It is commonly assumed that, with time, an initially labile memory is transformed into a permanent one via a process of consolidation. Yet, recent evidence indicates that memories can return to a fragile state again when reactivated, requiring a period of reconsolidation. In the study described here, we found that participants who memorized a story immediately after they had recalled neutral and emotional experiences from their past were impaired in their memory for the neutral (but not for the emotional) experiences one week later. The effect of learning the story depended critically on the preceding reactivation of the autobiographical memories since learning without reactivation had no effect. These results suggest that new learning impedes the reconsolidation of neutral autobiographical memories.
Citation: Schwabe L, Wolf OT (2009) New Episodic Learning Interferes with the Reconsolidation of Autobiographical Memories. PLoS ONE 4(10): e7519. doi:10.1371/journal.pone.0007519
Editor: Thomas Burne, University of Queensland, Australia
Received: August 20, 2009; Accepted: September 30, 2009; Published: October 21, 2009
Copyright: © 2009 Schwabe, Wolf. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: The authors have no support or funding to report.
Competing interests: The authors have declared that no competing interests exist.
Memories are built in stages. Initially, novel information is acquired and retained for a short period. This transient short-term memory is transformed into a lasting memory via a consolidation process depending on protein synthesis . For decades, it has been commonly assumed that consolidated memories are not subject to further modification. However, this view, has been challenged by animal studies suggesting that memories return to a fragile state when reactivated, making them susceptible to the same manipulations as the original consolidation process –. A time-limited process of reconsolidation appears to be necessary to render reactivated memories stable again , . Corroborating the animal studies, recent studies in humans show that memories are altered when a beta-blocker or a new learning task is given after their reactivation –.
Manipulating memory reconsolidation provides a unique opportunity to change unwanted, e.g. traumatic, memories in a favorable manner. Indeed, there is first evidence that the administration of a beta-blocker after the reactivation of traumatic memories may reduce emotional responding to the traumatic event in post traumatic stress disorder . Interestingly, a recent rodent study suggested comparable effects after a drug-free intervention, namely, learning new information after memory reactivation . Although there are human studies demonstrating altered memory when new material is learned after reactivation, these studies focused only on procedural memory , conditioning ,  or the memory for a list of items learned in the laboratory , . Whether these findings can be translated to memories for events people experienced in their “real” (i.e. every day) life is an important, but yet unanswered, question in reconsolidation research and its potential application in clinical, educational or legal settings.
Thus, the aim of the present experiment was to examine whether autobiographical memories can be modified by learning new episodic material after memories have been reactivated. Participants memorized an Indian folk tale (“The War of the Ghosts” by Bartlett ; interference task) after they had recalled neutral and emotional experiences from their past. In order to rule out non-specific effects of learning the tale, another group of participants learned the story without prior reactivation of autobiographical memories . A third group recalled the personal experiences but did not learn the story afterwards; a fourth group did neither reactivate the autobiographical memories, nor learn the story. We hypothesized that learning new information after the reactivation of autobiographical memories would reduce the richness of these memories.
Participants and general procedure
Ninety-six healthy young adults (48 men, 48 women; age: M = 23.5 years, SEM = 0.3 years), all students of the Ruhr-University Bochum, received either course credits or a financial reward (15€) for participation in this experiment.
We used a 2 (reactivation vs. no reactivation) ×2 (interference vs. no interference) factorial between-subject design resulting in four experimental conditions: reactivation+interference, reactivation only, interference only, neither reactivation nor interference (control group). Groups differed mainly with respect to the experimental manipulation on day 1. Participants in the reactivation+interference group completed an autobiographical memory test (see below) and then learned an unfamiliar story. Participants in the reactivation only group completed the autobiographical memory test but did not learn the unfamiliar story. In contrast, participants in the interference only group did not perform the autobiographical memory test but learned only the unfamiliar story. Participants in the control group did not come to the laboratory on day 1.
One week later, participants in the reactivation+interference and reactivation only groups recalled the autobiographical events they had described the previous week. The interference only and control groups completed the autobiographical memory test as did the other two groups the week before. We randomly assigned 12 men and 12 women to each of the four groups.
Autobiographical memories were reactivated by means of the autobiographical memory cueing test . Participants were presented 2 positive (happy, interested), 2 neutral (busy, concentrated) and 2 negative (sad, angry) adjectives in randomized order (see ). They were instructed to remember, in as much detail as possible, one specific event from their own past for each adjective. Events should have occurred at least 24 hours and at maximum 2 weeks in the past. Participants were told that solely the specificity of the event was important, not its content. There was a time limit of 4 minutes for each of the 6 adjectives. After participants had written down the events, they were asked to indicate when each event had occurred and to give each memory a title (which should help to refer to the events on day 2).
