Introduction
Disturbed sleep is increasingly being recognized as a major contributor to various deleterious health outcomes. Sleep deprivation is well known to impair cognition, mood, and memory, but more recent evidence indicates that sleep deprivation also has a major impact on cardiometabolic function. Obstructive sleep apnea (OSA), a common condition with well-established neurocognitive and cardiovascular sequelae, is covered elsewhere in this book ( Chapter 88 ). Other causes of sleep fragmentation such as central sleep apnea ( Chapter 89 ), periodic limb movements, and hypoventilation syndromes ( Chapter 85 ) are also being studied. The term sleep disruption is often used in a general sense to refer to conditions of sleep fragmentation (such as OSA) and sleep deprivation (in which total sleep time is reduced).
Short-Term Consequences with Potential Long-Term Sequelae
Neurocognitive Changes
Sleep disruption is associated with major neurocognitive changes that can affect performance in the short term and may impact long-term cognitive function. The cognitive dysfunction associated with sleep restriction for 28 hours has been shown to be roughly equivalent to that associated with consuming alcohol up to a blood alcohol level of 0.10%, which is above the legal driving limit in most states. Studies of partial and total sleep deprivation indicate a decline in many separate measures of broad cognitive performance, including slowed response time, decrease in short-term and working memory, reduced learning of cognitive tasks, and decreased situational awareness. Sleep deprivation decreases related executive function (via the pre-frontal cortex), which is responsible for working memory. For the individual who has been sleep deprived, this influence may result in decreased insight into the scope of a problem, decreased flexibility in thinking, a propensity for perseveration, and difficulty assimilating and integrating new information, as well as understanding the temporal order of information. Similarly, studies of cognitive function in patients with sleep disordered breathing (SDB) indicate that, compared with the general population, sleep apnea patients exhibit moderate to severe defects in sustained attention tasks (such as the psychomotor vigilance task), driving simulation, and working memory tasks requiring flexibility and insight in thought. SDB patients also demonstrate moderate deficits in verbal fluency, short attention tasks, decreased vigilance, and decreased intellectual function.
More recently, functional magnetic resonance imaging and positron emission tomography studies have been employed to evaluate the cognitive effects of sleep deprivation and fragmentation by evaluating metabolic fluctuations in glucose uptake in specific brain regions. Some data indicate that sleep deprivation results first in global decreases in cortical and subcortical structures. As people become cognitively impaired, metabolism decreases more specifically in the prefrontal cortex, thalamus, and posterior association cortices, which are believed to be responsible for supporting attention. At 24 hours of sleep deprivation, the thalamus is activated when given an attention-demanding task, which has been characterized as the need for increased “mental energy” to maintain attention during sleep deprivation. When given a verbal working memory task after 35 hours of sleep deprivation, increased parietal lobe activity is noted, which may be a compensatory mechanism to help improve overall declining working memory. Further studies are necessary to elucidate better the mechanisms of cognitive dysfunction associated with sleep deprivation.
Sleep and Immune Function
Disrupted and restricted sleep has also been implicated in proper functioning of the immune system. A prospective study by Patel and colleagues assessing nearly 57,000 female nurses (ages 37 to 57 years) found that both short and long sleep duration predicted increased risk of community-acquired pneumonia. Compared with those who slept for 8 hours, women who slept less than 5 hours had a relative risk of 1.4 (CI 1.1 to 1.8) of developing pneumonia. Women who slept more than 9 hours had a similar increased risk for developing pneumonia. A correlation was also found between perceived sleep quality and increased pneumonia risk. Another study evaluated 153 healthy volunteers (both men and women) and found that those who reported sleep of less than 7 hours were 2.9 times (95% CI 1.2 to 7.3) more likely to develop symptoms of a common cold following study-administered intranasal exposure to rhinovirus. Studies to elucidate the mechanism of impaired immunity have found that sleep-deprived humans have lower natural killer (NK) cell activity and IL-2 production, along with increased production of inflammatory biomarkers. In animals, chronic sleep deprivation has been found to reduce monocyte numbers, complement C3 levels, and spleen weight. Furthermore, studies of sleep-deprived animals have shown increased rates of bacteremia.
Along these lines, studies in healthy mice have shown that sleep deprivation ablates the response to influenza vaccination to such an extent that the sleep-deprived mice appeared to have no development of immunity after receiving the vaccine. A study of hepatitis B vaccination titers in humans showed that short sleep duration (as evaluated by actigraphy, i.e., activity monitors, and sleep journals) was associated with a lower secondary antibody response to the vaccine. Similarly, another study showed that one period of 24-hour sleep deprivation significantly reduced H1N1 antibody titers at 5 days postimmunization in healthy volunteers. This effect, however, was not prolonged because total antibody titers were not significantly different at 10, 17, and 52 days postimmunization. This study had a small sample size ( n = 24) and larger studies are necessary. The impact of sleep deprivation on immunity and vaccine titers is particularly relevant in critically ill patients in which iatrogenic infections may have devastating consequences. Further studies will need to quantify and mitigate the risk associated with sleep deprivation and infections in intensive care unit (ICU) patients.
