and lower beneficial influence on negative emotional memory than on
neutral memory.
Z. Tuinman
Zoë Tuinman
Date: February 2020 – June 2020 Deadline: 19 June 2020
Student number: 11600616
Abstract
While a lot is already known about the influence of sleep on memory, a debate in the literature remains about sleep’s influence on negative emotional memory. Specifically, whether this influence is greater, lesser or equal to the influence of sleep on neutral memories. To explore this, the current study looked at the difference in influence of sleep on neutral and negative emotional memory. Two aspects of memory were looked at, emotional tone and memory performance. This was done by showing participants (N=38) images before sleep that were then associated with either a neutral or a negative video. The memory for these images was tested both before and after sleep. Additionally, a valence rating of the images was obtained before and after sleep. During the experimental night, the participants were regularly woken up to collect their dreams. The results showed that the valence ratings of the images, though different pre-sleep, were similar for neutral and negatively associated items post-sleep. After sleep, both returned to neutral. The memory tests showed that memory performance for neutral and negative memory items were similar before sleep. After sleep, memory performance for both was higher than before sleep with the performance of neutral memory items being higher than that of negative items. These results suggests that sleep causes a loss of emotional tone, while also showing that sleep’s beneficial influence on memory performance is greater for neutral than for negative memories.
Keywords: sleep, neutral memory, emotional memory, emotional tone, memory performance
Introduction
Approximately, 7% of the world’s population will suffer from post-traumatic stress disorder (PTSD) at some point in their lives (Compean & Hamner, 2019). Additionally, of all the traumatized youth, roughly 8% will develop PTSD by the age of 18 (Herringa, 2017). When someone has either witnessed or experienced a traumatic event, PTSD can occur. Examples of such events include a natural disaster, a serious accident, a terrorist act, war or rape. The most common symptom of PTSD is having intense and disturbing thoughts and feelings related to the experience that lasts long after the actual traumatic event (Amerian Psyciatric Association, 2013). For example, patients may experience flashbacks and/or nightmares. Thus, PTSD patients have a problem with memory, specifically emotional memory.
An organism’s memory is crucial for selecting and improving behavior in an ever-changing environment (Rasch & Born, 2013). Memories are formed by a three-step process: encoding, consolidation and retrieval. First, an experience or stimulus forms a new memory trace, encoding, which is then stabilized during consolidation. Lastly, the organism can recall the experience or stimulus, which is called retrieval. It is assumed that when awake, the brain is optimized for encoding new information and memory retrieval, while during sleep the brain is optimal for consolidation (Rasch & Born, 2013). Therefore, one already heavily researched aspect of memory is the effect of sleep on memory. Multiple studies have shown that sleep is implicated in memory encoding and consolidation (Tempesta et al., 2016; Born & Wilhelm, 2012). Various studies using a sleep deprivation paradigm with a recognition task show that sleep after encoding results in enhanced memory performance as compared to wake. This benefit has been seen for declarative, procedural and episodic memory (van der Helm, Gujar et al., 2011). Additionally, in one study, participants had to memorize pictures with different valences and were given a recognition task after total sleep deprivation or a normal sleep night. This study showed that sleep deprivation resulted in an impaired ability to encode pictures (Kaida et al., 2015). One possible theory behind this beneficial role of sleep states that memories get reactivated during sleep and is called the two-stage model of memory trace formation. This reactivation results in the redistribution of the memory representation from intermediate-term storage in the hippocampus to long-term storage in the neocortex (Buzsáki, 1989; for a review see Diekelmann & Born, 2010). It is assumed that this reactivation prevents the decay of memory (Schäfer et al., 2019). Hence, these papers provide recent evidence on the idea that sleep and memory are connected.
However, less is known about sleep’s influence on emotional memory. Emotional memory consists of two aspects: emotional tone and memory performance. With regards to emotional tone, multiple studies have shown that after sleep the emotional tone of a memory is lost (Zohar et al., 2005; van der Helm, Yao et al., 2011). The emotional tone is the feeling one gets during a certain experience (Merriam-Webster, n.d.). A relevant theory related to emotional tone is the “sleep to remember, sleep to forget” model. This model states that the unique neurobiological state of sleep could lead to the de-potentiation of the emotional tone and simultaneously to the strengthening of a memory (Goldstein & Walker, 2014). Specifically, the activation of the amygdala-hippocampal network during rapid-eye movement (REM) sleep could promote long-term retention of the informational, salient aspects of the emotional experience. At the same time, during REM there is a suppression of adrenergic activity and this could decouple the emotional tone of the emotional experience (Tempesta, et al., 2018). Previously, sleep deprivation has been reported to magnify negative reaction to adverse experiences (Zohar, et al., 2005). Similarly, van der Helm and colleagues (2011) found that after sleep, and then specifically REM sleep, a decrease in amygdala activity was seen, which is associated with re-established prefrontal connectivity. These changes
went hand in hand with an overnight reduction in subjective reactivity (van der Helm, Yao, et al., 2011).