In retrospect, participants described events that were negative, positive or neutral, according to the presented adjective. Examples of negative events were conflict with a friend or the death of a beloved person; examples of positive events were a nice party with friends or a successful exam; examples of neutral events were a certain lecture or the preparation for a test.
Immediately after the completion of the autobiographical memory test, participants of the reactivation+interference group read Bartlett's  “War of the Ghosts”, a complex story that was most likely to be clearly distinct from the participants' remembered events; participants in the interference only group read the story immediately after their arrival at the laboratory. They were given five minutes to memorize the story. Memory for the story was tested at the end of the experiment on day 2.
The critical memory test took place one week after day 1. Participants in the reactivation+interference and reactivation only groups were presented the titles of the events they had described the week before and asked to remember as many details as possible of the referring event. In the interference only and control groups, participants completed the autobiographical memory test as did the other groups on day 1, except that they were instructed to remember events that were at least 1 week and at the most 3 weeks old. Again, participants had 4 minutes to describe the event associated with the title and adjective, respectively.
The autobiographical memories were assessed by two independent raters. One point was given for each detail (time, place, involved persons, weather, thoughts etc.) that was mentioned. The agreement between the two raters was very high (interrater reliability ricc = 0.94). Discrepancies were discussed until an agreement was reached.
Points were first summed up for each event and then averaged for the positive, neutral and negative events.
To control for possible mood-congruent or mood-dependent memory effects , , we asked participants to complete the MDBF, a German multidimensional mood scale , at the beginning of the two experimental sessions (participants in the control group completed the MDBF on day 2 only). This questionnaire measures three dimensions of subjective feeling (“elevated vs. depressed mood”, “wakefulness vs. sleepiness”, “calmness vs. restlessness”) on a 5-point rating scale ranging from “not at all” ( = 1) to “very much” ( = 5).
Age of the autobiographical memories
The age of the described memories was comparable in the four groups, F(3,92) = 1.63, p = .19, η2 = 0.05. Events occurred, on average, 13.6 days (SEM = 0.5 days) before experimental day 2. Memories were narrated from the first person perspective suggesting that these reflected experiences of participants' own past.
Effect of the interference task on the reconsolidation of autobiographical memories
Participants in the reactivation only and reactivation+interference groups did not differ in their memories on day 1 (12.8 vs. 12.4 details per event), F(1,46) = 0.10, p = .75, η2<0.01.
Learning “The War of the Ghosts” immediately after the reactivation of the autobiographical memories reduced memory for the neutral but not for the emotional experiences one week later. As shown in Figure 1, participants in the reactivation+interference group remembered significantly less details of the neutral events than participants of the reactivation only, interference only or control groups whereas the memory for the positive and negative events remained unaffected by the interference task after reactivation. This is supported by a reactivation (yes vs. no) × interference (yes vs. no) × emotionality of the events (positive vs. neutral vs. negative) × sex (men vs. women) ANOVA which yielded a significant three-way interaction between reactivation, interference and emotionality of the events, F(2,174) = 5.02, p<.01, η2 = 0.06. Follow-up ANOVAs indicated a significant reactivation × interference interaction for neutral, F(1,95) = 6.03, p<.02, η2 = 0.06, but not for positive or negative events, both F(1,95)<1.36, both p>.25, both η2<0.02. Bonferroni-corrected post-hoc tests showed that memory for neutral events was significantly poorer in the reactivation+interference group relative to the other three groups, all p<.01.
Neutral but not positive or negative autobiographical memories were reduced when participants learned new information after memory reactivation (* p<.01). Note that memories were about 2 weeks old when the memory test took place. Data represent means ± SEM.
Overall, women remembered more details than men, F(1,88) = 4.50, p<.04, η2 = 0.05. Moreover, memories tended to be more detailed for positive than for negative and neutral experiences, F(2,182) = 2.95, p<.06, η2 = 0.03. None of the other main or interaction effects reached statistical significance. The observed effect of the interference task after memory reactivation was not memory type specific (e.g. time, space, involved persons).
Memory for the interference task
Participants in the interference only and reactivation+interference groups performed equally well in the memory test for “The War of the Ghosts”, t(46) = 1.11, p = .27. On average, they remembered 13.7 (SEM = 0.8) facts of the story.