Sleep and the Intensive Care Unit
Besides effects on immunity, sleep deprivation has other important implications for ICU patients. The ICU setting leads to sleep disruption due to excessive environmental stimuli such as alarms, noise from health care providers and other patients, as well as routine interruptions for patient care. Sleep disruption has been hypothesized to contribute to delirium, impaired immune function, and prolongation of mechanical ventilation in critically ill patients. Delirium has been associated with prolonged ICU stays, as well as increased ICU mortality.
The relationship between disrupted sleep and delirium is controversial in part due to the difficulty in distinguishing cause and effect. Key features of delirium are also seen in sleep-disrupted patients, including inattention, fluctuating mental status, and cognitive dysfunction. A study by Trompeo and colleagues of mechanically ventilated patients examined the relationship of ICU delirium with sleep patterns. The authors prospectively followed intubated patients until their sedation had been discontinued for more than 24 hours and the patients were alert, cooperative, and ready to wean from the ventilator. At this point, they measured nocturnal polysomnograms (PSGs) on each patient for one night. Among patients with severe rapid eye movement (REM) sleep reduction (<6% REM), 73% had delirium. Among the patients with more than 6% REM sleep, only 9% had delirium. Although interesting, this study does not address whether the primary event was sleep deprivation or delirium. Further studies will be necessary to determine the mechanisms by which sleep disruption impacts delirium, as well as the potential impact of various interventions on clinical outcomes.
A study by Rompaey and colleagues randomized adult ICU patients to earplugs at bedtime (during the night shift) versus no earplugs and assessed self-reported sleep perception and delirium on the basis of the NEECHAM scale (a standardized test for confusion based on neurocognitive processing, behavior, and physiologic control). They found that the patients using earplugs had better perceived sleep, a lower incidence of confusion, and delayed onset of confusion. No difference was noted in the rate of delirium, a finding that is not surprising because it is unlikely that one intervention would eliminate delirium given the multifactorial nature of this syndrome. This study was encouraging in that it showed that a simple, inexpensive intervention may be beneficial as an adjunct to caring for the critically ill.
Ventilator dyssynchrony is also a likely culprit causing sleep disruption in the ICU. In one study in critically ill mechanically ventilated patients, better synchrony with the use of proportional assist ventilation versus standard pressure support ventilation improved sleep quality. Another study demonstrated that air leak during noninvasive ventilation was associated with disrupted sleep. This finding may be related to a disruption in respiratory pattern versus orofacial mechanoreceptors.
Currently, it is unclear what the implications of sleep loss are on pulmonary mechanics and ventilator weaning in ICU patients. A study in healthy men who were sleep deprived for 30 hours revealed a decrease in inspiratory muscle endurance and maximum voluntary ventilation. Sleep deprivation was initially thought to reduce chemoreceptor-mediated hypercapnic ventilator drive, but a more recent study by Spengler and colleagues showed that 24 hours of sleep deprivation did not affect the sensitivity of central chemoreceptors during resting ventilation. In critically ill tracheostomized patients undergoing a prolonged weaning from ventilation, sleep quality was similar whether on a ventilator or not overnight, but sleep quantity was greater on the ventilator; the authors recommended that patients in prolonged weaning will have greater sleep efficiency if ventilated at night. Further research is necessary to understand the implications of sleep deprivation on weaning from mechanical ventilation.
Poor sleep has been reported in up to 61% of critical care survivors. The impact of this finding is unclear, but poor sleep may contribute to depression, post-traumatic stress disorder, and possibly impaired exercise tolerance among ICU survivors.
Effect on Chronic Disease States
Metabolic
Basic science, translational, and epidemiologic studies indicate that diminished and disrupted sleep predisposes an individual to both obesity and diabetes via altered glucose metabolism, insulin resistance, and dysregulation of appetite control via the neuroendocrine system. Many studies have found a higher prevalence of diabetes within the OSA population with odds ratios from 1.4 to 2.2. Furthermore, restricted and disrupted sleep has been shown to predict glucose control in type 2 diabetes. A study by Aronsohn and colleagues performed an in-laboratory PSG and measured Hgb A1c in 60 diabetic patients. They found that compared with patients without OSA, the adjusted mean HbA1c was increased by 1.5% in patients with mild OSA, 1.9% in patients with moderate OSA, and 3.7% in patients with severe OSA.
