LIVING ALOFT: Human Requirements for Extended Spaceflight






[121] Various aspects of the space environment can adversely affect performance. Many of the environmental, physiological, and interpersonal factors potentially affecting the astronaut, and therefore his or her performance, are reviewed elsewhere in this text. Here, we concentrate on those factors which are directly related to his or her ability to perform work-related tasks.


Factors Affecting Work Capacity


Weightlessness-Several important aspects of work in space are directly affected by weightlessness. The accuracy of psychomotor performance in space is affected by the lack of the gravitational pull that normally weighs down a person on Earth. Simons (1964) has shown that during parabolic flight, the body tends to approach a fetal position as the limbs move toward equilibrium during relaxed posture. Such postural changes have been verified in spaceflight (Thornton, Hoffler, and Rummel, 1977). Because of this postural change, the ability to reach and position the arm and hand accurately can be adversely influenced. Until adaptation occurs, there is a tendency to overshoot a target to be grasped. However, since the phenomenon is recognized, accurate grasping responses are typically not included as mission-critical tasks, especially in the early phases of the mission.

As mentioned previously, fine motor movements may be adversely affected in 0-g.. This decrement in dexterity can pose potential problems for the manipulation of control panels, sensitive scientific experimental apparata, etc., although the problem tends to abate when adaptation to the 0-g environment has been achieved. Space system designers must ensure that switches are easy to [122] manipulate, do not require unnecessarily delicate tuning, and are positioned to allow maximum individual access without the chance of accidentally changing one setting while attempting to alter another.

The large metabolic cost required to conduct work in space is also an important aspect in assessing 0-g work capacity (see review by Christensen and Simons, 1970). Experience has demonstrated that metabolic rates during extravehicular activities can be truly dramatic, and in the early Gemini flights, some scheduled maneuvers had to be cancelled because of astronaut fatigue. Even for relatively easy tasks, increased effort is required. This extra effort is related in part to the unique character of working in space where, once momentum has been imparted, new action is required to terminate the original movement In addition, the astronaut finds it difficult to maintain his or her position in the absence of traction. Problems have also been encountered in moving when wearing a pressurized space suit (Covault, 1983). Although all of these issues have been effectively countered to some degree for short-term missions, the extent to which the increased workload may compound problems in more extended missions has not been determined.

Biomedical changes- Biomedical changes such as reduced cardiac activity and diminished musculoskeletal strength can affect work capacity. As discussed in chapter II, deterioration of these systems can jeopardize the health and functioning of crews and, if unchecked, can reduce the physical capacity of crews for extended high work output. For example, in-flight decreases approaching 10% were observed in exercise (vital) capacity in the Skylab 3 Pilot and the entire Skylab 4 crew (Sawin, Nicogossian, Rummel, and Michel, 1976). The extent to which countermeasures such as in-flight exercise, diet, and supplements will be sufficient to control these changes and to ensure optimal physical work capacity for long-duration missions is not clear.

Space sickness- The problems of space sickness pose an immediate threat to work capacity in space. Valuable mission time has already been lost during several flights because of the debilitating effects of space sickness, making this a high-priority concern for continued research.

One of the models used in space-related work capacity research involves the use of neutral buoyancy (submergence). Performance measurements taken during submergence have demonstrated many of [123] the same difficulties encountered in space. Performance involving manual manipulation (Streimer, Turner, Volkmer, and Guerin, 1970) and force-producing capabilities (Streimer, Turner, and Volkmer, 1966) are degraded in the underwater tractionless mode. Postural and equilibrium changes have also been noted. For instance, Morway, Lathrop, Chambers, and Hitchcock (1963) found that under some conditions subjects exhibited a significant bias in the upward direction when reaching for a target. This tendency may be analogous to the overreaching problem encountered in space (Simons, 1964), as previously noted.


Factors Affecting Work Schedules


During early spaceflight, it was assumed that duty schedules would be long and arduous; it was therefore necessary to investigate how wake-sleep cycles, other than the customary 24-hr day with its uninterrupted 8-hr sleep, might be applied to astronaut crews. This section outlines research pertinent to the issues of work, rest, and sleep cycles.