With regards to emotional memory performance, there is some controversy about whether sleep’s influence on emotional memory performance, specifically for negative emotional memories, differs from that on neutral memory performance. Some studies show that there is no difference between the influence of sleep on the two (Atienza & Cantero, 2008; Rihm & Rasch, 2015). Evidence for this is provided by Atienza and Cantero (2008), who showed that the same behavioral impairment was seen after sleep deprivation for both emotional and neutral image retrieval. This was done by observing the participants’ memory for pictures one week after either forty hours of sleep deprivation or normal sleep. Thus, this effect of sleep on memory consolidation was independent of valence of the memory. Another experiment added to this by using Pavlovian delay conditioning (Rihm & Rasch, 2015). Here, a neutral stimulus is associated with an aversive experience (US) because of the close temporal proximity of the two. This results in the, previously, neutral stimuli (now CS) bringing forth the response normally brought forth by the aversive experience (CR) (Bottary et al., 2020). In this study auditory fear conditioning was done with an unpleasant odor being the US. The researchers found no difference between recall performance in the paired and unpaired condition after sleep (Rihm & Rasch, 2015).
Meanwhile, other studies show that the effect of sleep on emotional memories differs from that on neutral memories (Hu et al., 2006; Groch et al., 2013; Payne et al., 2008; Hutchison & Rathore, 2015; Schäfer et al., 2019). The controversy is the direction of this effect. Some studies find that sleep has an extra beneficial role on emotional memories as compared to neutral memories. This has been demonstrated by Hu and colleagues (2006) who compared memory for negative and neutrally arousing pictures for a sleep as well as an awake group. The results showed that negative arousing pictures caused an improvement in memory accuracy for sleep more than awake subjects. Also, after sleep they found that the negative arousing pictures were better recognized than the neutrally arousing pictures (Hu et al., 2006). Especially REM-rich sleep enhanced consolidation of emotional memories better than neutral stimuli (Groch et al, 2013). This extra benefit has not just been shown for nocturnal sleep, but also for naps (Payne et al., 2008). A proposed theory to support this evidence is that co-activation of the hippocampus and the amygdala, a region heavily involved in emotion, is the underlying factor that explains the influence of emotion on long-term memory retention (Hutchison & Rathore, 2015). This theory states that the amygdala modulates hippocampal activity, resulting in the facilitation of preferential encoding of emotional memories and potentially the tagging for future consolidation (Hutchison & Rathore, 2015). fMRI studies show that during REM sleep the activation of the amygdala, hippocampus, striatum, medial prefrontal cortex and insula are increased. All of whom are strongly associated with emotional processing during wake (Hutchison & Rathore, 2015).
By contrast, there is also evidence that sleep may have a less strong effect on emotional memory, instead an extra beneficial effect was seen for neutral memories. A meta-analysis of multiple other studies concluded that neutral memory was better than emotional memory performance after both wakefulness and sleep (Schäfer, et al., 2019). This effect was seen when the authors included their estimates of a possible wake group when one was not included in the study. Yet when these estimates were not included, no significant results were obtained.
Six of the aforementioned studies use a negative emotional stimulus for the emotional condition, pictures or smells that are negative on their own. Another method for inducing a negative emotional memory is fear conditioning. To create comparable conditions, most studies include some stimuli that are paired with an unpleasant stimulus (CS+) while others that are not
(CS-). Fear conditioning is used to induce a negative emotional condition on some stimuli that were previously neutral. A fMRI study found that the fear conditioned response was strengthened after sleep, both the physiological and behavioral aspect, in humans (Menz et al., 2013). This meant an enhanced activation of the basolateral amygdala when recall was improved for the sleep group. These findings were accompanied by a correlation between REM sleep duration and the consolidation strength of the fear memory (Menz et al., 2013).
