Elsevier

Neuropsychologia

Volume 147, October 2020, 107561
Neuropsychologia

Neural substrates of long-term item and source memory for emotional associates: An fMRI study

https://doi.org/10.1016/j.neuropsychologia.2020.107561Get rights and content

Highlights

  • Long-term item and source memory for objects encoded with emotional and neutral background scenes was studied using fMRI.

  • Performance of emotional item and source memory increasingly relied on recollection versus familiarity.

  • Increased recollection-dependent memory was found along with increased activation within recollection-sensitive regions.

  • These findings suggest that memory for emotional associates is robust over time and mediated by recollection processes.

Abstract

Since Tulving's influential work on the distinction between familiarity and recollection-based retrieval, numerous studies have found evidence for differential contribution of these retrieval mechanisms on emotional episodic memory. Particularly, retrieval advantage for emotional, compared to neutral, information has been related to recollection-, but not familiarity-mediated processes. Neuroimaging studies suggest that this recollection-based retrieval for emotional information is related to stronger engagement of regions in the medial temporal lobe (MTL), posterior parietal cortex (PPC), and prefrontal cortex (PFC). In the present study, we investigated neural correlates related to long-term memory of neutral information that has been associated with emotional and neutral contexts, using functional magnetic resonance imaging (fMRI). During encoding, different neutral objects integrated with emotional or neutral scenes were presented. One week later, the encoded objects were intermixed with new ones and participants had to indicate whether the objects were previously seen or not, using the Remember/Know procedure (item memory). Furthermore, memory for the correct scene background category was also tested (contextual source memory). First, replicating previous findings, we observed a preference for recollection-dependent memory retrieval versus familiarity-dependent memory retrieval for those neutral objects encoded in emotional compared to neutral contexts. Second, consistent with these behavioral effects, objects encoded with emotional, compared to neutral, scenes produced larger memory-related activity in recollection-sensitive brain regions, including PPC and PFC regions. Third, correctly retrieved emotional compared to neutral contextual information was associated with increased activity in these brain areas. Together, these results suggest that memory for information encoded in emotional contexts is remarkably robust over time and mediated by recollection-based processes.

Introduction

A large amount of empirical evidence has robustly confirmed that there is a memory advantage for emotional events (Bradley et al., 1992; Dolcos et al., 2017; Dolcos et al., 2004, 2005; Weymar et al., 2009). For instance, when emotional and neutral pictures are presented and memory is subsequently tested, emotionally arousing scenes are better remembered than less arousing, neutral ones. This effect is found for free recall and recognition memory tasks (Bradley et al., 1992; Dolcos et al., 2004; Dolcos et al., 2005; Weymar et al., 2009), for different retention intervals (immediate and delayed; Ritchey et al., 2008; Wirkner et al., 2018; Schümann et al., 2017), and various stimulus materials (images, faces, sounds; Weymar et al., 2009; Maratos and Rugg, 2001; Righi et al., 2012; see for reviews, Dolcos et al., 2017, Dolcos et al., 2020; Weymar and Hamm, 2013). Furthermore, remembering an emotionally relevant event does not take place in isolation, but is typically accompanied by rich detailed information of the encoding episode (Ochsner, 2000; LaBar and Cabeza, 2006). According to Tulving's seminal work (Tulving, 1985), this mnemonic advantage is related to autonoetic consciousness which is expressed by means of recollection- (remembering specific spatial, temporal, or other contextual details of events), rather than familiarity-based retrieval (knowing that an event occurred but not being able to remember specific contextual details) (Dolcos et al., 2005; Ochsner, 2000; Sharot et al., 2004; Tulving, 1985; Weymar et al., 2010).

