A new approach to the construct of alertness

Abstract

Objective

There are no psychometric measures to evaluate the critical construct of alertness. We, therefore, developed two questionnaires to measure alertness, the Toronto Hospital Alertness Test (THAT) and the ZOGIM-A, and evaluated their psychometric properties.

Methods

We examined the correspondence between scores on the THAT and the ZOGIM-A in a sample of sleep clinic outpatients (n=96) with Maintenance of Wakefulness Test (MWT) results after an overnight sleep study, physiological sleep parameters, measures of subjective sleepiness, and two psychosocial variables (psychological well-being and emotional distress). Test–retest reliability was estimated based on responses from an independent sample of 295 sleep clinic outpatients who completed the instruments before and after an overnight sleep study.

Results

High values were observed for both the THAT (r_tt=.79) and the ZOGIM-A (r_tt=.70). Internal consistency reliability (coefficient alpha) was also high: .96 for THAT and .83 for ZOGIM-A. Although neither of the new scales correlated significantly with measures derived from the MWT or nocturnal physiological measures, the two alertness scales did correlate significantly and as hypothesized with subjective measures of sleepiness and other psychosocial measures.

Conclusions

These results support the perspective that subjective alertness cannot be reduced to the absence of sleepiness and corroborate the psychometric adequacy of the THAT and the ZOGIM-A as unique indices of alertness that complement objective data obtained via MWT and physiological indices of sleep architecture.

Introduction

Alertness and daytime sleepiness are complex behavioral and physiological states with a major impact on our daily lives [1]. It is philosophically and scientifically difficult to define states of consciousness. We know them intimately from the inside, but capturing them with external criteria is not an easy task. Alertness is one such concept. By alertness, we imply a state of responsivity to both introceptive and external stimuli. An alert individual is conscious: aware of environment, capable of focusing on a task, neither sleepy nor fatigued, and able to concentrate and be motivated. Alertness is a key neuropsychiatric construct (cf. memory, impulse control, and mood), and, without a system of measurement, we cannot make progress in understanding the modifiers of alertness. Alertness is often presumed to be the opposite or inverse of sleepiness. We will show why we think this is not the case. Similarly, a sleepy person who is motivated to stay awake may, for a time, achieve that objective but subjectively will not feel alert. In many medical conditions, including psychiatric ones, there appears to be a change in the level of alertness evinced by the patient. Changes in alertness are particularly obvious in patients with sleep disorders. It also applies in general medicine with conditions such as liver or kidney failure and multiple sclerosis. There are a number of medications that may have an impact on alertness. Traditionally, these are agents classified as stimulants, but more recently, a distinction has been made with “alertness-enhancing agents” being distinguished by not giving “a buzz.” At this time, alertness is measured by subjective report and partly based on studies using the Maintenance of Wakefulness Test (MWT) or studies where one or two items of an existing scale (e.g., the Profile of Mood States Scale) have been used to deduce an alertness metric. The modification of alertness may have an impact on recovery from medical ailments and may have medicolegal implications. Without measurement instruments to evaluate alertness, we cannot begin to grapple with these issues.

Functionalists have resorted to defining mental concepts in terms of function. In this vein, alertness at any moment can be defined as the capacity of the mind to respond appropriately to external and internal stimuli. This obviously means that alertness can be considered to be measured by degrees. A person is very alert if he notices any minor changes in his environment and adjusts his behavior accordingly. Conversely, someone is not alert if he fails to notice and respond to something important happening in his immediate environment. This means that one can have a degree of alertness when asleep. A mother can respond to her baby's cries appropriately even when asleep but is still not considered to be alert because she minimally responds to anything else. Impaired alertness and daytime sleepiness are two of the most common reasons for attending sleep disorder clinics [2], [3]. Studies indicate that daytime sleepiness and alertness are not opposite poles of a single continuum (as is often implied in the literature) but distinct central nervous system phenomena with different neural pathways and neurochemical mechanisms [4], [5].

The accepted standard for objective assessment of sleepiness is the Multiple Sleep Latency Test (MSLT), consisting of a series of four or five 20-min nap opportunities [6]. These tests are spaced at two hourly intervals across the day in order to assess the rapidity with which the participant falls asleep in a standardized sleep-conducive setting. Participants recline in bed in a quiet, darkened room. They are instructed to close their eyes and to try to fall asleep. Although the MSLT is regarded as a direct measurement of the underlying physiological potential for sleep to occur [7], no clear relationship has been observed between MSLT results and other measures of behavior—such as the ability to perform on the critical flicker fusion test and other psychomotor tasks—in nonpatients who do not display pathological MSLT scores [7]. Neither has a relation been found between MSLT results and individual sleep parameters [8]. Subjective rating scales such as the Epworth Sleepiness Scale (ESS) [9] shows modest correlation with the biologic MSLT [10].

By contrast, the MWT is designed to evaluate the ability to remain awake during nonstimulating situations. It is often viewed as a variant of the MSLT but it is conceptually different [10], [11], [12] in that patients/participants are asked to try to stay awake in a darkened room. This distinctiveness is corroborated by the fact that MSLT and MWT results typically correlate weakly [4], [13].