The controversy on the effect of sleep on emotional memory performance when compared to neutral memory performance is not yet fully resolved. That is why the current study looks at the question: What is the influence of sleep on negative emotional memory as compared to the influence of sleep on neutral memory? The hypothesis is that during sleep the emotional tone of the negative memory is reduced and that sleep will have an even greater beneficial effect on negative emotional memory performance than on the memory of neutral memories, leading to a better memory for negative emotional events. There is both physiological and behavioral evidence for this hypothesis while this is not the case for the idea of an equal or less strong effect of sleep on emotional memory performance when compared to neutral memory performance. Also, an even greater beneficial effect of sleep on negative emotional memory is seen more often in studies using fear conditioning. To test the hypothesis, participants are shown fragments of a video with a negative emotional tone or a neutral emotional tone. These videos are associated with images of which the memory, the ability to recognize that they have been shown before, will be tested later. During the night, the participants will be woken up and their dreams will be collected.
If sleep does cause emotional tone to get lost, the valence ratings of the images will be predominantly neutral for the negative memory and will stay around neutral rating for the neutral memories. Also, if sleep has a greater beneficial effect on negative emotional memory than neutral memories, a difference will be seen between the memory performance pre-sleep and post-sleep for both, with a higher memory performance score post-sleep for both neutral and negative memory items. Furthermore, the images that were preceded by the negative video will have a bigger difference between pre- and post-sleep memory performance than those preceded by the neutral. Thus, the improved memory performance between pre- and post-sleep will be larger for the negative memories than for the neutral ones in the direction of better memory performance.
Materials and methods
Participants
A total of 38 healthy psychology students (26 identified as females, 11 as males, 1 as other) were recruited from the university of Amsterdam. The median age of the participants was 20 years. After approval from the ethics committee had been obtained, an announcement had been placed on the university’s lab website. All participants were screened and did not suffer from any diagnosed sleep disorders, neurological disorders, psychiatric disorders, psychological disorders, cardiac problems or seizures. Additionally, the participants were proficient in English, had normal or corrected to normal vision and hearing and were not using any heavy medication. Furthermore, the participants were classified as good sleepers, meaning that they had no trouble falling asleep (< 45 min), did not work night shifts, were woken up by alarms, claimed they had better or equal to fair sleep quality, fell asleep around 23:00 hours and woke up around 8:00 hours.
Materials
For this study two three-minute fragments of two movies were used to trigger an emotional association with the images. One fragment came from the movie “March of the penguins” (Luc
Jacquet). This fragment was used as for the neutral condition (Talamini et al., 2013). The other fragment came from the movie “American History X” (Tony Kaye) and was used for the negative emotional condition (Schaefer et al., 2010). From these movies both a three-minute as well as a three-second fragment were taken. The videos functioned as the US and were used to create a neutral or a negative emotional memory.
For the tone-movie association two tones were used with different frequencies. One tone was high with a frequency of 1500 Hz. The other tone was a lower tone of 1000 Hz. These tones functioned as the CS+ and CS-.
The images used in this experiment were drawings of butterflies. These images were taken from the book ‘Uitlandsche kapellen’ by P. Cramer (Cramer & Stoll, 1779). First, 180 images were chosen and these were tested on arousal and valence. Next,120 butterfly images were selected with similar valence and arousal ratings. From these, 20 were used for the neutral condition, 20 for the negative emotion condition and the rest was used as foils in the memory tests (for examples see figure 1).
The online program Qualtrics XM was used to build the encoding task. Similarly, this program was used to collect the participants’ responses from multiple questionnaires, a screening, a consent form and both memory tests. A couple of self-report questionnaires were used to test for type of affect, anxiety, depression and sleep quality. These questionnaires consisted of Positive and Negative Affect Schedule (PANAS), Beck Depression Inventory (BDI-II), the Pittsburgh Sleep Quality Index (PSQI) and State-Trait Anxiety Inventory (STAI).
Figure 1: Examples of images of butterflies used in this experiment per condition. Design
This experiment used a serial awakening paradigm. Here participants are woken up several time in the night in order to collect their dream report. This type of paradigm does not interfere with memory consolidation (Schoch et al., 2018). The independent variables during this experiment were time and emotional valence, while the dependent variable was memory performance for one analysis and emotional tone for the other. The experiment used a within-subject design.
Procedure
To make sure that the participants met all the requirements, they had to fill out a screening questionnaire. When selected, an intake meeting was conducted.
