اثرات عصبی طولانی مدت تنظیم احساسات بر روی فرآیندهای حافظه اپیزودیک

Emotions can enhance memory which is on the one hand advantageous, but on the other hand may be detrimental in the long term, for example in the case of traumatic events. Although cognitive emotion regulation may reduce emotion experience and corresponding neural activation, at present little is known about its influence on long-term memory. We investigated memory for emotional pictures in healthy female subjects 1 year after voluntary emotion regulation using fMRI. Whereas memory performance was not affected by regulation, our data revealed a dissociation of brain regions involved in memory encoding and recognition depending on whether emotional engagement during encoding had been downregulated. Emotional engagement during encoding resulted in a long-term subsequent memory effect in mesolimbic brain regions and hippocampus, and in recognition-related activation in the amygdala. In contrast, when negative emotions had been downregulated during encoding memory performance was predicted by prefrontal activation. Our data suggest that memory for emotionally encoded stimuli is supported by emotional re-activation, whereas memory for successfully encoded items during emotion regulation is rather supported by recognition of features and cognitive contents. These results contribute to research on long-term effects of emotion regulation in everyday life and open new avenues to understand and possibly influence traumatic memory traces.

It is well known that long-term memory is improved for emotional events due to specific neural and hormonal mechanisms that are not engaged during memory for neutral material (Cahill et al., 1994, LaBar and Cabeza, 2006 and Phelps, 2004). Although the temporal development of the emotional modulation of memory processes is still unclear (Wolf, 2008), it is assumed that memory enhancement by emotion reflects the neuromodulatory influence of the amygdala on consolidation processes in the medial temporal lobe (MTL) through engagement of adrenergic stress hormones (McGaugh, 2004, Strange and Dolan, 2004 and van Stegeren et al., 2005). But what are the effects of emotion regulation on these well known processes? Although emotion regulation is an essential ability in everyday life, its effects on long-term memory processes have not been sufficiently researched. Brain imaging studies have shown that amygdala activation during encoding of emotional items is related to improved long-term memory (for a review see LaBar & Cabeza, 2006) due to a modulating effect on the MTL memory system ( Dolcos et al., 2004 and Richardson et al., 2004). In addition, enhanced memory for emotional material seems to be primarily influenced by arousal rather than valence characteristics of the stimuli ( Anderson et al., 2006 and Kensinger and Corkin, 2004) (but see also Mickley & Kensinger, 2008). In post-traumatic stress disorder (PTSD) where patients suffer from hyperarousal and re-experiencing phenomena such as flashbacks of traumatic experiences ( Etkin & Wager, 2007), it has been suggested that prolonged states of adrenergic activation lead to overconsolidation of traumatic memories ( Pitman et al., 2002 and Vaiva et al., 2003). In healthy subjects, administration of the beta-adrenergic antagonist propranolol during encoding has been shown to reduce amygdala-dependent memory for emotional events after minutes ( Strange, Hurlemann, & Dolan, 2003), hours ( Strange & Dolan, 2004) or weeks ( Cahill et al., 1994), presumably through a central adrenergic blockade. However, it is not known whether non-pharmacological, i.e., cognitive, emotion regulation can likewise influence long-term memory. Only a few studies have shown amygdala involvement during short-term (hours to days) recognition of emotional memories ( Kensinger and Schacter, 2005 and Sharot et al., 2004). Recently it was reported that even 1 year after incidental encoding memory for emotionally arousing items, compared to neutral non-arousing items, was enhanced and was accompanied by greater amygdala and hippocampal activation ( Dolcos, LaBar, & Cabeza, 2005). By using this long retention interval the authors are able to distinguish retrieval from early consolidation processes, which cannot be disentangled when testing memory after shorter delays of days or weeks. Here we were interested in the effects of emotion regulation on long-term episodic memory. Emotion regulation has been subject to a number of studies reporting significant reduction of amygdala activation during active regulation of negative events (for a review see Ochsner and Gross, 2005 and Phillips et al., 2008). However, effects of emotion regulation on long-term memory have so far only been investigated in a few behavioral studies showing diverging effects, depending on the regulation strategy used. Thought suppression ( Rassin, Merckelbach, & Muris, 1997) as well as cognitive reappraisal ( Dillon et al., 2007, Richards and Gross, 2000 and Richards and Gross, 2006) had no effect on memory for emotional items, whereas expressive suppression did have an effect ( Dillon et al., 2007, Richards and Gross, 2000 and Richards and Gross, 2006). In order to investigate the neural effects of emotion regulation on long-term memory, we studied memory in subjects 1 year after they participated in an emotion regulation experiment. We tested whether recognition was modulated by emotion regulation during encoding. Moreover, we also analyzed encoding data and investigated, whether brain activation during incidental encoding could predict memory for the respective item 1 year later, and whether this potential subsequent memory effect was modulated by emotion regulation—thus potentially resembling the pharmacological effects of beta-adrenergic blockade. Emotion regulation effectuates reduction of arousal, mostly reflected in reduction of amygdala activation ( Ochsner and Gross, 2005 and Phillips et al., 2008) and arousal seems to be a crucial factor of emotional memory by facilitating encoding and consolidation processes ( Hamann, 2001, Kensinger and Corkin, 2004 and McGaugh, 2004). Therefore, we expected to find effects of successful recognition as well as a subsequent memory effect during encoding in regions of the MTL, including the amygdala, only in cases where emotions were experienced in a natural way, but not when subjects regulated their emotions during encoding.

We assume that when regulating emotions related to negative and salient stimuli one reduces the saliency and arousal of the stimulus, e.g., by taking the position of a neutral observer. By this we “cognitivize” the stimuli thereby encoding features and evaluative identifications. When emotionally engaged during encoding, processing mechanisms were boosted that are related to a saliency-dependent dopaminergic modulation of hippocampal activation, thus fostering subsequent memory. During recognition, amygdala activation was stronger for those items successfully remembered. We suggest that activation of the amygdala during recognition may be indicative for a reinstantiation of emotional experience upon presentation of these stimuli. In contrast, when emotions are regulated during encoding, successful recognition does not rely on a reactivation of emotions but may depend on the recognition of features or cognitive contents, indicated by a performance dependent activation of the DLPFC. It is of crucial importance to investigate whether the learning of cognitive emotion regulation strategies could reduce, e.g., post-traumatic symptoms and re-experiencing of traumatic memories. It has been suggested that PTSD, in contrast to other anxiety disorders, may particularly involve dysfunctional emotion regulation (Etkin & Wager, 2007). New treatments, both psychotherapeutic and psychopharmacologic, aimed at enhancing emotion regulation abilities, may therefore lead in the best case to reduced traumatic re-experiencing during memory processing. Thus, our results contribute to research on the long-term effects of emotion regulation in everyday life as well as in psychotherapy and might increase our understanding dysfunctional accessibility of traumatic memory traces in affective disorders.