Chapter 6 Chronic Sleep Deprivation

Abstract

Chronic sleep deprivation, also referred to as chronic sleep restriction, is common, with a wide range of causes including shift work and other occupational and economic demands, medical conditions and sleep disorders, and social and domestic responsibilities. Sleep dose–response experiments have found that chronic sleep restriction to less than 7 hours per night resulted in cognitive deficits that (1) accumulate (i.e., become progressively worse over time as sleep restriction persists), (2) are sleep-dose sensitive (i.e., the less sleep that is obtained, the faster the rate at which deficits develop), and (3) do not result in profound subjective sleepiness or full self-awareness of the cumulative deficits from sleep restriction. The mechanisms underlying the sleep dose–response cumulative neurobehavioral and physiologic alterations during chronic sleep restriction remain unknown. Individual variability in neurobehavioral responses to sleep restriction appear to be as large as those in response to total sleep deprivation and as stable over time, suggesting a traitlike (possibly genetic) differential vulnerability to the effects of chronic sleep restriction or differences in the nature of compensatory brain responses to the growing sleep loss.

Chronic sleep restriction occurs frequently and results from a number of factors, including medical conditions (e.g., pain), sleep disorders, work demands (including extended work hours and shift work), and social and domestic responsibilities. Adverse effects on neurobehavioral functioning accumulate as the magnitude of sleep loss escalates, and the result is an increased risk of on-the-job errors, injuries, traffic accidents, personal conflicts, health complaints, and drug use.

Chronic sleep restriction, or partial sleep deprivation, has been thought to occur when one fails to obtain a usual amount of sleep.¹^,² Half a century ago, Kleitman first used the phrase sleep debt to describe the circumstances of delaying sleep onset time while holding sleep termination time constant.³ He described the increased sleepiness and decreased alertness in individuals on such a sleep–wake pattern, and proposed that those subjects who were able to reverse these effects by extending their sleep on weekends were able to “liquidate the debt.”³, p. 317

The term sleep debt is usually synonymous with chronic sleep restriction because it refers to the increased pressure for sleep that results from an inadequate amount of physiologically normal sleep.⁴ To determine the effects of chronic sleep loss on a range of neurobehavioral and physiologic variables, a variety of paradigms have been used, including controlled, restricted time in bed for sleep opportunities in both continuous and distributed schedules,⁵ gradual reductions in sleep duration over time,⁶ selective deprivation of specific sleep stages,⁷ and limiting the time in bed to a percentage of the individual’s habitual time in bed.⁸ These studies have ranged from 24 hours⁹ to 8 months⁶ in length.

Many reports published before 1997 concluded that chronic sleep restriction in the range commonly experienced by the general population (i.e., sleep durations of less than 7 hours per night but greater than 4 hours per night) resulted in some increased subjective sleepiness but had little or no effect on cognitive performance capabilities. Consequently, there was a widely held belief that individuals could “adapt” to chronic reductions in sleep duration, down to 4 to 5 hours per day. However, nearly all of these reports of adaptation to sleep loss were limited by problems in experimental design.⁸ Since 1997, experiments that have corrected for these methodological weaknesses have found markedly different results from those earlier studies, and have documented cumulative objective changes in neurobehavioral outcomes as sleep restriction progressed.¹⁰ This chapter reviews the cognitive and neurobehavioral consequences of chronic sleep restriction in healthy individuals.

Incidence of Chronic Sleep Restriction

Human sleep need, or, more precisely, the duration of sleep needed to prevent daytime sleepiness, elevated sleep propensity, and cognitive deficits, has been a long-standing controversy central to whether chronic sleep restriction may compromise health and behavioral functions. Self-reported sleep durations are frequently less than 8 hours per night. For example, approximately 20% of more than 1.1 million Americans indicated that they slept 6.5 hours or less each night.¹¹ Similarly, in polls of 1000 American adults by the National Sleep Foundation, 15% of subjects (aged 18 to 84 years) reported sleeping less than 6 hours on weekdays, and 10% reported sleeping less than 6 hours on weekends over the past year.¹² Scientific perspectives on the duration of sleep that defines chronic sleep restriction have come from a number of theories.

