How Sleep Helps Memory
Believe it! Sleep helps memory. Here are some reasons why.
Posted March 11, 2011
There is no longer any doubt. Sleep does improve the gelling or consolidation of memory for recently encoded information. Research is now focusing on how this happens and what other factors interact with the sleep effect. At least two processes seem to be at work: 1) sleep protects new memories from disruption by the interfering experiences that are inevitable during wakefulness, and 2) sleep consolidates memories according to their relative importance and the learner's expectations for remembering.
To test the idea, they asked 24 volunteers to memorize the two-dimensional loca-tion of 15 pairs of cards with pictures of animals and everyday objects. During the study time, they were also exposed continuously to a slightly unpleasant odor, which was intended to be an associational cue.
Forty minutes later, the volunteers were asked to learn a second, slightly different set of card pairs. This second task was to act as an interfering disruptor of the initial learning. The difference is that after the first memorization session, half of the group stayed awake and the other half took a nap. For 20 minutes during the break after the first study session, the odor cue was presented with the intent of helping to reactivate the memory of the first session. The awake group got the odor cue for 20 minutes just before starting the second learning session, while the sleep group got the odor cue dur-ing the last 20 minutes of the nap (dreaming did not occur, because it normally requires more than 40 minutes of sleep to start appearing).
When both groups were tested for recall of the first set of cards, the sleep group remembered much better (85% correct versus 60% for the awake group). The explana-tion begins with the knowledge that when temporary memories (as for the first card set) are recalled, they are vulnerable to being destroyed by new mental activity (as with the second card set). In this study, memory was reactivated in both wakefulness and sleep by the odor cue. Yet, the memorization processes that apparently persisted during sleep made the original memories more resistant to disruption. By the time of the second interfering task some 40 minutes later, much of the initial learning had gelled during sleep, but less so during wakefulness.
These authors also performed brain imaging that showed that the nap group had mostly completed a shift in activity from the temporary processing area (in the hippo-campus) to storage areas in the cortex. This was not true for the awake group. You might say that sleep enabled the information to be "uploaded from RAM to the hard drive" better than in the constant awake condition. Of course this computer metaphor breaks down in other respects. Biological memory is dynamic, readily degraded over time or changed by new experience. Also, recall of biological memory launches a recon-structive process whereby the memory can be reinforced or drastically altered.
The practical application, as I see it, is to take a short nap as soon as possible after trying to memorize something really important. For example, during a study session for a school exam, take a nap right away so that it has a better chance to consolidate than if you stayed awake and got exposed to many new interfering situations and stimuli.
Two new studies shed some light on prioritization of memory formation during sleep. We all have had the experience of improved memory if we know others expect us to remember. I guess such improvement occurs because we work harder at it, using more intensive rehearsal and perhaps using deliberate association strategies.
But we now find out from a recent study that the sleep effect on improving memo-ry formation benefits from the relevance of the learned information. Since sleep usually occurs significantly later than the learning and original encoding, this effect must arise from the consolidation process during sleep.
A recent study from this same German research lab has revealed that sleep helps memory formation the most if you know you will need the information later. That is, it seems that the brain prioritizes its consolidation operations during sleep to favor con-solidation of information that is most important. The study tested 193 volunteers for recall of a variety of memory tasks. Some subjects were exposed to the learning ma-terial early in the day, when there would be no sleep involved. The others were exposed to the same material late, just before the night's sleep. When subjects were told they would be tested later, they were more likely to remember if they had slept immediately after the learning. This was true for both procedural tasks (like finger-tapping sequences) or declarative tasks such as word matching or stating card-pair locations. Moreover, subjects who were told they would be tested later spent more total time in the deepest stage of Sleep (Stage IV) than did comparable subjects who were not told they would be tested later. Presumably, the brain is using Stage IV to accomplish this differential consolidation process.
In a recent study from a French group, the study focus was on sleep's apparent ability to prioritize memory formation based on prior instructions to remember or for-get items in a learning task. In the learning task, volunteers were shown 100 French words, one at a time. Fifty of these had accompanying instruction "to be remembered" and the other 50 "to be forgotten," presented in a pseudorandom sequence that pre-vented more than three words of the same type being presented consecutively. After the training session, subjects were divided into two groups, one which was sent home to continue their normal activities and to sleep on their usual schedule for the next three nights. The other group was denied the first night's sleep after training, where they stayed up all that night watching movies or playing games. Otherwise, this group was treated the same. On the fourth day, both groups were tested for recall with presentation the 100 of the original words and 100 new ones to serve as distracters. The task was to identify which words were in the original list.
Questionnaires revealed any strategies the subjects used in trying to remember "to be remembered" words and trying to ignore "to be forgotten" words. No subject intensively rehearsed the original items during the three-day interval, but of course casual rehearsal was going on. Generally, subjects made associations of "to be remembered" words with memories of personal events or with short stories or sentences. Mental images were much less used. Of course, no such rehearsals occurred with "to be forgotten" words.
Upon testing, both groups had about the same degree of correct recall for "to be remembered" words. But the sleep-deprived groups remembered more of the words they were not supposed "to be forgotten." Thus, it would seem that during sleep, the brain preserved its ability to remember words that were expected to be remembered and discriminated against remembering words that were unimportant. Recall that the instructions to remember or forget were given at the time of initial encoding. Thus, the brain must have preserved these instructions and followed them in the consolidation process during sleep. Though the authors did not mention it, the poor ability of sleep-deprived subjects to discriminate between the two categories of words could have ari-sen because being awake for a whole day after learning interfered with remembering and following instructions at the time of encoding.
Don't forget, if you have students in your life, have them check out
my new eBook, "Better Grades, Less Effort."
Diekelmann, S., Büchel, Born, J., and Rasch, Björn. 2011. Labile or stable: opposing con-sequences for memory when reactivated during wakefulness and sleep. Nature Neuroscience. Jan. 23. doi: 10.1038/nn.2744
Rauchs, G. et al. 2011. Sleep contributes to the strengthening of some memories over others, depending on hippocampal activity at learning. J. Neuroscience. 31 (7): 2563-2568.
Wilhelm, I. et al. 2011. Sleep selectively enhances memory expected to be of future re-levance. J. Neuroscience. 31 (5): 1563-1569.