At 21:00 hours the participant received an email containing a link to the first set of questionnaires and tasks. First, the participants filled out the PANAS mood questionnaire. Next, they watched a three-minute fragment of both the neutral movie (March of the penguins) and the negative emotional movie (American History X). Three-second clips were taken from these fragments and presented before the images during the following encoding task. During the
encoding task, participants were randomly shown forty blocks (see figure 2). Each block started with a fixation cross and the presentation of a tone (either the high or low tone). Next, either one of the three-second movie clips, an image and an emotional valence question about the image were shown sequentially. There were twenty negative blocks and twenty neutral blocks. On the basis of the tone presented, the participants would be able to predict whether the clip was going to be negative or neutral. The tone-movie association was randomized and counterbalanced across participants. At the end of the encoding task, the participants’ knowledge about this link was tested. Following the encoding task there were several questionnaires; these included PSQI, a general dream characteristics questionnaire, BDI and STAI-T. Afterwards, the participants had to complete the pre-sleep memory test. During this the participants were shown one of the images and were asked whether they had seen this image in the encoding task. Also, they were asked about their confidence in their answer. All the forty images from the encoding task were shown as well as forty foil images. Lastly, the participants would again fill out the PANAS mood questionnaire.
Figure 2: Example of one block of the encoding task. First a cross is shown. Next the
participants are presented with either a high or low tone which will predict which 3-second video clip will be played afterward. After the clip, an image of a butterfly is shown which the participant is asked to remember. At the end of the block, a valence question is asked. Which tone predicts what movie was randomized and counterbalanced across participants.
At 23:00 hours the participants were asked to go to sleep. Starting at midnight, the participants were awoken every hour by a phone call from the researcher. The participant had to answer the questions about what they were experiencing right before the call, including dreams and thoughts as well as feelings and impressions. The last awakening took place at 8:00 hours. After the last awakening, participants were asked to get up and start the last set of questionnaires and tasks. This last set consisted of the PANAS mood questionnaire, a sleep quality questionnaire, a memory test, a viewing of the movie scenes and an exit questionnaire. This second memory test
was identical to the first one, but with different foils and an additional valence question after each image. An overview of the procedure can be found in Figure 3.
Figure 3: overview of experiment. This figure shows the general overview of the whole
experiment. The experiment started with an intake meeting to explain the dream report to the participants and to obtain their consent. Next the participants would fill out multiple questionnaires, do the encoding task and the first memory task. Then the participants went to sleep. During the night, participants were regularly woken up and asked questions from the dream questionnaire. At 8:00 hours on day 2, participants were woken up and asked to fill out more questionnaires and another memory test.
Data analysis
After data collection, the d’-score for memory performance was calculated for each participant. For this score the amount of hits and the amount of false alarms were determined. Hits were defined as images previously seen in the encoding task and correctly identified as such. False alarms were defined as foils that were wrongly classified as previously seen. With these numbers the hit-rate and the false alarm rate were calculated and then, according to signal detection theory, the item recognition memory accuracy (d’-score) was calculated (Macmillan & Creelman, 1991).
Next, the outliers were removed. Outliers were defined as data points that did not fall within a range of three times the standard deviation from the mean. If a participant’s score was an outlier for either the memory performance or emotional tone data, they were completely deleted from the analysis. This was the case for three participants leading to 35 participants for data analysis.
The emotional tone from the images was evaluated by using paired t-tests, or a non-parametric alternative when needed, to compare valence within the pre-sleep memory test as well as to compare ratings pre-sleep and post-sleep. The effect of sleep on memory performance was evaluated using a two-way repeated measures ANOVA with the valence (neutral versus emotional memory) as well as the time (pre-sleep and post-sleep) as the within-subject factor. This statistical analysis was conducted with the program JASP (JASP Version 0.12.2). Additionally, the interaction effect was further evaluated using paired t-tests in R studio (R version 1.2.5033). The significance level was set to p=0.05.
Results
Emotional tone rating
To see if the emotional manipulation worked, the average valence ratings for the neutral and the emotional items were calculated per participant. Figure 4 shows the valence rating of the images for the images associated with the neutral video and those associated with the negative emotional video for both pre- and post-sleep. The ratings were on a scale from 0 to 100, with 0
meaning extremely negative and 100 meaning extremely positive. First, the emotional tone differed between the ratings for the neutral (mean±SD: 55.61±10.43) and the negative (37.91±12.70) images within the pre-sleep task (V = 620, p < .001) with the images associated with the negative video bringing forth a more negative rating than the ones associated with the neutral video. This shows that the experimental design was successful in attributing emotional valance to previously neutral image. As mentioned before, previous research has shown that the emotional tone of a memory is reduced after sleep. To investigate this, the valence rating for the images was compared over time. This analysis showed that emotional tone for the neutral images decreased to a more neutral rating from pre-sleep to post-sleep (50.94±7.83; V = 486.5, p = 0.005). At the same time the rating for the negative emotional images increased significantly after sleep (50.09±7.43), also towards a neutral rating (V = 46, p < .001). Furthermore, no significant difference in rating was seen between the valence groups post-sleep (V = 393.5, p = 0.103).
