Sleep

Sleep Enhances Emotional Memories

Related to the impact sleep has on declarative memories is the effect of sleep on emotional memories. As mentioned, these memories are often episodic in nature. However, they are special in that they carry an emotional undertone, which differentiates them from neutral episodic memories. This emotional content can afford a memory benefit on its own, but when added to sleep, this benefit is compounded.

Sleep and memory for emotional verbal stimuli

One of the first studies to investigate this impact of sleep on emotional memory used written descriptions as study material. Specifically, Wagner, Gais, and Born (2001) had people read and memorize descriptions of either emotional (e.g., someone committing a murder) or neutral (e.g., sculpture manufacturing) content for a later memory test. They further combined the study of this material with the early/late sleep design discussed earlier so that participants received either three hours of sleep early or late in the night after studying the descriptions. Two control groups remained awake during these intervals. They found that those who received late night sleep after learning these descriptions remembered significantly more of the emotional descriptions than the neutral ones, and this performance was far better compared to those who received early night sleep, early night wake, or late night wake.

Since REM sleep dominates late in the night, these results suggest that this stage of sleep may be particularly important for emotional memory consolidation. However, this differentiation between performance in the early and late sleep groups did not remain for a long-term memory test.

Four years after this study was conducted, Wagner and colleagues (2006) contacted the original participants to determine the strength of the remaining memory for the descriptions studied. They found that those who slept after originally learning the descriptions, regardless of whether the sleep occurred over the first half of the night or the second, remembered significantly more of the emotional descriptions than those who remained awake after learning. This was in contrast to memory performance for the neutral descriptions, which was not different between the sleep and wake groups.

At first glance, it would appear that specific sleep stages might not be important for emotional memory, at least not in the long run. However, as Wagner and colleagues (2006) point out, those in the early sleep group who originally did not show a similar enhancement of emotional memory as those in the late sleep group actually received the second REM-rich half of the night of sleep after the original memory test when they were allowed to return to sleep. The authors suggest that this might have allowed the memory trace to undergo appropriate selective consolidation processes during the ensuing REM period, and the impact of this occurrence was simply not evident until later testing. Thus sleep, and specifically REM sleep, appears to aid emotional memories, an effect which lasts for years to come.

Sleep and memory for emotional picture stimuli

Similar results of sleep on memory have been obtained for emotional pictorial stimuli as well. For example, Hu, Stylos-Allan, and Walker (2006) required participants to study lists of pictures that were either emotionally arousing or neutral for a later recognition test. Pictures were taken from the International Affective Picture System (IAPS; Lang, Bradley, and Cuthbert, 2008), a widely used database of emotional and neutral pictures with standardized arousal and valence ratings.

After a night of sleep or a day filled with wake, participants were tested for their memory of the pictures using the remember/know (R/K) paradigm. This method requires participants to make one of three memory responses. A “remember” response indicates a vivid memory of the stimulus, complete with contextual details. A “know” response, on the other hand, indicates familiarity of the stimulus without the ability to retrieve contextual details about its original presentation. Finally, a “new” response would be given to an item not seen previously.

The authors found that sleep only offered a benefit to memory for the emotionally arousing pictures given “know” responses. In other words, participants who slept were more likely to remember the emotional than the neutral pictures on the basis of familiarity. In contrast, those who remained awake during the retention interval did not remember emotionally arousing pictures any better than neutral ones. Although there was no memory benefit for “remember” judgments in those who slept, there was an increase in conservative responding, as measured by the calculation of response bias C, within this response category only after sleep, especially for emotionally arousing pictures.

This finding suggests that sleep may have increased confidence in these judgments, allowing for better differentiation of these items. However, others have also found that an entire night of sleep is not necessary for emotional memory enhancements. For example, Nishida and colleagues (2009) found an enhancement of memory for emotional pictures after just a 90-minute nap.

In their study, participants encoded separate sets of emotional and neutral IAPS pictures both four hours and 15 minutes prior to a memory test for both sets. Those in the sleep group obtained a 90-minute nap in the time between the encoding of the two lists, meaning that only the list encoded four hours prior to the test underwent a consolidation period filled with sleep in this group.

The authors found that those in the nap group exhibited improved memory for the emotional pictures encoded before the nap as compared to those encoded after the nap. Those who did not nap did not show this same memory difference. Further, they found that the amount of REM obtained during the nap, along with theta activity during this stage, was positively correlated with this emotional memory improvement, supporting previous findings that suggest an important role for this specific stage on emotional memory processing for both picture and verbal stimuli (Payne, Chambers, and Kensinger, 2012; Wagner, Gais and Born, 2001).

Interestingly, sleep has also been shown to impact memory for specific components of these pictures, rather than for the images in their entirety. Payne and colleagues (2008) found this to be the case in their study, which used pictures edited to contain negative or neutral objects placed on neutral backgrounds to investigate memory for these separate components (Figure 11.5).

Similar to the methods of Hu, Stylos,-Allan, and Walker (2006), participants encoded these scenes in their entirety (although a memory test was not expected in this case), and their memory for pictures was later tested after either a period of nocturnal sleep, or equivalent time of daytime wakefulness.

However, contrary to the methods of Hu and colleagues, Payne et al. (2008) did not test memory for the pictures in their entirety. They separated the objects from their respective backgrounds and tested recognition memory for these components separately with new components intermixed (i.e., foils).

Control groups tested only 30 minutes after encoding revealed that an overall trade-off occurred, whereby negative foreground objects were remembered better than neutral ones, but at the expense of the memory for their corresponding neutral backgrounds. Critically, they found that this result was actually enhanced after a period of sleep, revealing a boost in memory for negative objects above levels seen in the control groups (Figure 11.6).

Figure X6 Differences in memory performance for emotional and neutral scene components between 30-minute and 12-hour delays that spanned sleep or wake. Memory for objects and backgrounds (Bkg) generally decayed over time for both the wake and sleep groups, except for emotional objects, which were facilitated by sleep above 30-minute performance levels. Source: Current Directions in Psychological Science (2010), 19, 293.

Conversely, a period of wake resulted in a general decay of memory for the separate components.

A subsequent study revealed that this selective enhancement in memory for the negative objects was correlated with the amount of REM sleep obtained during the night (Payne, Chambers, and Kensinger, 2012). Thus, these results point toward a selective consolidation function of sleep, and REM sleep in particular, that works to pick out and enhance the most important parts of a scene. This function may have evolutionary significance, because it allows the enhancement of those emotional components of memories that might be critical to future survival, such as remembering a snake to be avoided later.