Mood effects cannot explain our results because groups had comparable MDBF scores on both experimental days, all F<1, all p>.46, all η2<0.03.
Retroactive interference by new learning shortly after the original learning is a well known phenomenon , . Here, we show for the first time that new learning after the reactivation of personal experiences impairs the subsequent memory for these experiences, suggesting reconsolidation blockade in autobiographical memory.
Previous studies that demonstrated interference effects in human memory reconsolidation included solely neutral material –. In the present experiment, we examined memory for both neutral and emotional experiences and found reconsolidation effects for neutral but not for emotional material. How can the differential sensitivity of reactivated neutral and emotional memories to new learning be explained? Imagining positive or negative events is a common method to induce positive and negative mood states, respectively . Thus, it appears likely that the retrieval of emotional memories altered the mood of the subjects. Neuromodulators, such as noradrenaline (which is known to play a crucial role in memory reconsolidation ), that were released during emotional memory retrieval may have affected the reconsolidation of neutral memories in particular, making the reconsolidation of these memories especially susceptible to the interfering influence of the rather disturbing “War of the Ghosts” story. Future studies should take this possibility into account and include subjective and objective (e.g. heart rate, salivary alpha amylase) measures of the arousal associated with the retrieval of the memories. Another potential explanation takes the strength of emotional memories into account. Emotionally arousing experiences are usually very well remembered . This is because they activate brain structures (in particular the amygdala) and elicit the release of numerous hormones and neurotransmitters (e.g. adrenaline and noradrenaline) that facilitate memory consolidation . Thus, emotional events were most likely better consolidated and therefore less sensitive to reconsolidation effects which are modulated by the strength of the memory . While beta blockers are potent enough to disrupt the reconsolidation of emotional and traumatic memories , , this does not seem to be the case for learning of new information. Our findings suggest that reconsolidation-based approaches to “erasing” previous emotional or traumatic memories that rely solely on cognitive interference might not work because only neutral memories would be affected.
In addition to the emotional tone of the memories, reconsolidation effects depend on several other parameters. One important factor appears to be the age of the memories. In this study, memories were about 1 week old when they were reactivated. There is some evidence that memories become less vulnerable to reconsolidation effects the older they are , . Hence, it can be speculated that memories from our more distant past are more likely to resist interfering influences when reactivated. Another factor relevant for memory reconsolidation is the reactivation context. It has recently been argued that new learning has to occur in the same spatial context as the original learning to modify the original memory trace . In this study, however, memories were only 2 days old and there is evidence that memories become less context-dependent with age . Our results suggest that travelling mentally back in time to recollect specific past episodes may be sufficient to render (autobiographical) memories susceptible to interference by new learning.
Memory reconsolidation has been viewed as a mechanism by which memories can be updated, either by the modification of the reactivated trace  or by the creation of a new version of the original trace . Such an update, however, requires some degree of similarity between the original and the novel learning episode , . In the present experiment, there was no obvious similarity between the recalled personal events and the Indian folk tale. There was also no indication of an updating of the autobiographical memories; none of the participants incorporated details of the folk tale into the recall of their own experiences. Our data suggest that memorizing the complex novel story claimed a considerable part of participants' limited cognitive processing capacities, leaving less cognitive resources for the reconsolidation of the reactivated memory. While the few previous studies that demonstrated reconsolidation effects in episodic memory ,  showed that new learning did not reduce the number of correctly recalled items but led to a mixture of old and new items, the present findings provide the first evidence that reactivated autobiographical memories can degrade when they are not fully reconsolidated.
While reconsolidation is generally seen as an adaptive mechanism enabling the updating of reactivated memories in the light of new experiences , the results presented here point to its potential side effect: They suggest that reactivated autobiographical memories are, similar to new memories, in danger of being lost.
We thank Melanie Führer, Florentina Flack and Mathias Winnicki for their help during data collection.
Conceived and designed the experiments: LS OTW. Performed the experiments: LS. Analyzed the data: LS. Wrote the paper: LS OTW.
- 1. Goelet P, Castellucci VF, Schacher S, Kandel ER (1986) The long and the short of long-term memory - a molecular framework. Nature 322: 419–422.
- 2. Nader K, Schafe GE, LeDoux JE (2000) Fear memories require protein synthesis in the amygdala for reconsolidation after retrieval. Nature 406: 722–726.
- 3. Przybyslawski J, Roullet P, Sara SJ (1999) Attenuation of emotional and non-emotional memories after their reactivation: role of beta adrenergic receptors. J Neurosci 19: 6623–6628.