Whether emotional information facilitates memory when perceived in the context of a more complex event or in relation to other pieces of information and/or source details, also termed source or relational memory, is less clear (see for reviews, Chiu et al., 2013; Dolcos et al., 2017; Mather and Sutherland, 2011; Mather et al., 2016; Murray and Kensinger, 2013). Some studies observed that retrieval of source information is facilitated when associated with an emotional, compared to neutral, stimulus (e.g., scenes or words). These effects have been shown for the recognition of colors, spatial location, and temporal order of the items (D’Argembeau and Van der Linden, 2005; D'Argembeau and Van der Linden, 2005; Doerksen and Shimamura, 2001; MacKay & Ahmetzanov, 2005; Mather et al., 2009; Mather and Knight, 2008; Nashiro and Mather, 2011; Rimmele et al., 2011; but see Koenig & Mecklinger, 2008; Mather and Knight, 2008; Maddock & Frein, 2009). Several studies also showed that memory for neutral objects or words integrated as a part of emotional cues (e.g., high arousing words or emotional scenes), are better retrieved than when encoded with neutral or less arousing cues (Martinez-Galindo, & Cansino, 2016; Guillet and Arndt, 2009; Luck et al., 2014; Maratos and Rugg, 2001; Pierce and Kensinger, 2011; Smith et al., 2004, Smith et al., 2005; Ventura-Bort et al., 2016). However, other studies have reported a negative impact of emotion on memory for neutral or emotional source cues (e.g. objects, words, frames, spatial location, or language) presented with emotional, compared to neutral, material (Bisby and Burgess, 2014; Bisby and Burgess, 2014; Ferré et al., 2019; Kensinger et al., 2007; Madan et al., 2017; Madan et al., 2012; Mather et al., 2006, 2009; Mather and Knight, 2008; Murray and Kensinger, 2012; Nashiro and Mather, 2011; Rimmele et al., 2011; Touryan et al., 2007). Altogether, these findings point to divergent effects of emotion on source memory, in some cases enhancing and in others decrementing memory retrieval.

The Arousal-Biased Competition (ABC) Theory (Mather et al., 2016; Mather and Sutherland, 2011) integrates these conflicting findings of emotion effects on source memory. According to this model, the modulatory effects of emotion on memory binding depend upon the attentional priority that the critical cue receives during learning (via bottom-up perceptual salience, top-down attentional focus, or via past experience with particular stimuli). Therefore, emotional arousal can have opposing effects on neutral associated cues, enhancing their memory if high-prioritized or integrated in the emotional event, or weakening their memory if low-prioritized or perceived as a competitor for resources against emotional cues (Mather et al., 2016; Mather and Sutherland, 2011).

Along with the priority-dependent effects, the retention interval is another important feature that may exert crucial influence on the effects of emotion on source memory (Mather et al., 2016; Mather and Sutherland, 2011). Indeed, increasing retention interval and thereby memory consolidation has been shown to potentiate the enhancing effect of emotion on episodic memories (McGaugh, 2004; Phelps and Sharot, 2008; Schümann et al., 2017; Sharot et al., 2004; Yonelinas and Ritchey, 2015). In line with this, longer retention intervals favor memory retrieval for items paired with emotional (particularly unpleasant) information in comparison to neutral material, likely by facilitating consolidation processes (Pierce and Kensinger, 2011). Thus, memory for neutral information and/or source details may be enhanced if integrated with emotional material, particularly when the retention interval is long enough to allow for consolidation processes to take place.

Findings from our studies using event-related potentials (ERPs) support these assumptions. When neutral objects are presented with emotional and neutral background scenes and participants are instructed to mentally integrate or connect both events (which gives to such associations high “attentional priority”), neutral objects associated with emotional scenes do not only undergo deeper encoding processing (Ventura-Bort et al., 2016a), but also facilitate long-term memory storage (Ventura-Bort et al., 2016b). Moreover, when ERPs are measured during a 1-week-delay recognition task, a larger late positive-going waveform over centro-parietal regions, compared to correctly identified new objects (ERP Old/New effect) is exclusively found for objects encoded in emotional background scenes (Ventura-Bort et al., 2016b; Ventura-Bort et al., 2017; see for similar results, Martinez-Galindo & Cansino, 2016; Maratos and Rugg, 2001; Meng et al., 2017). This effect was also observed during a long-term spontaneous retrieval task, suggesting that the memory enhancement is driven by the acquired relevance of the triggering event due to its association with emotional contextual information (background scene) at encoding (Ventura-Bort et al., 2019). Given that the late ERP Old/New effect has been associated with recollection-based retrieval (Rugg and Curran, 2007; MacLeod and Donaldson, 2017), our findings extended prior emotional item memory studies (Weymar et al., 2009; Weymar and Hamm, 2013), indicating that the long-term memory retrieval for emotional associates is also mediated by the process of recollection (Curran, 2000; Duarte et al., 2004; Rugg and Curran, 2007; Düzel et al., 2003; MacLeod and Donaldson, 2017).