Altered wake-sleep cycles- When wake-sleep cycles are altered from a 24-hr pattern, this change automatically produces a displacement in the patterning of normal sleep periods-a factor known to adversely affect sleep quality and quantity. In turn, such disruptions in sleep are known to affect performance adversely (as discussed in a later section of this chapter). Considerable research indicates that subjects placed on alternative sleep and wake schedules experience sleep problems (Carskadon and Dement, 1975; Weitzman, Nogeire, Perlow, Fukushima, Sassin, McGregor, Gallagher, and Hellman, 1974). Webb and Agnew (1975) observed subjects placed on 2:1 wake/sleep ratios totaling 9,12,18,30, or 36 hr. Both long days and short ones proved to be inefficient cycles, albeit for different reasons. On the 9-hr day, subjects experienced difficulties falling asleep, and woke up frequently. Under the 36-hr condition, most subjects awoke early and were unable to use the full 12- hr sleep allotted them. Such inefficiencies in sleep patterns as found with abnormal day lengths are known to adversely affect performance when sustained for more than a few cycles (Webb and Cartwright,1978). But within the desired 24-hr day, the question remains how best to schedule work and rest (or sleep) activities. Much of the work focusing on these issues was done by Chiles, Alluisi, and Adams, who carried out more than a dozen separate investigations over a 10-yr period. Performance was assessed using the multiple-task [124] performance battery described previously in this chapter and outlined by Alluisi (1967,1969).

Altered work-rest schedules- One question concerns the relationship of work-rest schedules to performance. Two 15-day studies using repeated schedules requiring 4-hr work followed by 2-hr rest (4:2 schedule) and a 30-day study requiring 4-hr work followed by 4-hr rest (4:4 schedule) addressed this question (Adams and Chiles, 1961, 1963a, 1963b). These investigators found that subjects in the 4:4 condition consistently maintained better performance than either of the 4:2 groups. From these studies Chiles et al. (1968) conclude that subjects working 12 hr/day on a 4:4 schedule can maintain a generally high level of performance when compared with subjects working 16 hr/day on a 4:2 schedule, and can do so for a longer period. However, they did find some interesting exceptions. Arithmetic tasks were performed about equally well by all groups, and some highly motivated subjects could work 16 hr/day on the 4:2 schedule with essentially no decrements over a period of at least 15 days, and possibly for as long as 30 days.

Stress- Even with nominally acceptable work schedules, performance reserves may be diminished and crew safety endangered in emergencies or during periods of stress. Performance degradation has been demonstrated in studies of the combined effects of sleep loss and work-rest schedules (Alluisi et al., 1964; Chiles et al., 1968). Subjects working on the 4:2 and 4:4 schedules were given additional tasks to complete during their regularly scheduled rest periods during days six and seven of a 12-day study. These additional tasks prevented their sleeping for periods of 40 to 44 hr. The results show a general trend toward greater decrements in those tasks which required either sustained concentration or systematic shifts in attention or mental set. The results further indicate that subjects in the 4:2 condition showed greater performance decrements resulting from sleep loss compared with subjects working on the 4:4 schedule, and did not recover as well following the sleep-loss period. These results strongly suggest that the more demanding the schedule, the more severe the performance impairment both during and after the depletion of performance reserves by the stresses of lost sleep and additional workload.

It seems clear that a high work/rest ratio is not likely to be practical for long-duration missions. We need to determine what specific work-rest schedules would be best over extended durations. Factors to be considered include the type and amount of the work to be [125] accomplished, the types and extent of environmental stressors present, the length of the mission, the level of task proficiency required by mission standards, etc. Several of these variables are considered in the following sections.


Factors Affecting Workload


Klein (1970) refers to workload as the requirements of task performance that can be specified without reference to any operator response or effort actually applied to satisfy those requirements. A clear distinction is made between the demands for action implied in the tasks and the capabilities necessary for effective response (see also Chiles and Alluisi, 1979). A second conceptualization of workload focuses on how much the operator has to do and/or how hard he or she must work to satisfy a set of demands (Cooper and Harper, 1969; Jenney, Older, and Cameron, 1972). A related emphasis is provided by Welford (1968) in characterizing effort as the intensity of work. A final conceptualization of workload emphasizes effort and accomplishment. Here the focus is on the total activity of the operator in completing the required tasks (Cantrell and Hartman, 1967).

For purposes of the present discussion, workload will be defined broadly as the total work demand placed on the operator, including the amount and intensity of operator effort required to accomplish the task. If we think of a continuum of workload, we can assume that some optimal work load exists for a given individual at a given level of skill. Thus, the optimal load varies within certain limits according to individual differences in capacity, skill, and reactions to situational factors. The extremes of this continuum have adverse implications for performance and they are of primary concern to us here.

Fatigue- If we assume that for a given individual there is an optimum level of task demand, then we may define "overload" as an increase in the task demand beyond the individual's optimum level. Similarly, when the demands of the task are lowered below some critical level, task underload results. Many investigators have attempted to unify research on the effects of these two extreme loading conditions through the hypothetical construct "fatigue." As with workload, there is no single definition of fatigue that is universally accepted. As Gartner and Murphy (1976) point out, at least five different conceptualizations of fatigue are present in the performance literature. We will employ one of these conceptualizations:

[126] Fatigue represents a reversible impairment of performance as a consequence of over- or underloading. Here, fatigue takes the form of lowered sensitivity, responsiveness, or capacity, which may show itself as a reduction in the amount of information that can be handled at any one instant, or in the amount that can be handled over a given period.