Figure 4: Valence rating of images for pre-sleep and post-sleep. Within pre-sleep a
significant difference was found between the ratings for the neutral and the emotional images. Furthermore, both the ratings for the neutral as well as for the negative emotional images went towards a neutral rating after sleep. The rating for the neutral images decreased post-sleep, more towards a neutral rating of 50. Whereas the rating for the negative emotional images increased over time, also towards a more neutral rating. In contrast with pre-sleep, post-sleep no significant difference in emotional tone was found between the valence groups. n.s. P > 0.05; *P < 0.05; **P < 0.01; ***P < 0.001.
Memory performance
For the memory performance the d’-score was calculated per participant for each time and valence group. Regarding this score, the higher the participant’s d’-score, the better the memory performance. Figure 5 shows the d’-score per time (pre- and post-sleep) for every valence group (negative and neutral). In table 1, the mean and the standard deviation can be seen for each group. The data was analyzed using a two-way repeated measures ANOVA with the within-subject factors being time and valence.
Figure 5: memory performance per valence of memory for pre- and post-sleep.
Average d’-score for the pre-sleep memory test and the post-sleep memory test for each valence group.
Negative Neutral Average
Pre-sleep 1.31 ± 0.40 1.34 ± 0.36 1.32 ± 0.37
post-sleep 1.55 ± 0.46 1.76 ± 0.42 1.65 ± 0.45
Average 1.43 ± 0.44 1.55 ± 0.44
Table 1: data from memory performance as mean±SD.
First, the main effect of time was significant (F(1,34) = 20.01, p < .001). The main effect is such that the d’-score is higher after sleep (1.65 ± 0.45) than before sleep (1.32 ± 0.37; table 1). Figure 6 shows the average d’-score when the scores of both the negative and neutral memory items are taken together for pre-sleep and post-sleep.
Figure 6: average d’-score of both negative and neutral memory taken together for pre-sleep and post-sleep. Main effect of time on the d’-score. It shows
that post-sleep the d’-score is significantly higher than it is pre-sleep. ***P < 0.001
Second, a main effect of valence was also found (F(1,34) = 5.01, p < 0.032; figure 7). This main effect shows that the overall d’-score for neutral memory items (1.55 ± 0.44) is significantly higher than the score for negative memory items (1.43 ± 0.44; table 1).
Figure 7: average d’-score of pre-sleep and post-sleep taken together for neutral as well as negative memory. This figure shows the main effect of valence
on the d’-score. The figure shows that the d’-score for the neutral memories is significantly higher than the d’-score for the negative memories. *P < 0.05
Additionally, an interaction effect was found between the valence and the time factor (F(1,34) = 5.45, p = 0.026). This interaction can be seen in the interaction plot in figure 8, by looking at the non-parallel lines of the valences over time. Follow-up analysis showed that for neutral memory items, the d’-score was higher for the post-sleep memory test than for the pre-sleep memory test (t(34) = -4.76, p < .001). Similarly, the d’-score for negative memory items was higher post-sleep as compared to pre-sleep (t(34) = -3.08, p = 0.004). Further, for the post-sleep test, neutral memory showed a significant higher d’score than negative memory items (t(34) = -3.18, p = 0.003). While no effect of valence was found for the pre-sleep memory test (t(34) = 0.51, p = 0.616). The results from these simple main effects are summarized in figure 9.
Figure 8: interaction plot of the factors valence and time on the score. This plot shows the average
d’-scores for neutral memory items along with those of the negative memory items for pre-sleep and post-sleep. An interaction effect was found between valence and time.
Figure 9: summary of the simple main effects of time and valence on the d’-score. This graph
shows the d’-score for neutral and negative memory items for both the pre- and post-sleep memory test. Analysis showed a significant difference in d’-score between neutral and negative memory items for the post-sleep memory test, but not for the pre-sleep memory test. Furthermore, this graph shows that both for the neutral memory items as well as for the negative memory items the post-sleep memory test showed a higher d’-score than the pre-post-sleep memory test. n.s. P > 0.05; *P < 0.05; **P < 0.01; ***P < 0.001
Discussion
In sum, this experiment found that the valence ratings of the negative items were more towards negative than those of the neutral items before sleep. This difference in valence ratings was severely reduced during re-evaluation of the images after sleep, with the valence rating of both the neutral and the negative items going to a neutral rating. Regarding the memory performance, this study found no difference in d’-score between the neutral and negative memory items pre-sleep. However, a difference in d’-score between the memory item groups was found post-sleep, with the neutral memory items showing a higher score than the negative ones. Additionally, d’-scores were higher for both neutral and negative memory items post-sleep as when compared to pre-sleep. These results lead to the conclusion that sleep removes the emotional tone of negative memories, while also increasing memory performance for negative memories. Yet this beneficial influence of sleep was higher for neutral memories than for negative memories.
