ABSTRACT

When we are asleep, we show levels of unconsciousness, are less vigilant and less responsive to outside stimuli. These behavioural characteristics of sleep are easily reversible during wakefulness, and they can be applied to many species. Sleep is best discerned from conscious rest by investigation of the electrical properties of the brain measured by the electroencephalogram (EEG). In mammals, sleep is broadly subdivided into ‘rapid eye movement’ (REM) and ‘non-rapid eye movement’ (nonREM) sleep, with a further specification of four levels (1 through 4) of nonREM sleep in primates. NonREM sleep is also known as slow wave sleep (SWS). Figure 27.1 shows the time course of a typical first 3 hours of human sleep. At the onset, we quickly move through the first three nonREM sleep stages, ending at stage 4, the deepest nonREM sleep stage. After remaining in nonREM sleep for some time, we leave deep sleep, and enter into REM sleep. We cycle through these stages and, during a typical night, each nonREM-REM sleep cycle lasts approximately 1.5 hours. Throughout the night, time spent in REM and nonREM sleep is not equally distributed. In the early night we spend more time in nonREM sleep, while in the latter part of the night REM sleep is more predominant than at the beginning of the night. Electrical activity measured on the scalp (electroencephalography) is the most common way of

measuring brain activity during sleep. Transforming the EEG into the frequency domain shows a slowing of the EEG with each stage of nonREM sleep. During nonREM sleep stages 3 and 4, the EEG frequency exhibits high amplitude, slow-frequency brain activity in the range 0.5-4 Hz, which is also known as delta sleep. During REM sleep, the EEG shows high frequency and irregular activity, and appears similar to wakefulness. Due to the similarity of the REM sleep EEG to the wakefulness EEG, REM sleep is sometimes referred to as ‘paradoxical sleep’. While the rapid eye movements and the EEG spectrum seem to indicate activity, behavioural activity and muscle tone in REM sleep is low. Sleep and wakefulness are driven by two processes that are fundamentally different. The first component

is a sleep-promoting process, and is quite intuitive: while we are awake, we build up a sleep debt. This is known as sleep homeostasis and, although not on a linear scale, being awake increases our sleep propensity. The second component that drives sleep and wakefulness cycles is our internal clock. Circadian (= approximately 24 hours) rhythms drive both sleep-and wake-promoting signals, and the interaction between the homeostatic and circadian components is what actually determines when our net sleep drive is high or low. Because these two processes are so strongly interlinked, it is important to also consider the circadian clock when interpreting sleep-dependent learning.