In structuring the work regime of the astronaut, the goal is to maintain an optimal workload so that fatigue does not occur. In the event that the reserve capacities of the astronaut are depleted because of either overloading during emergencies (e.g., continuous performance of highly complex tasks) or underloading during mission cruise phases (when few task demands are placed on the operator), we can anticipate that fatigue may be a problem. Gartner and Murphy (1976) suggest four broad categories of behavioral deterioration associated with such over- or underloading: motivation decrement, skill or proficiency decrement, psychological stress, and performance decrement. Fatigue can both slow performance and result in irregular or disordered performance (see review of the Effect of Loading, in Welford,1968).

One review (Beljan, 1972) of complex task performances of fatigued subjects revealed the following degradations: (1) range of attention is restricted; (2) responses to signals are made independently of the occurrence of the signals; (3) mood is changed, and impatience occurs or increases; and (4) orientation becomes less acute.

It is clear that fatigue is a potentially serious problem for astronaut crew performance and well-being. Management of this condition can be approached in several ways. The traditional approach used with airline crews has emphasized adequate rest, physical fitness, weight control, nutrition and diet, and moderation in the use of alcohol and tobacco in order to offset the negative effects of over- or underloading (Schreuder, 1966). Pharmacological agents, especially various amphetamine-like drugs, have also been used to reduce astronaut fatigue (Parin, Kosmolinskiy, and Dushkov, 1970). However, as discussed in chapter II, the long- term effects of drugs in space are not well understood.

There is an obvious need to explore how mission activities might be integrated to minimize fatigue, both with respect to the absolute amount of work required and to the sequential processing of tasks involving various levels of complexity and effort. The [127] delicacy of balancing workload was demonstrated in certain problems encountered in Skylab, where mission programmers may have planned too busy a schedule in their attempt to prevent the adverse effects of underloading (Cooper, 1976). Such a situation invites danger if a continuous misloading results in significant deterioration in performance levels.

It is also important to understand the proper sequencing of tasks. Simple tasks can be performed effectively at much higher levels of fatigue than more complex tasks (Welford, 1968). In designing a daily schedule, it might prove useful to assign complex tasks only at the beginning of the work day or following substantial rest periods. Simple tasks could then be assigned to follow complex tasks or could be completed at the end of the day when accumulated fatigue might be expected to be greatest.

From our review thus far, we can see that the type of work-rest schedule most beneficial for space missions must depend a great deal on the degree of loading involved in the tasks. Chiles and Adams (1961) posed the following guidelines regarding the duration of work periods as a function of loading:

Each of these recommendations will no doubt have relevance to various phases of extended missions. It would be desirable to develop a taxonomy of tasks and duty hours as they apply to each crew person at different points during the flight.

Thus far, we have assumed that over- and underloading result in the same state-"fatigue." However, various researchers take exception to this position. For example, Welford (1968) defines fatigue as resulting from conditions of overloading, but prefers the concept of [128] "monotony" as related to conditions of underloading. Monotony can occur when tasks require little cognitive processing or attention, are highly repetitious, are not complex, and/or have been extensively overlearned. Under these conditions, Welford predicts that performance may be generally poor as with overloading, but that the nature of performance degradation is fundamentally different. When the operator is overloaded, action is confused and judgments or graded responses tend to be undifferentiated. In contrast, when the operator is underloaded, attention tends to drift, signals are missed, and performance is lethargic. Whether or not fatigue and monotony are fundamentally different states, underloading related both to the task and the environment can be a potentially serious problem during extended missions.

The most extensively studied condition of task underloading involves the requirement for visual or auditory vigilance. One of the earliest of the laboratory studies of this phenomenon is the test employed by Mackworth (1950), where subjects attend to the pointer movements of a clock and respond appropriately to randomly presented anomolous movements. This and similar tasks, using a wide variety of signals such as a faint spot of light (Wilkinson, 1961), small changes in sound (Mackworth, 1950), or occasional features interspersed in a regularly presented series of digits (Bakan, 1953) have indicated significant performance decrement over relatively short periods of time. The decrements are shown not only by missed signals, but by slower responses to those that are detected (Walks and Samuel, 1961; McCormack, 1962; Surwillo and Quilter, 1964; Buck, 1966). Detection failures do not appear to be the result of looking away from the signals, since Mackworth, Kaplan, and Metlay (1964) have shown that signals are often missed even when the subject's attention is fixed on them. Nor is the performance decline the result of a simple lack of activity. Whittenburg, Ross, and Andrews (1956) repeated Mackworth's clock test, but required subjects to respond to every jump (making a different response for small and large jumps), and found that performance continued to show declines even when warnings of signals were given.