Theoretical Perspectives on Sleep Need and Sleep Debt

Basal Sleep Need

The amount of sleep habitually obtained by an individual is determined by a variety of factors. Epidemiologic and experimental studies point to a high between-subjects variance in sleep duration, influenced by environmental, genetic, and societal factors. Although not clearly defined in the literature, the concept of basal sleep need has been described as habitual sleep duration in the absence of preexisting sleep debt.¹³ Sleep restriction has been defined as the fundamental duration of sleep below which waking deficits begin to accumulate.¹⁴ Given these definitions, the basal need for sleep appears to be between 7.5 and 8.5 hours per day in healthy adult humans. This number was based on a study in which prior sleep debt was completely eliminated through repeated nights of long-duration sleep that stabilized at a mean of 8.17 hours.¹⁵ A similar value was obtained from a large-scale dose–response experiment on chronic sleep restriction that statistically estimated daily sleep need to average 8.16 hours per night to avoid detrimental effects on waking functions.⁴

Core Sleep versus Optional Sleep

In the 1980s, it was proposed that a normal nocturnal sleep period was composed of two types of sleep relative to functional adaptation: core and optional sleep.¹⁶^,¹⁷ The initial duration of sleep in the sleep period was referred to as core, or “obligatory,” sleep, which was posited to “repair the effects of waking wear and tear on the cerebrum.”^{16, p. 57} Initially, the duration of required core sleep was defined as 4 to 5 hours of sleep per night, depending on the duration of the sleep restriction.¹⁶ The duration of core sleep has subsequently been redefined as 6 hours of (good-quality, uninterrupted) sleep for most adults.¹⁴ Additional sleep obtained beyond the period of core sleep was considered to be optional, or luxury, sleep, which “fills the tedious hours of darkness until sunrise.”^{16, p. 57} This core versus optional theory of sleep need is often presented as analogous to the concept of appetite: Hunger drives one to eat until satiated, but additional food can still be consumed beyond what the body requires. It is unknown whether the so-called optional sleep serves any function.

According to the core sleep theory, only the core portion of sleep—which is dominated by slow-wave sleep (SWS) and slow-wave activity (SWA) on an electroencephalogram (EEG)—is required to maintain adequate levels of daytime alertness and cognitive functioning.¹⁶ The optional sleep does not contribute to this recovery or maintenance of neurobehavioral capability. This theory was strengthened by results from a mathematical model of sleep and waking functions (the three-process model) that predicted that waking neurobehavioral functions were primarily restored during SWS,¹⁸ which makes up only a portion of total sleep time. However, if only the core portion of sleep is required, it would be reasonable to predict that there would be no waking neurobehavioral consequences of chronically restricting sleep to 6 hours per night, and that cognitive deficits would be evident only when sleep durations were reduced below this amount. Experimental data have not supported this prediction.¹⁰ For example, findings from the largest sleep dose–response study to date, which examined the effects of sleep chronically restricted to 4, 6, or 8 hours of time in bed per night,⁴ found that cognitive performance measures were stable across 14 days of sleep restriction to 8 hours time in bed, but when sleep was reduced to either 6 or 4 hours per night, significant cumulative (dose-dependent) decreases in cognitive performance functions and increases in sleepiness were observed.⁴

It appears, therefore, that the “core” sleep needed to maintain stable waking neurobehavioral functions in healthy adults aged 22 to 45 years is in the range of 7 to 8 hours on average.¹⁸ Moreover, because extended sleep is thought to dissipate sleep debt caused by chronic sleep restriction,³ it is not clear that there is such a thing as “optional” sleep. There is, instead, recovery sleep, which may or may not be optional, although there have very few studies of the sleep needed to recover from varying degrees of chronic sleep restriction.

Adaptation to Sleep Restriction

One popular belief is that subjects may be acutely affected after initial restriction of sleep length and may then be able to adapt to the reduced sleep amount, with waking neurocognitive functions unaffected further or returned to baseline levels. Although several studies have suggested that this is the case when sleep duration is restricted to approximately 4 to 6 hours per night for up to 8 months,⁶^,⁹ there is also evidence indicating that the adaptation is largely confined to subjective reports of sleepiness but not objective cognitive performance parameters.⁴ This suggests that the presumed adaptation effect is actually a misperception on the part of chronically sleep-restricted people regarding how sleep restriction has affected their cognitive capability.

One factor thought to be important in adaptation to chronic sleep restriction is the abruptness of the sleep curtailment. One study examined the relationship between rate of accumulation of sleep loss, to a total of 8 hours, and neurobehavioral performance levels.¹⁹ After 1 night of total sleep deprivation (i.e., a rapid accumulation of 8 hours of sleep loss), neurobehavioral capabilities were significantly reduced. When the accumulation of sleep loss was slower, achieved by chronically restricting sleep to 4 hours per night for 2 nights or 6 hours per night for 4 nights, neurobehavioral performance deficits were evident, but they were of a smaller magnitude than those following the night of total sleep loss. A greater degree of neurobehavioral impairment was evident in those subjects restricted to 4 hours for 2 nights than in those subjects allowed 6 hours per night, leading to the conclusion that during the slowest accumulation of sleep debt (i.e., 6 hours per night for 4 nights), there was evidence of a compensatory adaptive mechanism.¹⁹

