The role of sleep in memory

Why do we need sleep?

A lot of theories have been thrown up over the years as to what we need sleep for (to keep us wandering out of our caves and being eaten by sabertooth tigers, is one of the more entertaining possibilities), but noone has yet been able to point to a specific function of the sleep state that would explain why we have it and why we need so much of it.

One of the things we do know is that young birds and mammals need as much as three times the amount of sleep as adult birds and mammals. It has been suspected that neuronal connections are remodeled during sleep, and this has recently been supported in a study using cats (Cats who were allowed to sleep for six hours after their vision was blocked in one eye for six hours, developed twice as many new or modified brain connections as those cats who were kept awake in a dark room for the six hours after the period of visual deprivation).

Certainly a number of studies have shown that animals and humans deprived of sleep do not perform well on memory tasks, and research has suggested that there may be a relationship between excessive daytime sleepiness (EDS) and cognitive deficits. A recent study has found that for seniors at least, EDS is an important risk factor for cognitive impairment.

The effect of sleep on memory and learning

Some memory tasks are more affected be sleep deprivation than others. A recent study, for example, found that recognition memory for faces was unaffected by people being deprived of sleep for 35 hours. However, while the sleep-deprived people remembered that the faces were familiar, they did have much more difficulty remembering in which of two sets of photos the faces had appeared. In other words, their memory for the context of the faces was significantly worse. (The selective effect of sleep on contextual memory is also supported in a recent mouse study – see below)

While large doses of caffeine reduced the feelings of sleepiness and improved the ability of the sleep-deprived subjects to remember which set the face had appeared in, the level of recall was still significantly below the level of the non-sleep-deprived subjects. (For you coffee addicts, no, the caffeine didn’t help the people who were not sleep-deprived).

Interestingly, sleep deprivation increased the subjects’ belief that they were right, especially when they were wrong. In this case, whether or not they had had caffeine made no difference.

In another series of experiments, the brains of sleep-deprived and rested participants were scanned while the participants performed complex cognitive tasks. In the first experiment, the task was an arithmetic task involving working memory. Sleep-deprived participants performed worse on this task, and the fMRI scan confirmed less activity in the prefrontal cortex for these participants. In the second experiment, the task involved verbal learning. Again, those sleep-deprived performed worse, but in this case, only a little, and the prefrontal areas of the brain remained active, while parietal lobe activity actually increased. However, activity in the left temporal lobe (a language-processing area) decreased. In the third study, participants were given a "divided-attention" task, in which they completed both an arithmetic and a verbal-learning task. Again, sleep-deprived participants showed poorer performance, depressed brain activation in the left temporal region and heightened activation in prefrontal and parietal regions. There was also increased activation in areas of the brain that are involved in sustained attention and error monitoring.

These results indicate that sleep deprivation affects different cognitive tasks in different ways, and also that parts of the brain are able to at least partially compensate for the effects of sleep deprivation.

Sleep deprivation mimics aging?

A report in the medical journal The Lancet, said that cutting back from the standard eight down to four hours of sleep each night produced striking changes in glucose tolerance and endocrine function that mimicked many of the hallmarks of aging. Dr Eve Van Cauter, professor of medicine at the University of Chicago and director of the study, said, "We suspect that chronic sleep loss may not only hasten the onset but could also increase the severity of age-related ailments such as diabetes, hypertension, obesity and memory loss."

Should we draw any conclusion from the finding that sleep deprivation increased the subjects’ belief that they were right, especially when they were wrong, and the finding that chronic sleep deprivation may mimic the hallmarks of aging? No, let us merely note that many people become more certain of their own opinions as they mature into wisdom.

Is sleep necessary to consolidate memories?

This is the big question, still being argued by the researchers. The weight of the evidence, however, seems to be coming down on the answer, yes, sleep is necessary to consolidate memories — although maybe for only some types of memory. Most of the research favoring sleep’s importance in consolidation has used procedural / skill memory — sequences of actions.

From this research, it does seem that it is the act of sleep itself, not simply the passage of time, that is critical to convert new memories into long-term memory codes.

Some of the debate in this area concerns the stage of sleep that may be necessary. The contenders are the deep "slow wave" sleep that occurs in the first half of the night, and "REM" (rapid eye movement) sleep (that occurs while you are dreaming). Experiments that have found sleep necessary for consolidation tend to support slow-wave sleep as the important part of the cycle, however REM sleep may be important for other types of memory processing.

Sleep studies cast light on the memory cycle

Two new studies provide support both for the theory that sleep is important for the consolidation of procedural memories, and the new theory of what I have termed the "memory life-cycle".

In the first study, 100 young adults (18 to 27) learned several different finger-tapping sequences. It was found that participants remembered the sequence even if they learned a second sequence 6 hours later, and performance on both sequences improved slightly after a night's sleep. However, if, on day 2, people who had learned one sequence were briefly retested on it and then trained on a new sequence, their performance on the first sequence plummeted on day 3. If the first sequence wasn't retested before learning the new sequence, they performed both sequences accurately on day 3.

In another study, 84 college students were trained to identify a series of similar-sounding words produced by a synthetic-speech machine. Participants who underwent training in the morning performed well in subsequent tests that morning, but tests later in the day showed that their word-recognition skill had declined. However, after a full night's sleep, they performed at their original levels. Participants trained in the evening performed just as well 24 hours later as people trained in the morning did. Since they went to bed shortly after training, those in the evening group didn't exhibit the temporary performance declines observed in the morning group.

On the basis of these studies, researchers identified three stages of memory processing: the first stage of memory — its stabilization — seems to take around six hours. During this period, the memory appears particularly vulnerable to being “lost”. The second stage of memory processing — consolidation — occurs during sleep. The third and final stage is the recall phase, when the memory is once again ready to be accessed and re-edited. (see my article on consolidation for more explanation of the processes of consolidation and re-consolidation)

The researchers made a useful analogy with creating a word-processing document on the computer. The first stage is when you hit “Save” and the computer files the document in your hard drive. On the computer, this takes seconds. The second stage is comparable to someone coming and tidying up your word document — reorganizing it and tightening it up.

The most surprising aspect of this research is the time it appears to take for memories to initially stabilize — seconds for the computer saving the document, but up to six hours for us!

See news reports on sleep's role in memory

See news reports on the effects of sleep deprivation