Performance decrements in response to routine tasks occur in a variety of situations. For example, Adams and Chiles (1961) found that performance of monitoring and vigilance tasks indicated decrements over a 15-day confinement period, and Altman and Haythorn (1967) found poor performance on monitoring tasks among two-man groups confined for 10 days. Burns and Gifford (1961) found that decrements occurred when subjects had to operate routine tasks [129] without apparent consequence. However, when the vigilance tasks were embedded in a multiple-task performance battery, and occurred with active tasks like arithmetic computations, performance appeared to be no different than for the active tasks. If anything, the vigilance performances deteriorated less and recovered more quickly than did active task performance (Alluisi, Coates, and Morgan,1977).

It is clear that we must guard against the problem of task underload if optimal performance is to be attained and sustained in space. There appear to be several ways of minimizing the decrement in vigilance. Duty periods should be no more than 30 min if monitoring or vigilance comprises the sole task or a major part of the job to be accomplished within that time frame. Making the display and monitoring signal more intense appears to be a useful aid (Mackworth, 1950; Broadbent, 1958). Periodic changes of activity (Bevan, Avant, and Lankford, 1967), brief rest periods (Bergum and Lehr, 1962), or the presence of others in the room (Fraser,1953; Bergum and Lehr, 1963; Williams, Kearney, and Lubin, 1965) all help to sustain satisfactory vigilance levels.

In considering the design and scheduling of tasks to be accomplished by spacecrews, it will be important to determine whether task underload may become a problem during particular phases of the mission. Although certain tasks, such as those requiring monitoring or vigilance, could give rise to underloading problems during flight, it is more likely that underloading difficulties will be related to the sheer repetitiveness of the tasks demanded of spacecrews, and the degree to which they are so overlearned as to no longer be interesting.

Uniform environmental conditions-Although the term "loading" generally refers to the demands of the task, a broader definition encompassing environmental stimulation can also be useful. Research strongly indicates that performance can be expected to deteriorate when the environmental input is at a reduced level. Studies of extreme sensory or perceptual deprivation conducted primarily in the 1950s and 1960s (see Zubek, 1974 for overview and synthesis) demonstrated significant impairment in certain types of performance.

Bexton, Heron, and Scott (1954) and Scott, Bexton, Heron, and Doane (1959) found that performance on measures of verbal fluency, anagrams, and various numerical tasks was significantly impaired during several days of perceptual deprivation. No deficits [130] were found for measures of digit span, analogies, or associative learning. Zubek, Aftanas, Hasek, Sansom, Schludermann, Wilgosh, and Winocur (1962) report that, during 7 days of perceptual isolation, impairment occurred in arithmetic problem-solving, numerical reasoning, and recognition, but no significant changes were found in digit span, rote-learning, recall, and verbal reasoning. Impairment of visually cued reaction time following 2 days of perceptual deprivation has been demonstrated in studies by Nagatsuka and Suzuki (1964). Further, auditory (Zubek, Pushkar, Sansom, and Gowing, 1961) and visual (Zubek et al., 1961, 1962) vigilance have deteriorated during experiments involving low levels of sensory input. These results confirm that a reduced or nonvarying work environment can produce dramatic decrements in various aspects of performance.

Less extreme deprivation experiments have also shown that isolation in a nonvarying environment can sometimes adversely affect performance. For example, space-cabin simulator studies conducted at the USAF School of Aviation Medicine (Flinn, Monroe, Cramer, and Hagen, 1961; McKenzie, Hartman, and Welch, 1961; Morgan, Ulvedal, and Welch, 1961) found that two-man groups isolated for 17 days or for 30 days showed a gradual increase in average response time to a 14-task multiple-test battery. Similar findings demonstrating minor changes in performance have been reported from sealed-submarine experiments (Fawcett and Newman, 1953), fallout-shelter studies (Hammes, 1964), and in laboratory-isolation investigations (Altman and Haythorn, 1967).

Subjective reports of crews stationed in the polar regions frequently include complaints that long wintering-over periods affect memory and concentration (Mullin, 1960), but no systematically obtained data involving actual mission tasks have been presented to verify or deny that significant performance impairments have occurred. However, there are indications that long-term environmental underloading can disrupt performance. For example, crews in the polar regions have demonstrated markedly low involvement with intellectual activities, and many persons slated for isolated duty indicate an intent to occupy themselves with intellectual pursuits, but few people actually follow through (Rohrer, 1961; Gunderson and Nelson, 1963). It seems probable that some degree of performance attenuation will occur among crews isolated for long periods, perhaps owing to decrements in morale and motivation rather than to any cognitive or psychomotor deterioration. Research involving limited environments is certainly warranted in order to explore more [131] fully just how an unchanging environment affects performance under long-term conditions.