In the present study, we followed up on this ERP research and used functional magnetic resonance imaging (fMRI) to identify the associated brain substrates related to the long-term memory enhancing effects of emotional contextual information, and their link to recollective processes. As in our ERP studies, we specifically focused on episodic memory retrieval. In brain imaging studies, episodic retrieval is associated with greater activation in various brain regions, including the medial temporal lobe (MTL), which encompasses the hippocampus (HC) and associated parahippocampal cortices (e.g., perirhinal cortex, PrC and parahippocampal, PHC), the prefrontal cortex (PFC), including the medial and orbital areas, and the posterior parietal cortex (PPC), including the angular gyrus (AG), retrosplenial/posterior cingulate cortex (PCC) and precuneus (PCUN)/cuneus (see reviews by Davachi, 2006; Diana et al., 2007; Eichenbaum et al., 2007; Ranganath, 2010; Ranganath and Ritchey, 2012). These anatomical regions seem to have both specific and shared (or complementary) functions, including their involvement in recollection- and familiarity-based processes. Within the MTL, activity in the PrC may be crucial for familiarity-based memory, whereas activity in the PHC and HC may be necessary for recollection-based retrieval (Diana et al., 2007; Charan Ranganath and Ritchey, 2012; Michael D. Rugg and Vilberg, 2013). For instance, in a detailed review of eight studies from their own group, Rugg et al. (2012) found that HC activation consistently increased with the amount of information retrieved, both during item recognition and source recognition tasks, suggesting that HC is modulated by the quantity of retrieved information (i.e. recollection processes). Furthermore, the successful retrieval of associated source information has been not only related to HC but also to PHC activity (Düzel et al., 2003; see for review Diana et al., 2007).

Other regions beyond the MTL, such as the AG, PCC, PCUN (Vilberg and Rugg, 2007, 2009a, 2009b; see for reviews, Cabeza et al., 2008; Rugg and King, 2018; Sestieri et al., 2017; Vilberg and Rugg, 2008; Wagner et al., 2005; Wheeler and Buckner, 2004), and medial PFC (e.g., Schlichting and Preston, 2015) have also showed strong activation during retrieval, particularly when memory was based on recollection-rather than familiarity-related processes. Supporting the conjoint participation of these regions in recollection-based retrieval, it has recently been observed that functional connectivity between these areas is associated with recollection-based memory (King et al., 2015), providing evidence for the existence of the so-called recollection-sensitive network (Rugg and Vilberg, 2013).

Of note, the activation of PPC regions seems to be closely related to the aforementioned recollection-sensitive late ERP Old/New effect, as suggested by studies using combined fMRI-ERP (Hoppstädter et al., 2015; Vilberg et al., 2006) and ERP source localization analysis Weymar et al., 2010, Weymar et al., 2011). Along this line, prior brain imaging studies using immediate and long-term retrieval also suggest that —as for the ERP Old/New effect (e.g., Weymar et al., 2009, Weymar et al., 2011)— the recognition of emotional information (e.g., scenes) is associated with enhanced PPC activity (Keightley et al., 2011; V. Sterpenich et al., 2009). Furthermore, emotion-specific retrieval effects have been also observed in MTL and PFC regions (Keightley et al., 2011; Dolcos et al., 2005; Sterpenich et al., 2009; see for review LaBar and Cabeza, 2006), as part of a putative recollection network (Gilmore et al., 2015; Rugg and Vilberg, 2013). However, the scarce evidence for the emotion effects on the neural substrates underlying immediate source memory retrieval is mixed. Whereas Smith and colleagues observed that the retrieval of neutral objects encoded in emotional backgrounds elicited stronger MTL and PFC activation than objects encoded in neutral background scenes (Smith et al., 2004, Smith et al., 2005), other studies have found opposing effects. For instance, memory for neutral words integrated with other emotional words, compared to neutral ones, showed larger activation in the PPC regions, but lower in MTL regions (Murray and Kensinger, 2014). Similarly, memory for information paired with emotional cues was related to a decreased activation of the MTL regions, compared to when paired with neutral cues (Bisby et al., 2015).