The finding that the emotional tone is reduced after sleep is in line with the hypothesis. However, an alternative explanation for the finding that the emotional ratings returned to natural after sleep could be because of the showing of the videos before the valence question pre-sleep, but not post-sleep. The valence question pre-sleep was asked during the encoding task and was preceded by the negative or neutral video. Post-sleep the valence question was asked during the memory task without the images being preceded by the videos (figure 10). Therefore, the decrease in emotional tone could also be due to the not showing of the videos before the image. Nonetheless, the valence question was asked specifically about the image and not the video or how the participant felt at that moment.
Figure 10: timing of the valence question. This figure shows that what preceded the valence question
differs between pre-sleep and post-sleep. Pre-sleep, the video that created either a neutral or negative emotional state were shown before the valence rating. Post-sleep, only the image was shown before the valence question, not one of the video.
The main effect of time on the d’-score that was found in this experiment was in line with the hypothesis that sleep has a beneficial effect on memory. This result could also be explained by the fact that when the participants do the post-sleep memory test, they have seen the images from the encoding task twice before while during the pre-sleep memory test they have only seen them once before. Re-representing items generally leads to an improved memory performance, called the Hebb repetition effect (Johnson & Miles, 2019). However, this alternative explanation is unlikely, because this effect has been shown when items were repeated at least ten times (Johnson & Miles, 2019). In this experiment the items were only repeated a maximum of 3 times, meaning that the reported results cannot be explained by such effect.
Moreover, the main effect of the valence of the items on the d’-score that was found seems to be mostly driven by the difference post-sleep. When looking at the difference in d’-score between the neutral and negative memory items, the results show that the score pre-sleep is very similar. At the same time, post-sleep d’-scores did differ significantly from each other with the neutral memory items giving a higher d’-score. Therefore, taking these pre- and post-sleep groups together for the main effect will give a significant difference between the neutral and negative memory items. However, it should be noted that this difference is driven by the difference post-sleep.
Not in line with the hypothesis was the result that post-sleep the neutral memory items showed a greater improvement in memory performance. Expected was that the negative emotional memory items would show a greater improvement in memory performance. One possible explanation is that the hypothesis should be rejected and that sleep has a greater benefit on memory performance for neutral items as compared to negative items. Another possible explanation is that this deviation from the hypothesis is due to the serial awakening paradigm employed in this experiment. As stated before, the awakening should not impair memory consolidation (Schoch et al., 2018). However, that experiment used neutral and positive items, not negative items. Further, this same experiment did find that during the serial awakening night the amount of REM sleep decreased significantly when compared with a night without awakenings. As mentioned before, Hutchison and Rathore (2015) proposed a theory that during REM sleep an increase in activity can be seen in the areas in the brain associated with emotional processing. One of the areas is the amygdala. This area has been proposed to modulate hippocampal activity, resulting in a
preferential encoding of emotional memories. It could be that because of the reduced amount of REM sleep, this process did not take place as much as usual. This reduced amount of REM could have led to a lesser benefit of sleep on negative emotional memory performance than normally and then the neutral memories which are not dependent on this mechanism. Although the amount of REM was not a measure in this study and can thus not be shown, it is expected that the limitations are the same as found by Schoch and colleagues (2018). On the other hand, REM sleep has also been implicated in the decoupling of the emotional tone of a memory, but this process was still observed in this study.
The emotional tone findings were in line with previous research which also found a loss of emotional tone across sleep. Moreover, the overall improvement after sleep is in line with previous research. Taken together these two findings provide evidence for the “sleep to remember, sleep to forget” model. The finding that sleep had a greater beneficial influence on neutral memory when compared to negative emotional memory was only partly in line with previous research. The results from this experiment were in line with the findings from the meta-analysis done by Schäfer and colleagues (2019), but not with the studies that claimed that emotional memories are better remembered after sleep than neutral memories or those that found no difference between the two. Further, this study is not in line with the fear conditioning results that also showed an extra beneficial effect of sleep on fear memories. Regarding the controversy in the literature, these findings contribute to the notion that sleep has a less strong beneficial effect on emotional memories as compared to neutral memories.