Variety of work available- As Kubis and McLaughlin (1967) summarize, overlearning of the mission duties; extensive training in the diagnosis of a means to deal with simulated system failures; and comprehensive training in egress, escape, and survival techniques all have been emphasized as means of building confidence, reducing and controlling anxiety, and minimizing the potential problems that might occur in flight. Without question, this approach has been a valuable facet of astrononaut training, given the emergencies that have occurred and the skill with which the astronauts have responded. However, in anticipation of future longer-duration missions, some attention should be devoted to the potential problems that such a comprehensive training approach may generate.

Given the extreme importance that isolated individuals place on stimulating, challenging, and meaningful work, we must ensure that the training program, with its rigorous overtraining goals, does not reduce the interest value of mission responsibilities, adding to the potential for monotony and boredom in a long-term mission. Although crews need to be fully prepared, we must remember that being well trained exacerbates the dangers of boredom and makes the designing of jobs to be interesting and intrinsically motivating even more important. Apparently we cannot count on the anticipation or excitement of spaceflight to offset monotony; in the Mercury project, Gordon Cooper fell asleep inside the capsule during a long hold before liftoff (Wolfe, 1979).

One strategy for addressing monotony and boredom in spaceflight might be to overtrain crewmen on those tasks that are crucial for mission safety and success, while leaving noncritical procedures to be scheduled for in-flight learning. If it were possible to schedule some training exercises and academic learning during the flight itself, the additional challenge and stimulation of meaningful work would be available. This possibility would seem particularly appropriate to the cruise phases of long-duration missions, where only routine duty tasks are required, where the environment may be relatively unstimulating, and where few important mission goals are involved.

Crewmembers might also be given greater flexibility in determining their own work schedules. Both American and Russian space crews have repeatedly asked to have some options in the day-to-day organization of their work and leisure schedules-options allowing [132] them more control over the effectiveness of their work (Cooper, 1976; Gazenko, Myasnikov, loseliani, Kozerenko, and Uskov,1979). American crews have argued that rigidly structured schedules lead to confusion and as a result, experiments may not be conducted as carefully as desired. A varied approach to work scheduling seems worthy of trial; however, there appear to be no plans to institute this type of arrangement in the Shuttle Program (Morris, 1975; Watters and Stead man, 1976).

One further possibility for varying crew responsibilities is the rotation of crew duties. If each crewmember is held responsible for the same duties during an extended flight, the chances of boredom will increase; rotating the assignment of noncricital duties could be used to minimize the chances of performance decline. Research is needed to investigate this and the other possible performance- enhancement techniques.


The Effects of Desynchronosis


It is well documented that many functions of the human body, including temperature, metabolic rate, and endocrine output among others, are regulated on a near 24-hr basis (Conroy and Mills, 1970; Colquhoun, 1971). Alterations in these cycles occur on Earth in unusual environments or through displaced or disrupted sleep or work, with potentially detrimental effects. Disruption in the body's circadian activity, known as desynchronosis, can produce physical symptoms such as malaise, insomnia, appetite loss, and nervous stress (Hauty and Adams, 1966a,1966b,1966c).

Desynchronosis problems associated with sleep or work schedules have been observed in space. For example, up to and including the Apollo 9 mission, it was common practice to employ staggered sleep schedules for crewmembers. This scheduling resulted in shifts of as much as 6 to 10 hr from the Earth-based time during which sleep normally occurred. Such disruption of the established cycle was described as producing "a most unsatisfactory situation in flight" (Berry, 1969). Even during later flights when crewmembers slept simultaneously, hectic schedules required displacement of the sleep phase by several hours, leading to complaints of fatigue and sleep problems. For future extended missions, the problems of adjusting work and sleep schedules to the demands of mission requirements will be an issue of importance. We need to understand how alterations in circadian rhythm can affect performance and what potential areas of research should be favored.

[133] Studies of desynchronosis on Earth have included a number of different situations and techniques. Experiments with constant illumination (Webb and Agnew, 1974), day lengths extending beyond 24 hr (Webb and Agnew, 1975), industrial shift workers whose daily wake-sleep cycle is often displaced from the norm (Aschoff, 1978; Reinberg, Vieux, Ghata, Chaumont, and La Porte, 1978; Monk, Knauth, Folkard, and Rutenreanz, 1978), bedrest patients (Winget, Vernikos-Danellis, De Roshia, and Cronin, 1974; Winget, De Roshia, and Sandler, 1979), etc., have all contributed to our understanding of desynchronosis. Alluisi and Morgan (1981) recently reviewed the effects of temporal factors on human performance and productivity. Although a discussion of each of these topics is beyond the scope of this book, one clearly relevant area, transmeridian flights, will be reviewed. The transmeridian flight situation is similar to that experienced by astronauts, in that phase shifts accompany translocation. Also, the published studies provide performance data that permit some generalization to the spaceflight situation.