It is possible but not scientifically resolved that different objective neurobehavioral measures may show different degrees of sensitivity and adaptation to chronic sleep restriction. For example, in the largest controlled study to date with statistical modeling of adaptation curves, cognitive performance measures showed little adaptation across 14 days of sleep restriction to 4 or 6 hours per night, compared with 8 hours per night,⁴ whereas waking EEG measures of alpha and theta frequencies showed no systematic sleep dose–dependent changes over days.¹⁴ Consequently, different neurobehavioral outcomes showed markedly different responses to chronic sleep restriction, with neurocognitive functions showing the least adaptation, subjective sleepiness measures showing more adaptation, and waking EEG measures as well as non–rapid eye movement (non-REM), SWS measures showing little or no response.⁴^,¹⁴ The reliability of the latter findings may depend on the dose of restricted sleep and other factors.

Two-Process Model Predictions of Sleep Restriction

Biomathematical models of sleep–wake regulation have been used to make predictions about recovery in response to various sleep durations. The basis of almost all current biomathematical models of sleep–wake regulation is the two-process model of sleep regulation.²⁰ This model proposes that two primary components regulate sleep: (1) a homeostatic process that builds up exponentially during wakefulness and declines exponentially during sleep (as measured by slow-wave energy or delta power in the non-REM sleep EEG), and (2) a circadian process, with near–24-hour periodicity.

Since its inception, the two-process model has gained widespread acceptance for its explanation of the timing and structure of sleep. Its use has extended to predictions of waking alertness and neurobehavioral functions in response to different sleep–wake scenarios.²¹ This extension of the two-process model was based on observations that as sleep pressure accumulated with increasing time awake, so did waking neurobehavioral or neurocognitive impairment, and as sleep pressure dissipated with time asleep, performance capability improved during the following period of wakefulness. In addition, forced-desynchrony experiments revealed that the sleep homeostatic and circadian processes interacted to create periods of stable wakefulness and consolidated sleep during normal 24-hour days.²² Hence, it was postulated that waking cognitive function (alertness variable A) could be mathematically modeled as the difference between the quantitative state for the homeostatic process (S) and the quantitative state for the circadian process (C), and thus A = S − C. Accordingly, predictions for changes in the neurobehavioral recovery afforded by chronically restricted sleep of varying durations could be made on the basis of sleep–wake times and circadian phase estimates, using the quantitative version of the two-process model.

The validity of the various biomathematical models based on the two-process model, and their ability to predict actual experimental results of the neurobehavioral effects of chronic sleep restriction have been evaluated in a blind test.²³ Because all current models are based on the same underlying principles as the two-process model, all yielded comparable predictions for neurobehavioral functioning in scenarios involving total sleep deprivation or chronic partial sleep restriction. All models accurately predicted waking neurobehavioral responses to total sleep deprivation. However, they all failed to adequately predict sleepiness and cognitive performance responses during chronic sleep restriction.⁴^,²³ Hence, it appears that the extension of the two-process model to prediction of waking alertness²¹ does not account for the results of chronic sleep restriction. Because the two-process model has had a profound theoretical influence on predictions of sleepiness based on total sleep deprivation data, its failure to capture the dynamic changes in neurobehavioral measures during chronic sleep restriction suggests that additional biological factors are relevant to the brain’s response to chronic sleep restriction.

Effects of Chronic Sleep Restriction

The effects of sleep loss may be quantified in a number of different ways, using a wide range of physiologic, neurocognitive, behavioral, and subjective tools. Many early studies examining the effects of chronic sleep restriction on cognitive performance were conducted outside a controlled laboratory setting, with little or no control over potentially contaminating factors, such as the level of napping, extension of sleep periods, diet, stimulant use (e.g., caffeine, nicotine), activity, or exposure to zeitgebers (environmental time cues). The majority of these studies concluded that there were few or no detrimental effects on waking neurobehavioral capabilities, or subjective effects of the sleep restriction. For example, restriction of nocturnal sleep periods to between approximately 4 and 6 hours per night for up to 8 months produced no significant effects on a range of cognitive outcomes, including vigilance performance,⁹ psychomotor performance,⁶ logical reasoning, addition, or working memory.⁹ In addition, few effects on subjective assessments of sleepiness or mood were reported.⁹

Later studies, however, with far greater experimental control and appropriate control groups, have demonstrated significant cumulative sleep dose–response effects on a wide range of physiologic and neurobehavioral functions, which we summarize here.