Here, we investigate for the first time whether during long-term memory retrieval, in which emotional effects are more prominent (McGaugh, 2000), neutral cues previously associated with emotional arousing events engage brain areas typically associated with episodic memory (i.e. MTL, PFC, and PPC) more strongly than cues previously related to low arousing neutral events. To address this, we presented neutral objects integrated with emotional or neutral scenes and, one week later, the same encoded objects together with new ones were presented to participants, who had to indicate whether the objects were seen during encoding (item memory), and which background category each object was paired with (contextual source memory). To directly assess the contribution of recollection- and familiarity-based memory, Tulving's Remember/Know paradigm was used (Tulving, 1985). In line with prior studies (Ventura-Bort et al., 2016b, 2017, 2019), we expected that emotion would increase recollection-based memory for objects paired with emotional backgrounds, as well as for contextual source information. In accordance with the behavioral performance, we expected larger activation in regions of the recollection memory network for successfully retrieved items. As in prior imaging studies testing long-term memory of emotional scenes (i.e. item memory; Sterpenich et al., 2009; Dolcos et al., 2005), we expected to find larger activation in MTL, PFC, and PPC regions for successfully retrieved objects paired with emotional backgrounds compared to their neutral counterparts. Similarly, given that recognition memory for contextual source cues is also guided by recollection processes (Ventura-Bort et al., 2017) and has been associated with greater activation in regions of the recollection memory network (Rugg et al., 2012; Rugg and Vilberg, 2013; Rugg and King, 2018), we expected increased activation of MTL, PFC, and PPC regions during the retrieval of contextual cues, especially for emotional background scenes in comparison to their neutral counterparts.

Section snippets

Participants

A total of thirty-two healthy students (29 women, 3 men; mean age = 22.68) from the University of Greifswald participated in this study for course credits or financial compensation. All participants had normal or corrected-to-normal vision and were native German speakers. Each individual provided written informed consent for a protocol approved by the Review Board of the German Psychological Society. Retrieval data from three participants (all women) could not be analyzed due to technical

Item memory

Table 1 summarizes the mean (standard deviation) memory accuracy for item and source memory as a function of context category.

After assuming independence of redundancy (Yonelinas and Jacoby, 1995), significant differences in item memory were found, showing larger hits for Know than for Remember judgments, t (28) = 4.45, p < .001, d = 0.82 (see Table 1). This difference, however, was not found for d' indices, (d' Know vs. d' Remember: t < 1).

Discussion

In the present study, we investigated long-term item and source recognition memory for neutral objects encoded with emotional and neutral scenes. Overall, we observed similar item memory performance for emotional and neutral associates. However, when memory quality was considered, memory for emotional, compared to neutral, associates was more dependent on recollection-based memory. Similarly, correct source memory was more likely driven by recollection processes, particularly for emotional

Credit author statement

MW conceived the present idea. All authors contributed to the design. CV-B programmed the experiment and analyzed the data. All authors discussed the results and contributed to the final manuscript

Acknowledgements

This research was supported by a grant from the German Research Foundation (DFG, WE 4801/3–1) to MW. F.D. was supported by a Helen Corley Petit Scholarship in Liberal Arts and Sciences and an Emanuel Donchin Professorial Scholarship in Psychology from the University of Illinois. We are grateful to Anna Josephine Thamm, Livia Viva Verena Welsch, and Sophie Kalweit for their assistance in data collection, and to Yuta Katsumi for assistance with the analyses using the MTL mask from Dolcos Lab. We

References (126)

  • E.A. Kensinger et al.

    Effects of emotion on memory specificity: memory trade-offs elicited by negative visually arousing stimuli

    J. Mem. Lang.

    (2007)
  • C.R. Madan et al.

    Emotional arousal impairs association-memory: roles of amygdala and hippocampus

    Neuroimage

    (2017)
  • C.R. Madan et al.

    Emotional arousal does not enhance association-memory

    J. Mem. Lang.

    (2012)
  • E.J. Maratos et al.

    Neural activity associated with episodic memory for emotional context

    Neuropsychologia

    (2001)
  • X. Meng et al.

    The impact of emotion intensity on recognition memory: valence polarity matters

    Int. J. Psychophysiol.

    (2017)
  • M. Roy et al.

    Ventromedial prefrontal-subcortical systems and the generation of affective meaning

    Trends Cognit. Sci.

    (2012)
  • Michael D. Rugg et al.

    Event-related potentials and recognition memory

    Trends Cognit. Sci.

    (2007)
  • Michael D. Rugg et al.

    Ventral lateral parietal cortex and episodic memory retrieval

    Cortex

    (2018)
  • Michael D. Rugg et al.

    Item memory, context memory and the hippocampus: fMRI evidence

    Neuropsychologia

    (2012)
  • M.L. Schlichting et al.