Transmeridian flight- In 1966, Hauty and Adams used reaction time, decision time, critical flicker fusion, and numerical ability to assess the effects of translocation of subjects to a new time zone. Measurements taken following long westward flights showed that reaction time and decision time were both significantly impaired during the first day. Although similar measures taken following eastward flights tended to show decrements, the observed impairment did not prove to be statistically significant.

More extensive studies have been conducted by Klein. In one study, Klein, Bruner, Holtman, Rehme, Stolze, Steinhoff, and Wegmann (1970) measured pilot performance in a flight simulator. Pilots were tested in Germany, then in the United States, and again in Germany. Measurements of the magnitude of performance alteration showed that the change in performance level was more adversely affected and of higher significance following eastward flight from the U.S. to Germany than following westward flight from Germany to the U.S. In a second study (Klein, Wegmann, and Hunt, 1972), eight subjects were measured, again following both westward and eastward flights. In this experiment, the performance measures used included a reaction time test, a symbol cancellation test, a digit summation test, and a complex psychomotor performance test known as the Kugel test. Klein et al. (1972) found that after eastbound flight, it took 12 days for performance on the Kugel test to recover, whereas it took 9 days for the simpler performance tasks to readapt. Recovery [134] time for the Kugel test following westward travel was 10 days, whereas the simpler tasks required 6 days.

Klein and his colleagues do not believe that there is any difference in psychophysiological response due specifically to the direction of flight. Rather, they ascribe the greater performance decline found in their studies in eastward travel to factors which, for a variety of reasons, lead to greater fatigue. Although some suggestions have been offered for distinguishing fatigue effects from circadian effects, it remains extremely difficult to separate the two when abnormal sleep is involved (Dodge, 1982), and in most studies to date, fatigue and circadian effects are compounded. (For a discussion of this and related issues, see Alluisi, 1977.)

From these and other transmeridian flight studies (e.g., Fort, 1969; Conroy and Mills, 1970; Berkhout, 1970; Wright, Voget, Sampson, Knapik, Patton, and Daniels, 1983) several inferences can be drawn regarding the relationship between desynchronosis and performance. First, decrements in performance are found for a range of behaviors from simple reaction time to cumulative exercise capacity Second, the more complex the task, the more detrimental is desynchronization upon task- completion efficiency. Whereas simple tasks in both the Hauty and Adams (1966a, 1966b) and the Klein et al. (1972) studies were affected by shifts in circadian rhythm, the greater deterioration was clearly associated with complex tasks. Third, wide individual differences in adaptability occur. Some individuals may adapt to a phase shift within 1 day, and others appear to adapt only after extended periods of time (Strughold, 1952). Finally, the majority of research indicates that although performance is affected by desynchronosis, the time course for adaptation is less than that for the body rhythms themselves.

Some important research questions can be drawn from these studies. For example, with few exceptions, there have been essentially no studies of the combined effects of monotony, desynchronosis, and other temporal factors (Alluisi and Morgan, 1981), conditions that are likely to prevail in spaceflight. Another issue involves the possible differential effects associated with flights headed in different directions, and the underlying mechanisms of circadian-rhythm phase shifting. Controlling for fatigue would help to isolate potential direction effects. Also important is the question of individual differences in phase-shift adaptability. If, as it appears, wide differences exist in the degree to which individuals can adjust to [135] changes in circadian rhythms, it is important to understand the behavioral and physiological correlates of these differences.

Zeitgebers- A factor related to circadian rhythmicity concerns those variables that regulate the 24-hr clock. Such variables are referred to as zeitgebers and consist of a wide range of physical, temporal, and social "cues" that serve to entrain sleep and wakefulness to a particular rhythm. Very few physical cues to diurnal cycles exist in space. Story Musgrave, Mission Specialist on STS-6, reports that time has little meaning in spaceflight except as it relates to elapsed time or work schedules (personal communication). When zeitgebers are not present, certain rhythms become "free running" and as a result may vary from a 24-hr schedule. Perhaps the most significant diurnal cue is the day/night cycle. In space, this cue is at best lacking, and at worst it can introduce distortions by driving rhythms in deleterious ways. For instance, the path of the orbiting Space Shuttle exposes crews to multiple sunrises and sunsets in a single Earth day. Before we attempt to reproduce or replace Earth zeitgebers under spaceflight conditions, we must determine which cues are involved with what specific rhythms. We can assume that artificial lighting will serve as a strong cue to periodicity. Also, social cues revolving around daily meals, work-rest schedules, and evening leisure activites can be important. Indeed, some research indicates that social cues may be even more important than lighting conditions for regulating work performance (Aschoff, 1978). Given that circadian rhythms and the issues that surround them have important implications for performance in space, methods of entraining rhythms to the most advantageous wake-sleep cycles constitute an important research area.

Performance as a circadian function-The performance levels of different tasks have been found to fluctuate during the day. It appears that performance may be subject to cyclic regulation of circadian periodicity in much the same way as many physiological functions. This finding has important implications for operators in space.