The results from this study can be used to add to the information about sleep and its effect on emotional memory. This information can then be used to help patients suffering from PTSD. The way this can be done is by integrating the information into an effective treatment related to sleep. For example, one characteristic of PTSD is sleep disturbance (Germain et al., 2007). This experiment showed that after sleep the emotional tone of negative memories goes towards neutral. Thus, helping PTSD with sleeping better might also decrease other symptoms. This information could not only be used for helping PTSD patients, but also patients with other maladaptive emotional memories, such as phobias.
Further research can look into whether the serial awakening did indeed cause the decrease in sleep’s benefit on emotional memories, as speculated above. This can be done by replicating the study, but instead of performing a serial awakening paradigm during the night have people sleep through the whole night. Additionally, further research can look more into sleep composition, percentage REM and NREM sleep, and whether this composition effects the emotional memory consolidation. Further, seeing as this research asked the emotional tone question during the encoding with the videos shown right before the images, further research can replicate this study but ask the emotional tone rating also during the pre-sleep memory test. This can show better the evolution of the emotional tone across sleep.
In conclusion, this research showed that sleep has a different beneficial effect on memories depending on their valence. This adds to the already known information in the debate in the literature. Although more research is needed to explain these differences in findings, this research provides a new piece to the puzzle.
References
Amerian Psyciatric Association. (2013). Diagnostic and Statistical manual of Mental Disorders, Fifth edtion. https://www.psychiatry.org/patients-families/ptsd/what-is-ptsd.
Atienza, M., & Cantero, J. L. (2008). Modulatory effects of emotion and sleep on recollection and familiarity. Journal of Sleep Research, 17(3), 285-294.
https://doi.org/10.1111/j.1365-2869.2008.00661.x
Born, J., & Wilhelm, I. (2012). System consolidation of memory during sleep. Psychological Research, 76, 192-203. https://doi.org/10.1007/s00426-011-0335-6
Bottary, R., Seo, J., Daffre, C., Gazecki, S., Moore, K. N., Kopotiyenko, K., Dominguez, J. P., Gannon, K., Lasko, N. B., Roth, B., Milad, M. R. & Pace-Schott, E. F. (2020). Fear extinction memory is negatively associated with REM in insocmnia disorder. Sleep, zsaa007. https://doi.org/10.1093/sleep/zsaa007
Buzsáki, G. (1989). Two-stage model of memory trace formation: A role for “noisy” brain states . Neuroscience, 31(3), 551-570. https://doi.org/10.1016/0306-4522(89)90423-5
Compean, E., & Hamner, M. (2019). Posttraumatic stress disorder with secondary psychotic features (PTSD-SP): Diagnostic and treatment challenges. Progress in
Neuro-Psychopharmacology and Biological Psychiatry, 88, 265-275. https://doi.org/10.1016/j.pnpbp.2018.08.001
Cramer, P., & Stoll, C. (1779). Aanhangsel van het werk, De uitlandsche kapellen :
voorkomende in de drie waereld-deelen Asia, Africa en America. Amsteldam: Chez Barthelmy Wild.
Diekelmann, S., & Born, J. (2010). The memory function of sleep. Nature Reviews Neuroscience, 11, 114-126. https://doi.org/10.1038/nrn2762
Germain, A., Shear, M. K., Hall, M., & Buysse, D. J. (2007). Effects of a brief behavioral treatment for PTSD-related sleep disturbances: A pilot study. Behaviour Research and Therapy, 45(3), 627-632. https://doi.org/10.1016/j.brat.2006.04.009
Goldstein, A. N. & Walker, M. P. (2014). The Role of Sleep in Emotional Brain Function. Annual Review of Clinical Psychology, 10, 679-708. https://doi.org/10.1146/annurev-clinpsy-032813-153716
Groch, S., Wilhelm, I., Diekelmann, S., & Born, J. (2013). The role of REM sleep in the
processing of emotional memories: Evidence from behavior and event-related potentials. Neurobiology of Learning and Memory, 99, 1-9.
https://doi.org/10.1016/j.nlm.2012.10.006
Herringa, R. J. (2017). Trauma, PTSD, and the Developing Brain. Current Psychiatry Reports, 19(10), 69. https://doi.org/10.1007/s11920-017-0825-3
Hu, P., Stylos-Allen, M., & Walker, M. P. (2006). Sleep Facilitates Consolidation of Emotional Declarative memory. Psychological science, 17(10), 891-898.
https://doi.org/10.1111/j.1467-9280.2006.01799.x
Hutchison, I. C., & Rathore, S. (2015). The role of REM sleep theta activity in emotional memory. Frontiers in Psychology, 6, 1439. https://doi.org/10.3389/fpsyg.2015.01439 Johnson, A. J., & Miles, C. (2019). Visual Hebb Repetition Effects: The Role of Psychological
Distinctiveness Revisited. Frontiers in Psychology, 10, 17. https://doi.org/10.3389/fpsyg.2019.00017
Kaida, K., Niki, K., & Born, J. (2015). Role of sleep for encoding of emotional memory. Neurobiology of learning and memory, 121, 72-79.
