Horne (1998) argues that there are two modes of sleep, “Core” and “Optional” sleep. Core sleep refers to an essential component, that is; necessary for the maintenance of normal psychological performance; made up largely by delta electroencephalonographic (EEG) activity, or slow wave sleep (SWS), and also about half of our nightly rapid eye movement (REM) sleep; largely concentrated in the first three or four sleep stages (i.e. the initial 4-5 hours of sleep).
Optional sleep basically constitutes the remainder of sleep and is appears to be more dispensable. It is extremely important to note that Horne makes this division on the assumption that it is for the brain and behaviour (in particular the cerebrum), that sleep appears to be the most vital and the only organ which shows signs of restitution. Circumstantial evidence to support this notion is provided by; 1) pointing to the fact that the brain is unable to relax outside of sleep – electrical activity during relaxed wakefulness indicates that the cerebrum continues to work at a high rate; and 2) sleep deprivation studies have pointed to a more profound impairment to cerebral function than to any other part of the brain or body, where functions generally ‘slow down’ more than they do become impaired (e.g. Horne, 1978, 1988; Johnson, 1982).
Accordingly, sleepiness can be subdivided into Core and Optional components. For a ‘normal or healthy sleeper1’ optional sleepiness would mostly prevail when the individual sleeps for less time than their usual quota, but still enough to satisfy their core sleep requirement (e.g. 6 hours for an 8-hour sleeper). If however this core sleep requirement is not satisfied, then core sleepiness ensues, as manifested by behavioural and performance decrements. Let us now look at the different tools for measuring sleepiness.
Overview: methods for measuring Sleepiness
There are numerous methods for investigating sleepiness available to the sleep researcher. A brief overview will be provided here with a view to introducing the most commonly used methods and providing a simple description of the key features of each. These methods can generally be sub-divided into ‘Subjective’ methods ‘Objective’ methods. However, it should be noted that the author wishes to pay particular attention to the multiple sleep latency test (MSLT). He proposes that the MSLT provides a particularly good basis for understanding sleepiness and its measurement, but moreover by looking at its implementation in research, will also allow us to look at the interplay between the methods and the methodology; a recurring theme of this essay.
SUBJECTIVE MEASURES
Visual Analogue Scales This is the simplest test of subjective sleepiness. It is not in itself a recognised measure, but is commonly used in conjunction with the psychomotor vigilance task (PVT- see below). It comprises a 100mm line, with ‘Very Alert’ and ‘Very Sleepy’ at the beginning and end of the line respectively, upon which participants are asked to graphically mark where he/she believes they should be on the scale according to what they are feeling at that moment in time.
Epworth Sleepiness Scale (ESS) This test was developed by Johns (1991), as a measure of ‘trait’ sleepiness. Thus it measures a person’s general level of daytime sleepiness. A self-administered questionnaire is used to capture the participant’s ratings, on an ordinal scale of 0 (never) to 3 (high chance), the chances of dozing off or falling asleep when in eight different situations that Johns claims are encountered commonly in daily life (e.g. sitting and reading, watching TV etc). The ESS is a measure of ‘trait’ sleepiness (a general value for the participant) and not ‘state’ sleepiness (a more situation- and temporal-specific value). The absence of repeated measures means that the test assumes that the participant’s assessment of their general sleepiness would not vary. Further criticism is levelled at the ESS for failing to refer to the time of day or to the context in which each situation occurs.2
Stanford Sleepiness Scale (SSS) The SSS was developed by Hoddes and colleagues (1973) and is another state measure, consisting of a scale from 1 to 7 on which participants rate themselves according to how they feel at the time. Some sleep researchers claim that this is the best validated of subjective sleepiness scale (Roehrs, et al., 2000), but the test is not without its criticism. The main criticism of the test is that it is culturally biased due to the language used to describe sleepiness, i.e. while ‘woozy’ and ‘foggy’ might be terms used by Americans, they would not appear to form part of, for example, a British person’s everyday vocabulary.
Furthermore, Horne (2003) has recently commented that the semantics of the SSS might account for some of its peculiar findings. Referring to Maclean, et al.’s work (1989), Horne points out that the terms “alert” and “relaxed” tap into different major mood factors on the Profile of Moods States (POMS- Lorr, et al., 1967; McNair, et al., 1967), which might be indicative that the SSS, a ‘gold standard’, might not be a ‘pure’ scale of sleepiness.
Karolinska Sleepiness Scale (KSS) The KSS is another subjective test, very similar to the SSS. It was validated by �kerstedt and Gillbert (1990), whose EEG results indicated that during wakefulness, subjective sleepiness is reflected in increased activity in the alpha and theta bands. The main difference to the SSS is that the semantics used in the KSS adhere to just one of the two dimensions that the work of Maclean, et al. indicated the SSS appeared to be measuring; namely the “alertness-sleepiness” dimension. As such the KSS can be viewed as a ‘purer’ scale of sleepiness.