Scientists have shown numerous ways in which sleep is related to memory. In a study conducted by Turner, Drummond, Salamat, and Brown working memory was shown to be affected by sleep deprivation. Working memory is important because it keeps information active for further processing and supports higher-level cognitive functions such as decision making, reasoning, and episodic memory. Turner et al. allowed 18 women and 22 men to sleep only 26 minutes per night over a 4-day period. Subjects were given initial cognitive tests while well rested and then tested again twice a day during the 4 days of sleep deprivation. On the final test the average working memory span of the sleep deprived group had dropped by 38% in comparison to the control group.
Memory also seems to be affected differently by certain stages of sleep such as REM and slow-wave sleep (SWS). In one study cited in Born, Rasch, and Gais multiple groups of human subjects were used: wake control groups and sleep test groups. Sleep and wake groups were taught a task and then tested on it both on early and late nights, with the order of nights balanced across participants. When the subjects' brains were scanned during sleep, hypnograms revealed that SWS was the dominant sleep stage during the early night representing around 23% on average for sleep stage activity. The early night test group performed 16% better on the declarative memory test than the control group. During late night sleep, REM became the most active sleep stage at about 24%, and the late night test group performed 25% better on the procedural memory test than the control group. This indicates that procedural memory benefits from late REM-rich sleep whereas declarative memory benefits from early SWS-rich sleep.
Another study conducted by Datta[30 indirectly supports these results. The subjects chosen were 22 male rats. A box was constructed where a single rat could move freely from one end to the other. The bottom of the box was made of a steel grate. A light would shine in the box accompanied by a sound. After a 5 second delay an electrical shock would be applied. Once the shock commenced the rat could move to the other end of the box, ending the shock immediately. The rat could also use the 5-second delay to move to the other end of the box and avoid the shock entirely. The length of the shock never exceeded 5 seconds. This was repeated 30 times for half the rats. The other half, the control group, was placed in the same trial but the rats were shocked regardless of their reaction. After each of the training sessions the rat would be placed in a recording cage for 6 hours of polygraphic recordings. This process was repeated for 3 consecutive days. This study found that during the post-trial sleep recording session rats spent 25.47% more time in REM sleep after learning trials than after control trials. These trials support the results of the Born et al. study, indicating an obvious correlation between REM sleep and procedural knowledge.
Another interesting observation of the Datta study is that the learning group spent 180% more time in SWS than did the control group during the post-trial sleep-recording session. This phenomenon is supported by a study performed by Kudrimoti, Barnes, and McNaughton. This study shows that after spatial exploration activity, patterns of hippocampal place cells are reactivated during SWS following the experiment. In a study by Kudrimoti et al. seven rats were run through a linear track using rewards on either end
The rats would then be placed in the track for 30 minutes to allow them to adjust (PRE), then they ran the track with reward based training for 30 minutes (RUN), and then they were allowed to rest for 30 minutes. During each of these three periods EEG data were collected for information on the rats' sleep stages. Kudrimoti et al. computed the mean firing rates of hippocampal place cells during pre-behavior SWS (PRE) and three 10-minute intervals in post-behavior SWS (POST) by averaging across 22 track-running sessions from seven rats. The results showed that 10 minutes after the trial RUN session there was a 12% increase in the mean firing rate of hippocampal place cells from the PRE level, however after 20 minutes the mean firing rate returned rapidly toward the PRE level. The elevated firing of hippocampal place cells during SWS after spatial exploration could explain why there were elevated levels of SWS sleep in Datta's study as it also dealt with a form of spatial exploration.
The different studies all suggest that there is a correlation between sleep and the many complex functions of memory. Harvard sleep researchers Saper and Stickgold point out that an essential part of memory and learning consists of nerve cell dendrites sending information to the cell body to be organized into new neuronal connections. This process demands that no external information is presented to these dendrites, and they suggest that this may be why it is during sleep that we solidify memories and organize knowledge.
Posts