For example, performance as represented by the average number of simple sums computed in a 48-min period follows a cyclic pattern coincident with diurnal changes in body temperature (Colquhoun, Blake, and Edwards, 1968). Similar patterns have been observed for auditory vigilance tasks measuring the average number of signals detected and the average response time. Also, performance rhythms have been found within the workday portion of the 24-hr cycle.

[136] Kleitman (1963), regarded as a leading authority in this area, has concluded that there is a recognizable general pattern in the fluctuations of performance levels throughout the work day which in many cases can be related to diurnal variations in body temperature. Results indicate that performance on many tasks shows a morning rise with a peak somewhere in the afternoon, followed by a fall. Although the exact timing of these points is disputed by various investigators (see Colquhoun, 1971, for example), all appear to agree that there may be recognizable patterns in the fluctuations of performance levels during the work day. In their review of temporal factors in human performance and productivity, Alluisi and Morgan (1981) conclude that whether or not performances show the circadian fluctuations, the underlying physiological rhythm is a bodily factor that must be taken into account. This relationship poses some important research questions for space mission planners seeking to optimize work output and accuracy.

For example, from a rhymicity perspective, it may prove advantageous to schedule delicate and sensitive tasks after the "morning rise," and to relegate less demanding or less critical tasks for the period following the "afternoon fall." This recommendation fits in well with our discussion of task overloading/underloading and the need to balance work load across the day. Such a schedule would require constructing a taxonomy of tasks most appropriate for different phases during the daily work schedule, which in turn requires a greater understanding than is currently available of what types of requirements and tasks are most affected by circadian periodicity.


Sleep Disturbances


Implicit in the study of effective work-rest conditions for spaceflight is the need to identify relevant factors associated with sleep, and to determine how these factors may affect performance during waking hours.

In confined, isolated environments such as habitation in the polar regions, sleep difficulties have been a repeated problem. For example, Natani, Shurley, Pierce, and Brooks (1970), commenting on the rather severe alterations in sleep observed among the early Antarctic exploring parties, noted that conditions of insomnia and changes in the quality and pattern of sleep occurred, although the actual quantity of sleep did not vary. Such findings have been confirmed in investigations of personnel stationed for many months in Arctic and Antarctic outposts. Here, sleep quality deteriorates and [137] the men suffer from a dramatic form of insomnia in which the dream phase of sleep declines significantly, the deeper stages of sleep decrease or disappear altogether, and only the lightest stages of sleep are actually reached. The restorative benefits of sleep may be severely limited under such conditions. The subjective reports that are associated with such disturbances indicate that persons are physically uncomfortable, withdrawn, easily annoyed, and preoccupied with vague physical complaints, and have varying degrees of depression.

Findings of quantitative and qualitative sleep loss in other isolated and confined conditions suggest that it is important to determine how sleep may be affected in the confined, isolated quarters of space. A basic question is whether "normal" sleep patterns can ever be achieved in space. The astronauts' often choppy schedules may decrease sleep duration or displace the time when sleep occurs. For example, during the Gemini VII mission, both crewmen felt fatigued, yet averaged only 5.3 hr of sleep a night throughout the flight. They slept less than 5 hr a night during the final 4 days (Berry, Coons, Catterson, and Kelly, 1966), and sleep was particularly poor on the first night (a fact reported by numerous astronauts). These problems appear to be related to high noise and vibration levels, general tension during the mission, staggered sleep schedules, and bed designs that were inadequate to ensure comfortable sleeping positions. Similar difficulties have been reported by the cosmonauts (Oberg, 1981).

During the later Apollo missions, the amount and quality of sleep were still unresolved issues. Certainly there were occasions when crews experienced sleep difficulties, but these difficulties may be explained by the prevailing physical conditions. For example, the crew of Apollo 14 experienced a large displacement in the normal terrestrial sleep cycle, a phase shift of some 7 to 11.5 hr later than they were used to. The crew had difficulty sleeping, probably owing to the lack of kinesthetic sensations and to muscle soreness in the legs and lower back. Throughout the mission sleep was intermittent and never exceeded 2-3 hr of deep and continuous sleep. In the lunar module, crewmen slept little. Difficulty in finding a place to rest their heads, discomfort of the pressure suit, and a 7° tilt of the module, all combined to make sleep difficult (Strughold and Hale, 1975).

Some encouraging results can be noted when the results of the Skylab missions (Frost, Shumate, Salamy, and Booher, 1977) are examined. During the 59- and 84-day missions, there was no [138] significant decrease in the average amount of sleep time. A 1-hr decrease in sleep time was observed on the 28-day mission, but this was strictly voluntary and not due to insomnia. Sleep latency, or the length of time needed to fall asleep, was relatively long in the early portions of the 84- day mission, but returned to values typical of pre and postflight conditions by the latter half of the mission. No significant change in sleep latency was seen in the 59-day mission, whereas the in-flight latencies were actually significantly lower than preflight latencies during the 28-day mission.