Sleep Architecture

Sleep restriction alters sleep architecture, but it does not affect all sleep stages equally. Depending on the timing and duration of sleep, and the number of days it is reduced, some aspects of sleep are conserved, occur sooner, or intensify, and other aspects of sleep time are diminished. For example, studies examining sleep architecture during chronic periods of sleep restriction have demonstrated a consistent conservation of SWS at the expense of other non-REM and REM sleep stages.^4,^24,²⁵ In addition, elevations in SWA, derived from spectral analysis of the sleep EEG in the range of 0.5 to 4.5 Hz, during non-REM sleep have also been reported during and after chronic sleep restriction.⁴^,²⁵

Because of the conservation of the amount of SWS and SWA during restricted sleep protocols, independent of sleep duration (e.g., 8 hours of time in bed or 4 hours of time in bed), it has been proposed that, with regard to behavioral and physiologic outcomes, these phenomena provide the recovery aspects of sleep. It remains to be determined whether the lack of SWS and SWA response to chronic restriction of sleep to 4 hours a night, relative to steady increases in physiologic and neurobehavioral measures of sleepiness,⁴ can account for the latter deficits. Consequently, although SWS and non-REM SWA may be conserved in chronic sleep restriction (to 4 to 7 hours per night), they do not appear to reflect the severity of daytime cognitive deficits or to protect against these deficits, raising serious doubts about SWS and non-REM SWA being the only aspects of sleep critical to waking functions in chronic sleep restriction.⁴

Sleep Propensity

With the development and validation of sleep latency measures as sensitive indices of sleep propensity,²⁶ the effects of chronic sleep restriction could be evaluated physiologically. Objective EEG measures of sleep propensity, such as the multiple sleep latency test²⁶ (MSLT) and the maintenance of wakefulness test²⁷ (MWT), are frequently used to evaluate sleepiness (see Chapters 4 and 143).

The daytime MSLT ²⁶ has been shown to vary linearly after 1 night of sleep restricted to between 1 and 5 hours of time in bed.²⁶ Progressive decreases in daytime sleep latency have been documented (i.e., increases in sleep propensity) across 7 days of sleep restricted to 5 hours per night in healthy young adults,²⁸ a finding confirmed in a later study using the psychomotor vigilance test (PVT).⁸

Dose–response effects of chronic sleep restriction on daytime MSLT values have been reported in a controlled laboratory study in commercial truck drivers.²⁴ A significant increase in sleep propensity across 7 days of sleep restricted to either 3 or 5 hours per night was observed, with no increase in sleep propensity found when sleep was restricted to 7 or 9 hours per night.²⁴ Similarly, sleep propensity (as measured by the MWT²⁹) during 7 days of sleep restriction to 4 hours per night was reported to increase, especially in subjects whose sleep was restricted by advancing sleep offset.³⁰

An epidemiologic study of predictors of objective sleep tendency in the general population ³¹ also found a dose–response relationship between self-reported nighttime sleep duration and objective sleep tendency as measured by the MSLT. Persons reporting more than 7.5 hours of sleep had significantly less probability of falling asleep on the MSLT than those reporting between 6.75 and 7.5 hours per night (27% risk of falling asleep), and than those reporting sleep durations less than 6.75 hours per night (73% risk of falling asleep).³¹ Although the MWT has been used less in experimental settings than the MSLT, it has also been found to increase in experiments in which adults were restricted to 4 hours for sleep for 7 nights,³⁰ and for 5 nights.³² All of these studies suggest that chronic curtailment of nocturnal sleep increases daytime sleep propensity.

Oculomotor responses have also been reported to be sensitive to sleep restriction.³³ Eyelid closure and slow rolling eye movements are part of the initial transition from wakefulness to drowsiness. Eye movements and eye closures have been studied during sleep-loss protocols, under the premise that changes in the number and rate of movements and eyelid closures are a reflection of increased sleep propensity and precursors of the eventual onset of sleep.³⁴ It has been demonstrated experimentally that slow eyelid closures during performance are associated with vigilance lapses and are sensitive indices of sleep deprivation, and slow eyelid closures have been found to be a sign of drowsiness while driving.³³

Increased slow eye movements attributed to attentional failures have been reported to be increased by reduced sleep time in medical residents.³⁵ Sleep restriction has also been found to decrease saccadic velocity and to increase the latency to pupil constriction in subjects allowed only 3 or 5 hours of time in bed for sleep over 7 nights.³⁶ These changes in ocular activity were positively correlated with sleep latency, subjective sleepiness measures, and accidents on a simulated driving task.³⁶

Only gold members can continue reading. Log In or Register to continue