By instructing participants to remain awake while sitting in a chair in a quiet dimly lit room, the MWT has face validity as an index of the tendency to fall asleep unintentionally during soporific activities where individuals are attempting to stay awake [14]. The MWT is used in many sleep laboratories and sleep clinics to examine patients with disorders associated with sleepiness and has been shown to be useful in detecting valid treatment effects [17].

The MWT is typically initially undertaken 2 h after awakening with four or five 20- to 40-min trials every 2 h throughout the day. The patient is instructed to stay awake. A standard montage is used for polysomnographic recording. Average sleep latency scores between 12 and 19 min are interpreted to indicate “moderately impaired alertness,” whereas scores under 12 min indicate “severely impaired alertness” [17].

In MWT studies, recording and scoring protocols with varied definitions of sleep onset and trial duration have been used [14], [15], [16].

The absence of normative information concerning the MWT until relatively recently and the fact that no standardized procedure has been adopted for using the MWT have led to a relative paucity of information as to how this test might best be used. We (inter alia) conducted a study to address this need and to gather preliminary normative data for the procedure [14]. Similar results have been found subsequently by an Australian group [18].

The Alpha Attenuation Test (AAT) [19], [20] is another physiological measure used to measure sleepiness. The AAT is estimated by calculating a ratio of eyes-closed to eyes-open EEG alpha power (power spectrum in the 14–18 Hz EEG). This measure is the Alpha Attenuation Coefficient and is considered an objective measure of physiological sleepiness [21], but no formal evaluation as to whether it may be more indicative of impaired alertness rather than a measure of sleepiness has been undertaken. It offers the potential of an instantaneous measure of alertness but has, to date, barely been used in field studies. We found that it was useful in a study of patients with narcolepsy [21].

There are no other widely accepted physiological measures of either alertness or sleepiness, although there are many tests of daytime performance [22], [23]. These tests often require motor dexterity on which an alert patient such as one with Parkinson's disease may do poorly, although, by a widely understood concept of alertness, such an individual may have no impairment. These issues have been recently reviewed [10].

Physiological tests, such as the MWT and AAT, are costly and require considerable time and resources. Comparatively, more efficient questionnaire measures are easier to administer, less expensive, and may offer valuable assessment information about the subjective experience of alertness not captured by the more labor-intensive physiological indices. Several scales to assess subjective experiences of daytime sleepiness currently exist, including the Stanford Sleepiness Scale (SSS) [24] and the ESS [9], [25]. To our knowledge, however, no scale to specifically measure the subjective experience of alertness currently exists.

The present article is predicated on the hypothesis that subjective alertness is independent of the experience of sleepiness. Because no measurement instruments have been developed specifically to address the problem of alertness, we developed two self-report questionnaires: the Toronto Hospital Alertness Test (THAT) and the ZOGIM-A [26].

The THAT is a rating scale of alertness that assesses a range of activities such as ability to concentrate, think of new ideas, or focusing on the task at hand. The ZOGIM-A describes a range of activities that might impact on alertness and inquires about how a person may function differently if he or she were “more alert.” Although both tests (see Appendix A Toronto Hospital Alertness Test, Appendix B ZOGIM-A) address the concept of alertness, the subtle differences in focus are intended to tap complementary facets of this important but understudied construct. It would also be informative to determine whether one of these two instruments was more closely related to an objective measure of alertness, namely, the MWT.

The THAT is a 10-item self-report index designed to measure perceived alertness during the past week. Respondents rate all items on a 6-point scale, ranging from 0 (not at all) to 5 (all the time). Items include the degree to which the respondent felt alert, fresh, energetic, able to concentrate, able to think of new ideas, clarity of vision for noting details, able to focus on the task at hand, and that mental facilities were at peak level. Two final items, “feeling that extra effort was needed to maintain alertness” and “in a boring situation I would find my mind wandering,” are reverse scored to control for acquiescence response set. All items are considered equally important, yielding an additive score ranging from 0 (very low alertness) to 50 (very high alertness). The instrument is presented in Appendix A.

The ZOGIM-A is a 10-item self-report scale designed to measure (a) the impact of various influences on subjective alertness (e.g., sleep loss, exercise, caffeine, vacation, forgetting worries), (b) the anticipated benefits of increased alertness (i.e., ability to organize day-to-day activities, completion of tasks, creativity, and frequency of mistakes), and (c) the proportion of daytime experience during which the respondent functions with a high level of alertness. Respondents rate each item on a 5-point scale, ranging from 1 (extremely) to 5 (not at all). Scores on all 10 items are added, yielding a single summative index, ranging from 0 (impaired alertness) to 50 (high alertness). The ZOGIM-A is presented in Appendix B. In formulating this scale, we recognized that it may comprise three independent facets. This awaits further analysis. Future studies will be required to show a relation (or lack thereof) of these scales and scales measuring fatigue and sleepiness in specific situations.

Both scales were developed at the University Health Network, Ontario, Canada, based on clinical observations and available literature. Initial testing in a group of 50 patients repeatedly assessed (on six occasions) with these two scales over a 6-month period showed that they were easy to use, required little instruction, and showed variation in scores across individuals, time, and treatment circumstances [27].