Significantly, the most marked changes in sleep occurred not in flight but upon return to Earth (with alterations more of quality than quantity). This suggests that readaptation to a 1 g environment may be more disruptive to sleep than adaptation to a 0-g environment.

Performance correlates of sleep loss- Against this background of sleep disturbances in space, let us consider how sleep loss on Earth has affected performance. The two-man Space Cabin Simulator Studies (Cramer and Flinn, 1963) showed that sleep loss directly affects the performance of tasks requiring sustained concentration or vigilance. Performance also can be affected indirectly, since there are strong relations among mood, attitude, motivation, and performance efficiency. Other studies have confirmed these results. For example, Wilkinson (1969) kept a group of men awake at night for various lengths of time over a 6-wk period. After each experimental night, the men were kept busy for most of the next 15 hr while auditory vigilance was tested at five different intervals lasting 1-hr each. The results indicated that following a full night's sleep (7.5 hr), about 65% of the signals were correctly detected. When a sleep debt of 5.5 hr occurred (i.e., sleep lasted only 2 hr), performance during the following day was significantly reduced. Interestingly, when sleep was reduced to 5 hr/night for two consecutive nights, a similar deterioration in performance was noted. These findings indicate that even relatively minor sleep loss, when accumulated over several evenings, can adversely affect performance in a manner similar to acute sleep loss on a single night. Wilkinson also found that sleep loss had a significant impact on decisionmaking as measured by a five-choice display panel task. A single night without sleep reduced the rate of response and increased the latency of response. Similar results have been found for tasks involving psychomotor performance on a rotary pursuit task and on a reversed-digit writing task (Chambers, 1964).

[139] In another study, Wilkinson (1963) found that a night of sleep did not entirely compensate for a night without sleep. Twenty-four men were deprived of sleep and required to keep busy for 24 hr, the sleep deprivation portion of the experiment ending in early evening. Most of the men then slept for about 2 hr and later went to bed at the usual time, resulting in about 9.5 hr of sleep by the second day. The subjects were then tested on a vigilance monitoring task or on the five-choice decision task. With both tasks there was a significant deterioration in performance compared with the control condition after normal sleep. The effect was more pronounced during morning trials than during afternoon trials, indicating that by the end of the second day the effects of sleep loss from the previous evening had begun to dissipate. In addition, Alluisi, Coates, and Morgan (1977) present data clearly showing that the effect of sleep loss interacts with the circadian rhythm, as does the recovery from sleep loss.

Sudden awakening-One final issue regarding the relationship between sleep and performance entails the ability of subjects to function optimally immediately upon awakening. During some of the past space missions, events have occurred that required the crew to arise from sleep and effect repairs immediately. How good is performance immediately after sudden awakening? If less than optimum, what is the time required to reach normal waking levels of performance?

Langdon and Hartman (1961) investigated this problem by assessing the performance task of turning off lights at a desk console. Subjects were awakened either at midnight or between 3 and 4 a.m. In this study, the subjects required not less than 7 min after being awakened to achieve normal efficiency in performing the task. Similar findings are reported by Webb and Agnew (1964), who awakened subjects from the deepest stage of sleep. These and other findings (Wilkinson and Stretton, 1971 ; Kleitman, 1949) suggest that performance upon first awakening is degraded, a point which should be taken into account in scheduling activities, when possible.

From the results reported above, we must conclude that sleep disturbances could pose a significant hazard to spacecrews. Efforts must be made to minimize such disturbances and to ensure that wake-sleep cycles do not adversely affect sleep. There are some unanswered questions surrounding the issue of sleep as affected by weightlessness. A fundamental question concerns how, over the long term, sleep requirements change in space. Although early mission [140] crews experienced great difficulties in sleeping, the later crews of Skylab reported very few problems. In fact, some of the Skylab astronauts reported that they thought sleep was actually better in space than on Earth, since they required fewer hours of sleep and felt less fatigue during the workday. Perhaps the lack of gravity and consequent diminution in the forces acting on the body (particularly the muscles) may be responsible for this interesting effect.

Relevant and potentially important research can be conducted on sleep issues without going into space. For several years levitation via airflow (Scales, Winter, and Block, 1965) has been used for the treatment of hospitalized severe burn victims (Whitaker, Graham, Parsi, and Olson, 1976; Demling, Perea, Maly, Moylan, Jarrett, and Balish, 1978). Since the degree of pressure placed on the body using air is minimal, air levitation might provide a good analogy to the sleeping conditions of weightlessness. Less exotic, but also potentially useful, would be similar research employing water levitation techniques.