recognition A 3 (regulation: Decrease, Watch, Increase) x 2 (sex) ANOVA was conducted on corrected recognition. There was a marginally significant interaction between regulation and sex, F(2,60) = 2.49, p = .09. No main effects were significant. Post-hoc analyses were conducted for men and women separately. For men, there was a simple main effect of regulation, F(2,32) = 5.02, p < .05. Follow-up analyses revealed that men recognized more pictures from the Increase condition (M = 25.82, SEM = 3.89), t(16) = 2.74, p < .05, and from the Decrease condition (M = 26.55, SEM = 3.67), t(16) = 2.65, p < .05, than those from the Watch condition (M = 18.95, SEM = 3.99), yielding a significant quadratic trend across the Decrease, Watch and Increase conditions, F(1,16) = 9.35, p < .008 ( Fig. 4). For women, there was no significant simple main effect of regulation, F(2,28) = .39, ns. The means for differential recognition scores in each condition, separately for ... Fig. 4. The means for differential recognition scores in each condition, separately for men and women. Error bars indicate the standard error of the mean. Figure options 3.4.3. Regression analyses To better characterize the relation between physiological changes and modulation of recognition memory due to emotional reappraisal we conducted hierarchical multiple regression analyses. We first derived regulation-specific changes in memory and physiological responses by calculating differences between each regulation condition and the control condition (Increase – Watch, Watch-Decrease). Increase-specific and decrease-specific changes in recognition were entered in the analyses as the dependent variables. Corresponding changes in EMGs and SCRs, and sex were entered as the predictor variables in Step 1. Two-way interactions (i.e. EMG changes x sex and SCR changes x sex) were entered in Step 2. To minimize multicollinearity, all predictor variables were mean-centered. Decrease specific changes: The regression analyses indicated that men had greater changes in memory due to reappraisal than women (β = .48, t = 3.01, p < .005). Overall, enhancement in memory was predicted by the degree of reductions in EMGs (β = − .24, t = 2.17, p < .05) and marginally by the reductions in SCRs (β = −.22, t = 1.79, p < .09). None of the two-way interaction terms were significant (ps > .2). Increase specific changes: There were no main effects (ps > .1) or interactions (ps > .3) related to increase-specific changes. 3.5. Self-report questionnaires 3.5.1. Regulation success A 3 (Regulation: Decrease, Watch, Increase) x 2 (Sex) ANOVA was conducted to assess how confident men and women participants were in their ability to comply with each regulation instruction condition. Results indicated that the main effect of regulation was significant, F(2,60) = 5.44, p < .01. Follow-up t tests revealed that participants felt less successful in complying with decrease instructions (M = 3.44, SEM = .16) compared to the Watch instructions (M = 4.16, SEM = .14), t(31) = 3.37, p < .001 ( Fig. 5). However, participants reported that they were equally successful in complying with the Increase (M = 3.81, SEM = .16) and Watch instructions, t(31) = 1.54, ns. No main effect of sex nor interaction effect was found, F(1, 30) = .46, ns and F(2,68) = 1.07, ns, respectively. The means for success ratings in each regulation condition for men and women. ... Fig. 5. The means for success ratings in each regulation condition for men and women. Error bars indicate the standard error of the mean. Figure options 3.5.2. Post-hoc arousal ratings A 2 (Emotion: Negative, Neutral) X 2 (Sex) analysis of variance on arousal ratings confirmed that, as expected, the negative pictures were rated significantly more arousing than the neutral pictures, F(1, 30) = 848.09, p < .0001. No main effect of sex, F(1,30) = .36, ns, or an interaction effect between sex and emotion was observed, F(1,30) = .007, ns. The means and standard deviations of post-hoc arousal ratings are presented in Table 2. Table 2. Post-hoc mean arousal ratings by stimulus type and regulation condition. Valence Neutral Negative Watch Decrease Watch Increase Mean S.D. Mean S.D. Mean S.D. Mean S.D. Men N = 17 1.73 .49 4.51 .65 4.69 .53 4.57 .67 Women N = 15 1.58 .27 4.46 .87 4.57 .65 4.57 .64 Table options A 3 (Regulation: Decrease, Watch, Increase) x 2 (Sex) ANOVA was conducted on arousal ratings for negative pictures to examine whether picture assignment to each regulation conditions were equally arousing across regulation conditions. There was no main effect of regulation, F(2, 60) = 1.26, ns, sex, F(1, 30) = .08, ns, or interaction effect, F(2,60) = .26, ns.