Indira Gurubhagavatula, Barry Fields and Ilene M. Rosen

Characteristics of Normal Sleep

Healthy human sleep is divided into four stages. Three are non-rapid-eye-movement (non-REM or NREM) stages (N1, N2, N3), whereas the fourth is known as rapid-eye-movement (REM) sleep. NREM sleep is characterized by progressive slowing of the frequency of brain wave activity detected by electroencephalography (EEG), with high-amplitude, slow waves (0.5 to 2 Hz frequency), also known as delta waves, predominating in N3 sleep. The major skeletal muscle groups remain active in NREM sleep. N1 is a transitional stage between wakefulness and deeper stages of sleep (N2 and N3). We spend 5% to 8% of sleep in N1, 45% to 65% in N2, and about 15% to 25% in N3.

Healthy sleepers spend about 20% of the night in REM sleep, which is characterized by a mixed-frequency, low-amplitude EEG that bears a striking resemblance to wakefulness. The major skeletal muscle groups are paralyzed, except for the eye muscles, diaphragm, tensor tympani, and posterior cricoarytenoid muscles. REM sleep periods occur in an ultradian rhythm, with the first REM onset occurring approximately 90 minutes after sleep onset. A typical first REM period may last for 20 to 30 minutes, with successive REM periods becoming longer and occurring at approximately 90-minute intervals. REM sleep therefore predominates the last third of the night, whereas NREM predominates the first third.

By contrast, the typical sleep pattern of a resident on call is characterized by frequent sleep interruptions because of calls regarding patient care. Overall sleep time is reduced, with disproportionately greater time spent in stage N1 and less in stages N3 and REM sleep. A delayed sleep onset time and early rise time also disproportionately reduce REM sleep. This fragmented, REM-deprived sleep pattern can lead to significant deficits in memory acquisition and performance of newly learned tasks.

Measuring Sleepiness

Whereas fatigue or weariness results from prolonged physical or mental exertion, illness, or medications, sleepiness describes the tendency to fall asleep. Fatigue may improve with short breaks from work or rest without sleep, but sedentary conditions often unmask sleepiness. That is, a short break in a quiet room may alleviate the sense of fatigue in a medical resident, but make his or her sleepiness even more obvious.

Sleepiness can be measured using subjective tools, like the Epworth Sleepiness Scale. This 8-item, self-administered questionnaire quantifies sleepiness on a scale of 0 to 24, with a score > 10 representing sleepiness and a score > 15 suggesting pathologic sleepiness. Otherwise-healthy medical housestaff demonstrate Epworth scores well in the pathologic range.

Determinants of Alertness

An individual’s level of alertness waxes and wanes throughout a 24-hour period. The greater the number of hours an individual spends awake, the greater the tendency toward sleepiness (Figure 106.1). This “sleep pressure,” which dissipates after sleep, is known as the homeostatic “Process S.” Process S is simultaneously counterbalanced by the circadian process, or Process C, which is driven by an internal clock residing in the suprachiasmatic nucleus. This wake-promoting internal clock operates with a periodicity of about 24.2 hours. A small circadian dip in alertness occurs in the late afternoon, and a larger nadir occurs at night, when the clock withdraws its wake-promoting signal. Because both Processes C and S favor sleepiness in the later evening and overnight, humans typically spend these hours sleeping.

Numerous external factors impact an individual’s performance, including environmental factors, such as light, noise, and temperature; situational factors, such as immediacy and urgency, stress, or boredom; pharmacologic factors, such as the use of caffeine or other stimulants or alcohol; and individual factors, such as genetically determined tolerance to lower amounts of sleep. Despite these factors, if sleep need is not met, the homeostatic drive for sleep builds to the point at which the circadian system can no longer maintain wakefulness. As a result, the individual may experience micro-sleeps or sleep attacks despite attempts to stay awake. Indeed, firefighters have fallen asleep near burning wildfire embers because of the powerful homeostatic drive for sleep after several days of sleep deprivation.

Sleep-Related Determinants of Performance

Four characteristics of sleep determine performance: (1) nightly sleep duration; (2) number of hours awake; (3) perturbation of the circadian phase, as occurs when an individual works nightshifts, and (4) sleep inertia. The third correlates with a misalignment of Processes C and S, as mentioned previously. Sleep inertia refers to the decrement in performance in the first 15 to 30 minutes after waking, with impairments observed in short-term memory, counting, cognitive processing speed, and number fact and lexical retrieval. These skills are critical for ICU residents, who may be awakened for emergencies and asked to synthesize, sort, and integrate voluminous data and deliver urgent decisions. Sleep inertia increases with the depth of sleep that precedes waking; hence, longer naps lead to sleep inertia more often than shorter ones.

Effects of Sleep Deprivation

Sleep deprivation results from insufficient sleep time, with neurobehavioral, cognitive, physiologic, and epidemiologic consequences. On a neurobehavioral level, microsleeps intrude into wakefulness involuntarily. Increased lapsing (errors of omission) and false responses (errors of commission) have also been described. Additionally, time-on-task decrements are observed, whereby the ability to sustain attention and respond quickly and accurately becomes unstable the longer an individual remains awake and attempts a repetitive task. Cognitive impairments are seen in learning and recall, working memory, and executive function. There is loss of the speed/accuracy trade-off: individuals take longer and still commit more errors with sleep deprivation.

Some individuals are far more vulnerable to cognitive deficits than others. Unfortunately, the ability of individuals to self-rate their degree of impairment tends toward inaccuracy; anesthesia residents misidentify EEG-confirmed sleep in 50% of trials. These effects can lead to workplace injuries. Indeed, concentration lapses and fatigue were the most commonly reported contributing factors among interns with percutaneous injuries. Such injuries were more frequent after extended shifts (mean of 29.1 consecutive work hours) versus non-extended shifts (mean of 6.1 consecutive work hours). Attention failures, medical errors, and personal injuries have been shown in sleep-deprived housestaff specifically during their ICU rotations.

Numerous physiologic perturbations of sleep deprivation include increased cortisol, cytokines, and C-reactive protein; leukocytosis; insulin resistance; on EEG, reduced sleep latency, increased slow waves, increased slow eye movements during sleep, and greater frequency of slow eyelid closures. Epidemiologic surveys reveal an increased likelihood of mortality, diabetes, insulin resistance, and cardiovascular disease in sleep-deprived individuals.

Performance decrements result not only from acute, total sleep deprivation, but from repeated nights of partial sleep loss as well, in a dose-responsive fashion. Therefore, a week of partial sleep deprivation, even without “extended” shifts, can result in more lapses than 2 consecutive days of total sleep deprivation.