    Memory integration: neural mechanisms and implications for behavior

    Current Opinion in Behavioral Sciences

    (2015)
  • A.P.R. Smith et al.

    fMRI correlates of the episodic retrieval of emotional contexts

    Neuroimage

    (2004)
  • J. Spaniol et al.

    Event-related fMRI studies of episodic encoding and retrieval: meta-analyses using activation likelihood estimation

    Neuropsychologia

    (2009)
  • H.C. Barron et al.

    Online evaluation of novel choices by simultaneous representation of multiple memories

    Nat. Neurosci.

    (2013)
  • J.R. Binder et al.

    Where is the semantic system? A critical review and meta-analysis of 120 functional neuroimaging studies

    Cerebr. Cortex

    (2009)
  • J.A. Bisby et al.

    Negative affect impairs associative memory but not item memory

    Learn. Mem.

    (2014)
  • J. Bisby et al.

    Opposing effects of negative emotion on item and associative memory are predicted by changes in amygdala and hippocampal activity

    Soc. Cognit. Affect Neurosci.

    (July 2015)
  • M.M. Bradley et al.

    Remembering pictures: pleasure and arousal in memory

    J. Exp. Psychol. Learn. Mem. Cognit.

    (1992)
  • Margaret M. Bradley et al.

    Imaging distributed and massed repetitions of natural scenes: spontaneous retrieval and maintenance

    Hum. Brain Mapp.

    (2015)
  • M.B. Brodeur et al.

    Bank of standardized stimuli (BOSS) phase II: 930 new normative photos

    PloS One

    (2014)
  • M.B. Brodeur et al.

    The bank of standardized stimuli (BOSS): comparison between French and English norms

    Behavior Research Methods, 1980

    (2012)
  • R.L. Buckner et al.

    The brain's default network

    Ann. N. Y. Acad. Sci.

    (2008)
  • R. Cabeza et al.

    The parietal cortex and episodic memory: an attentional account

    Nature Reviews Neuroscience

    (2008)
  • G.A. Carlesimo et al.

    Recollection and familiarity in the human thalamus

    Neuroscience and Biobehavioral Reviews

    (2015)
  • A.E. Cavanna et al.

    The precuneus: a review of its functional anatomy and behavioural correlates

    Brain

    (2006)
  • Y.-C. Chiu et al.

    On opposing effects of emotion on contextual or relational memory

    Front. Psychol.

    (2013)
  • E. Ciaramelli et al.

    Subjective recollection independent from multifeatural context retrieval following damage to the posterior parietal cortex

    Cortex

    (2017)
  • E. Ciaramelli et al.

    Top-down and bottom-up attention to memory: a hypothesis (AtoM) on the role of the posterior parietal cortex in memory retrieval

  • X.D.V. Clewett et al.

    Locus coeruleus activity strengthens prioritized memories under arousal

  • T. Curran

    Brain potentials of recollection and familiarity

    (2000)
  • A. D'Argembeau et al.

    Influence of emotion on memory for temporal information

    Emotion

    (2005)
  • A.R. Damasio

    Time-locked multiregional retroactivation: a systems-level proposal for the neural substrates of recall and recognition

    Cognition

    (1989)
  • J. Decety et al.

    The Role of the Right Temporoparietal Junction in Social Interaction: How Low-Level Computational Processes Contribute to Meta-Cognition

    (2007)
  • S. Dehaene et al.

    Three parietal circuits for number processing

  • S. Doerksen et al.

    Source memory enhancement for emotional words

    Emotion

    (2001)
  • F. Dolcos

    Linking enhancing and impairing effects of emotion—the case of PTSD

    Front. Integr. Neurosci.

    (2013)
  • F. Dolcos et al.

    Neural correlates of emotion-attention interactions: From perception, learning, and memory to social cognition, individual differences, and training interventions

  • F. Dolcos et al.

    Emerging directions in emotional episodic memory

    Front. Psychol.

    (2017)
  • F. Dolcos et al.

    Remembering one year later: role of the amygdala and the medial temporal lobe memory system in retrieving emotional memories

    Proc. Natl. Acad. Sci. U.S.A.

    (2005)
  • H. Eichenbaum et al.

    The medial temporal lobe and recognition memory

    Annu. Rev. Neurosci.

    (2007)
  • P. Ferré et al.

    Emotional content and source memory for language: impairment in an incidental encoding task

    Front. Psychol.

    (2019)
  • Cited by (0)

    View full text