- 4. Debiec J, LeDoux JE, Nader K (2002) Cellular and systems reconsolidation in the hippocampus. Neuron 36: 527–538.
- 5. Sara SJ (2000) Retrieval and reconsolidation: toward a neurobiology of remembering. Learn Mem 7: 73–84.
- 6. Nader K, Hardt O (2009) A single standard for memory: the case for reconsolidation. Nat Rev Neurosci 10: 224–234.
- 7. Kindt M, Soeter M, Vervliet B (2009) Beyond extinction: erasing human fear responses and preventing the return of fear. Nat Neurosci.
- 8. Hupbach A, Hardt O, Gomez R, Nadel L (2008) The dynamics of memory: context-dependent updating. Learn Mem 15: 574–579.
- 9. Hupbach A, Gomez R, Hardt O, Nadel L (2007) Reconsolidation of episodic memories: A subtle reminder triggers integration of new information. Learn Mem 14: 47–53.
- 10. Walker MP, Brakefield T, Hobson JA, Stickgold R (2003) Dissociable stages of human memory consolidation and reconsolidation. Nature 425: 616–620.
- 11. Galluccio L (2005) Updating reactivated memories in infancy: I. Passive- and active-exposure effects. Dev Psychobiol 47: 1–17.
- 12. Brunet A, Orr SP, Tremblay J, Robertson K, Nader K, et al. (2008) Effect of post-retrieval propranolol on psychophysiologic responding during subsequent script-driven traumatic imagery in post-traumatic stress disorder. J Psychiatr Res 42: 503–506.
- 13. Monfils MH, Cowansage KK, Klann E, LeDoux JE (2009) Extinction-reconsolidation boundaries: key to persistent attentuation of fear memories. Science 324: 951–955.
- 14. Bartlett F (1932) Remembering: A study in experimental and social psychology. Cambridge: Cambridge University Press.
- 15. Nader K (2003) Neuroscience: re-recording human memories. Nature 425: 571–572.
- 16. Williams JMG, Broadbent K (1986) Autobiographical memory in suicide attempters. J Abnorm Psychol 95: 144–149.
- 17. Buss C, Wolf OT, Witt J, Hellhammer D (2004) Autobiographic memory impairment following acute cortisol administration. Psychoneuroendocrinology 29: 1093–1096.
- 18. Blaney PH (1986) Affect and memory: A review. Psychol Bull 99: 229–246.
- 19. Eich E (1995) Searching for mood dependent memory. Psychol Sci 6: 67–75.
- 20. Steyer R, Schwenkmezger P, Notz P, Eid M (1994) Testtheoretische Analysen des Mehrdimensionalen Befindlichkeitsfragebogens (MDBF). Diagnostica 40: 320–328.
- 21. Postman L, Underwood BJ (1973) Critical issues in interference theory. Mem Cogn 1: 19–40.
- 22. Bäuml KH (1996) Revisiting an old issue: Retroactive interference as a function of the degree of original and interpolated learning. Psychonom Bull Rev 3: 380–384.
- 23. Westermann R, Spies K, Stahl G, Hesse FW (1998) Relative effectiveness and valididty of mood induction procedures: a meta-analysis. Eur J Soc Psychol 26: 557–580.
- 24. McGaugh J (2000) Memory-a century of consolidation. Science 287: 248–251.
- 25. Roozendaal B, McEwen BS, Chattarji S (2009) Stress, memory and the amygdala. Nat Rev Neurosci 10: 423–433.
- 26. Suzuki A, Josselyn SA, Frankland PW, Masushige S, Sliva AJ, et al. (2004) Memory reconsolidation and extinction have distinct temporal and biochemical signatures. J Neurosci 24: 4787–4795.
- 27. Milekic M, Alberini CM (2002) Temporally graded requirement for protein synthesis following memory reactivation. Neuron 36: 521–525.
- 28. Wiltgen BJ, Silva AJ (2007) Memories for context become less specific with time. Learn Mem 14: 313–317.
- 29. Nader K (2003) Memory traces unbound. Trends Neurosci 26: 65–72.
- 30. Moscovitch M, Nadel L (1998) Consolidation and the hippocampal complex revisited: in defense of the multiple-trace model. Curr Op Neurobiol 8: 297–300.
- 31. Marsh RL, Landau JD, Hicks JL, Bink ML (1998) On reducing retroactive interference. Am J Psychol 111: 175–190.