Berntsen, D. & Rubin, D. (2008). The reappearance hypothesis revisited: Recurrent involuntary memories after traumatic events and in everyday life. Memory & Cognition, 36, 449-460.

 

Background

This study is concerned with recurrent involuntary autobiographical memories, which the authors define as repetitive memories of personal experience that come to an individual’s mind without an attempt being made to think about them.  Typically, this is an area focused on in a clinical setting with regard to Posttraumatic Stress Disorder (PTSD); however, previously it has not been studied in nonclinical populations.  This brings up the question of are these memories specifically tied to traumatic or negative emotional events. 

 

There is also the question of what exactly is meant by the word recurrent.  Typical literature views memories as reconstructive, which basically refers to the idea that every time you remember something it is recreated, and is not identical to the initial event.  This is opposite to claims in PTSD literature that says these memories are fixed and repeat themselves because of their traumatic/emotional nature.  To answer these questions the authors performed three studies.  The first study examined individuals that experienced the same traumatic event to check for the presence of Recurrent Involuntary Memories (RIMs).  The second study, examined the general population for instances of RIMs.  The third study examined whether the RIMs were exact copies of previous memories or if they differed each time.

 

In order to study these things, the authors created a basic framework for RIMs, which is based highly on the framework of normal autobiographical memories.  According to the authors, at a basic level these types of memories occur when they are highly accessible in memory.  Therefore, these memories are rather common and should not occur in just clinical populations.  Certain things will make it more likely for a RIM to occur.  These things include more recent events, emotionally intense events, emotionally positive events, and positive events from young adulthood.  Based on this, recent traumatic events should be likely to result in RIMs due to their recent nature and highly emotional content.  Furthermore, individuals with the highest emotional response and who are closer to the traumatic event will suffer from more RIMs than individuals farther away.  Also within this framework, the authors referred to the concept that an individual’s memory of the same event varies over time.  As a result, they assert that while the overall event in a RIM is the same, the details will be different each time.

 

STUDY 1

            The goal of this study was to examine the incidence of RIMs among individuals who had experienced the same recent traumatic event.  It was predicted that participants who experienced more RIMs would report higher levels of fear during the event.  Furthermore, the RIMs would refer to those memories that caused the most arousal, and would vary based on the participant’s proximity to the event.

 

Methods

             The authors mailed questionnaires to Danish tourists who had survived the tsunami in Thailand and Sri Lanka in 2004.  The questionnaires included questions about proximity to the wave, trauma criteria from the DSM for trauma (A1 and A2), the Centrality of Event scale (CES) which measures how central the traumatic event was to the course of their lives, the Posttraumatic Stress Disorder checklist (PCL) which measures PTSD symptoms, overall memory of the event, and incidence of recurrent memories.  They also requested open ended recollections of both recurrent memories and the individual’s clearest memory of the event.

 

Results

            The authors found that 51% of the participants had experienced or witnessed some sort of life danger or serious injury (A1) and 47% of the participants had experienced intense fear, helplessness, or horror (A2).  Also, 32% had acknowledged both of these which meets the DSM criteria for a traumatic event.  Furthermore, 9.5% of the participants obtained a score in the critical range on the PCL.  As was hypothesized, proximity to the wave was positively correlated with their scores.  The authors also found that 41% of the participants acknowledged RIMs and those that had recurrent memories scored higher on PTSD symptoms.  A relationship was also found between proximity to the wave and incidence of RIMs, where the closer to the wave the individual was, the more likely they were to have had a RIM.

            With regard to the description of recurrent memories, all were scored as emotionally negative.  The most common categories for memories were memories of escaping, when the wave hits, and searching for loved ones.  According to the authors, the recurrent memory was related to whatever the individual experienced as their most emotionally intense moment.  So for those that were most directly threatened by the wave, their most common memory involved escaping, while those that were not directly threatened remembered most when the wave hit, and those who heard about the wave from a safe distance mostly remembered the worst case scenarios.

           

Discussion

            The results supported the hypothesis.  Participants who reported having RIMs experienced higher levels of PTSD symptoms and exposure to tragic events.  Furthermore, the content of the RIMs coincided with the proximity to the event and tended to be about the most emotional moments.  This supported both the model the authors gave as well as the standard PTSD theories.  On a somewhat side note, the results did not seem to support warning signal theory which refers to the idea that the stimuli directly proceeding a traumatic event had the most emotional impact.

 

STUDY 2

            The second study focused on RIMs in the general population rather than just in a population that had experienced a major trauma.

 

Methods

            They contacted a considerable amount of people in Demark and asked them apparently 6 questions.  The questions are centered around whether they experience RIMs or recurrent dreams.

 

Results

            The researchers found that 53% of the participants had RIMs and 29% had recurrent dreams.  There was a slight decrease in RIMs with age, and a larger decrease in recurrent dreams.  They also found that most RIMs were on the positive side, with few being rated as negative, and more often than not they tended to be intense rather than not intense.

 

Discussion

            The results were consistent with the initial framework given by the authors.  This is contrary to the concept in the clinical realm that RIMs are a reaction to traumatic/negative experiences.

 

STUDY 3

            The third study used a prospective method to determine if the RIMs were identical to each other, or just related to the event’s main idea but with different details.  Typically, clinical research points to the RIMs as having the same exact details each time, however, the authors predict that the details will be different.

 

Methods

            Participants were nine individuals who had created diaries of their RIMs in a previous study by Berntsen.  They were asked to record RIMs and explain how it related to their most traumatic event.  The events included assaults, diseases, and death of a friend.  All of the participants met criteria for PTSD.

 

Results and Discussion

            The researchers found that the majority of RIMs were not duplicates and referred to different details and parts of the event.  Also of interest, the one case where the memories were identical, the individual was experiencing the same cue as when she experienced the trauma.  This is contrary to the theories of clinical research.

 

Overall Discussion

            Most clinical theories focus on RIMs as being based on highly distressing or traumatic events.  However, the study showed that RIMs are not based solely on negative events, but can rather be explained by the same theories as normal autobiographical memory.  In the first study, the memories referred to the most emotionally arousing parts of the event which tended to be the most traumatic.  This is consistent with clinical views; however, the authors explained it more in terms of emotional intensity and recency based on the findings of the second study.  The second study found that RIMs among the general population had the same qualities as normal autobiographical memory, which counteracts the typical clinical views.  The third study expanded on this difference from clinical views by showing the RIMs to differ in detail for the most part about the same event.  As a result, according to the authors, a better way to account for PTSD memories would be through normal autobiographical memory research than just focusing on traumatic memories