Macmillan, N. A., & Creelman, C. D. (1991). Detection theory: A user's guide. New York: Cambridge University Press.
Menz, M. M., Rihm, J. S., Salari, N., Born, J., Kalisch, R., Pape, H. C., Marshall, L. & Büchel, C. (2013). The role of sleep and sleep deprivation in consolidating fear memories. NeuroImage, 75, 87-96. https://doi.org/10.1016/j.neuroimage.2013.03.001
Merriam-Webster. (n.d.). Feeling tone. Retrieved from: https://www.merriam-webster.com/dictionary/feeling%20tone
Payne, J. D., Stickgold, R., Swanberg, K., & Kensinger, E. A. (2008). Sleep Preferentially Enhances memory for Emotional Components of Scenes. Psychological Science, 19(8), 781-788. https://doi.org/10.1111/j.1467-9280.2008.02157.x
Rasch, B., & Born, J. (2013). About sleep's role in memory. Physiological Review, 93(2), 681-766. https://doi.org/10.1152/physrev.00032.2012
Rihm, J. S., & Rasch, B. (2015). Replay of conditioned stimuli during late REM and stage N2 sleep influences affective tone rather than emotional memory strength. Neurobiology of Learning and Memory, 122, 142-151. https://doi.org/10.1016/j.nlm.2015.04.008
Schäfer, S. K., Wirth, B. E., Staginnus, M., Becker, N., Michael, T., & Sopp, R. (2019). Sleep's impact on emotional memory: A meta-analysis of whole-night, nap, and REM sleep effects. bioRxiv. https://doi.org/10.1101/732651
Schaefer, A., Nils, F., Sanchez, X., & Philippot, P. (2010). Assessing the effectiveness of a large database of emotion-eliciting films: A new tool for emotion researchers . Cognition and Emotion, 24(7), 1153-1172. https://doi.org/10.1080/02699930903274322
Schoch, S. F., Cordi, M. J., Schredl, M., & Rasch , B. (2018). The effect of dream report collection and dream incorporation on memory consolidation during sleep. Journal of sleep Research, 28(1), e12754. https://doi.org/10.1111/jsr.12754
Talamini, L. M., Bringmann, L. F., de Boer, M., & Hofman, W. F. (2013). Sleeping Worries Away or Worrying Away Sleep? Physiological Evidence on Sleep-Emotion Interactions. PLoS ONE, 8(5), e62480. https://doi.org/10.1371/journal.pone.0062480
Tempesta, D., Socci, V., Coppo, M., Dello loio, G., Nepa, V., De Gennaro, L., & Ferrara, M. (2016). The effect of sleep deprivation on the encoding of contextual and non-contextual aspects of emotional memory. Neurobiology of Learning and memory, 131, 9-17.
https://doi.org/10.1016/j.nlm.2016.03.007
Tempesta, D., Socci, V., De Gennaro, L., & Ferrara, M. (2018). Sleep and emotional processing. Sleep Medicine Reviews, 40, 183-195. https://doi.org/10.1016/j.smrv.2017.12.005
van der Helm, E., Gujar, N., Nishida, M., & Walker, M. P. (2011). Sleep-Dependent Facilitation of Episodic Memory Details. PLoS One, 6(11).
https://doi.org/10.1371/journal.pone.0027421
van der Helm, E., Yao, J., Dutt, S., Rao, V., Saletin, J. M., & Walker, M. P. (2011). REM Sleep De-potentiates Amygdala Activity to Previous Emotional Experiences. Current Biology, 21(23), 2029-2032. https://doi.org/10.1016/j.cub.2011.10.052
Zohar, D., Tzischinsky, O., Epstein, R., & Lavie, P. (2005). The Effects of Sleep Loss on Medical Residents' Emotional Reactions to Work Events: a Cognitive-Energy Model. Sleep, 28(1), 47-54. http://doi.org/10.1093/sleep/28.1.47