An initial sleep debt study looked at healthy male volunteers and subjected them to 4 hours of sleep nightly for 6 days followed by 7 nights of 12 hours in bed. The subjects underwent IV glucose tolerance tests on days 5 and 6 and after 7 nights of rest. They found that the acute response to insulin was diminished by 30% in the sleep-restricted compared with the well-rested state. Furthermore, their disposition index (a product of the acute response to insulin and insulin sensitivity) was 40% lower during sleep restriction. A low disposition index indicates a higher risk of type II diabetes, and these patients in fact had disposition indices in the range similar to those reported in epidemiologic studies of patients at higher risk for type II diabetes (i.e., Hispanic women with prior gestational diabetes). A proposed mechanism for this finding may be via sympathoexcitation and release of counter-regulatory hormones associated with sleep disruption.
Cardiovascular Disease
Multiple studies have documented a relationship between clinical cardiovascular disease and sleep disruption. A large, prospective 10-year cohort study followed more than 70,000 U.S. female health care workers with no known heart disease at baseline to evaluate the incidence of coronary heart disease and its relationship to self-reported daily sleep duration. Fascinatingly, this study showed that short sleep and long sleep were independently associated with a modest increase in incidence of coronary heart disease. This finding mimics the bimodal distribution seen in studies of sleep deprivation and immune dysfunction. The age-adjusted relative risks for individuals reporting fewer than 5 hours per night, 6 hours per night, 7 hours per night, and 9 hours per night were 1.8 (1.3 to 2.4), 1.3 (1.1 to 1.6), 1.1 (0.9 to 1.3), and 1.6 (1.2 to 2.1), respectively.
A similar study of more than 98,000 Japanese men (42%) and women (58%) aged 40 to 79 years investigated cardiovascular mortality in relation to self-reported sleep duration. The study group had a median follow-up of 14.3 years from 1988 to 1990 through 2003. Compared with a sleep duration of 7 hours, a sleep duration of 4 hours was associated with increased mortality from cardiovascular disease in women (hazard ratio of 2.3), as well as an increase in mortality from all causes among both men and women (hazard ratios of 1.3 for men and 1.3 for women). Interestingly, an association was not seen for reduced sleep and cardiovascular mortality in men. Another study examined the relationship between short sleep duration and incident coronary calcification in healthy middle-aged adults. Computed tomography performed in 2000-2001 and 2005-2006 in the cohort revealed that longer sleep duration was associated with a reduced incidence of coronary artery calcification with an adjusted odds ratio of 0.7 per extra hour of sleep. This conclusion was made after accounting for potential confounders (age, sex, race, education, apnea risk, smoking status) and mediators (lipids, blood pressure, body mass index, diabetes, inflammatory markers, alcohol consumption, depression, hostility, self-reported medical conditions). Thus considerable epidemiologic data support a strong association between reduced sleep duration and coronary disease.
Disrupted sleep has a role in the pathogenesis of hypertension, perhaps secondary to disrupted balance of sympathovagal tone. Prospective studies have demonstrated that nocturnal blood pressure is a better predictor of cardiac risk than daytime blood pressure. One prospective study evaluated cardiovascular outcomes for more than 5000 patients with hypertension over a median period of 8.4 years and found that nocturnal blood pressure was the strongest predictor of cardiovascular mortality and that an increase of 10 mm Hg in mean nocturnal blood pressure corresponded with a 21% increase in cardiovascular mortality. Another prospective study evaluated the effect of 24 hours’ sleep deprivation on systolic and diastolic blood pressure in 8 healthy normotensive young adults (mean age 24) versus 8 healthy normotensive older adults (mean age 64) and found a significant increase in both diastolic and systolic blood pressure in the elderly group. OSA has a significant correlation with both systemic and pulmonary arterial hypertension. There is strong evidence from animal studies that OSA contributes to systemic hypertension through mechanisms of intermittent hypoxia, sympathetic activation, and alterations in the renin-angiotensin system.
Disrupted sleep has also been shown to predict arrhythmias. OSA has been documented to have an association with atrial fibrillation, nonsustained ventricular tachycardia, and complex ventricular ectopy. Even in healthy young adults without OSA, a single night of sleep deprivation has been associated with increased atrial electromechanical delay, a marker of risk for various arrhythmias such as atrial fibrillation.
Mechanisms
The pathophysiologic consequences of sleep disruption are related to multiple alterations that take place following attenuated or disrupted sleep, including low-grade systemic inflammation, increased oxidative stress, dysautonomia, and endothelial dysfunction ( Fig. 87-1 ).
