Sleep and Health

Sleep and Health provides an accessible yet comprehensive overview of the relationship between sleep and health at the individual, community and population levels, as well as a discussion of the implications for public health, public policy and interventions. Based on a firm foundation in many areas of sleep health research, this text further provides introductions to each sub-area of the field and a summary of the current research for each area. This book serves as a resource for those interested in learning about the growing field of sleep health research, including sections on social determinants, cardiovascular disease, cognitive functioning, health behavior theory, smoking, and more.

• Highlights the important role of sleep across a wide range of topic areas
• Addresses important topics such as sleep disparities, sleep and cardiometabolic disease risk, real-world effects of sleep deprivation, and public policy implications of poor sleep
• Contains accessible reviews that point to relevant literature in often-overlooked areas, serving as a helpful guide to all relevant information on this broad topic area

• Language: English
• Publisher: Academic Press
• Release date: Apr 17, 2019
• ISBN: 9780128153741

Book preview

things.

Part I

General concepts in sleep health

Chapter 1

The basics of sleep physiology and behavior

Andrew S. Tubbsa; Hannah K. Dollishb; Fabian Fernandezb; Michael A. Grandnerc    a Department of Psychiatry, University of Arizona, Tucson, AZ, United States

b Department of Psychology, University of Arizona, Tucson, AZ, United States

c Sleep and Health Research Program, Department of Psychiatry, University of Arizona College of Medicine, Tucson, AZ, United States

Abstract

Sleep is an essential element of human health, supporting a wide range of systems including immune function, metabolism, cognition, and emotional regulation. To understand everything that sleep does, however, it is necessary to understand what sleep is. This chapter provides that foundation by discussing the conceptualization, physiology, and measurement of sleep.

Keywords

Sleep need; Sleep ability; Sleep opportunity; Sleep physiology; Polysomnography; Actigraphy; Non-REM sleep; REM sleep; Circadian rhythms; Sleep measurement

Introduction

Sleep is an essential element of human health, supporting a wide range of systems including immune function, metabolism, cognition, and emotional regulation. To understand everything that sleep does, however, it is necessary to understand what sleep is. This chapter provides that foundation by discussing the conceptualization, physiology, and measurement of sleep.

The definition of sleep

Sleep is a naturally recurring and reversible biobehavioral state characterized by relative immobility, perceptual disengagement, and subdued consciousness. As a predictable and easily reversible phenomenon, sleep is distinct from states of anesthesia and coma, which typically involve the absence or suppression of neural activity. Additionally, proper sleep involves a dynamic interaction between voluntary decisions and involuntary biological activities. Turning off the lights, reducing noise, and laying down are voluntary behaviors, but the result is an involuntary increase in melatonin and a series of shifts in the activity patterns of the brain throughout the night. Sleep ultimately depends on this collaboration between behavior and biology, and a deficit in either will disrupt sleep.

Conceptualizing sleep as a health behavior

A health behavior is an action (or omission) by an individual that impacts their health. Conceptualizing sleep as a health behavior is useful because it highlights how behavior and neurobiology interact, and how individuals can modify their health through sleep. Viewed in this way, sleep can be divided into three processes: sleep need, sleep ability, and sleep opportunity. These processes are diagramed in Fig. 1.1.

Fig. 1.1 The three process model of sleep.

Sleep need is the biological requirement for sleep, or the minimal amount of rest the body requires to prepare for the next day. This need is defined by individual genetics and physiology and does not change after losing a night of sleep or oversleeping on the weekends. Unfortunately, there is no standard method for measuring sleep need. While epidemiological studies suggest an average of 7–8 h for healthy adults, some individuals naturally need more sleep (e.g., children and adolescents), while some need less. Sleep need represents the core motivation for engaging in sleep, and consistently failing to meet this need can promote cardiometabolic disease, impair cognitive functioning, and increase risk for psychiatric disorders.

The only way to satisfy sleep need is to sleep. The amount of sleep an individual can achieve is known as sleep ability and is approximated by total calculated sleep time. Unlike sleep need, sleep ability can change from one night to the next depending on life circumstances. Stress, a cold, or the death of a loved one can reduce sleep ability, while one night of sleep deprivation can increase sleep ability the following evening. Thus, while sleep ability cannot be directly controlled, it can be influenced by behavior.

Whereas sleep ability is correlated to the amount of time one spends sleeping, sleep opportunity is the amount of time the person makes available for sleep. Sleep opportunity is measured by the amount of time the person stays in bed (although time spent in habitual bedtime activities like reading a book could theoretically be incorporated). Unlike the two previous processes, sleep opportunity is under conscious control and is the most vulnerable to environmental factors. This is illustrated by a trauma resident on a 24-h call. Fatigue accumulates over the course of the shift, slowly increasing the resident’s sleep ability. However, sleep opportunity is negligible, since the resident must be ready at a moment’s notice to respond to a life-threatening crisis.

While these three sleep processes can be described independently, in the real-world they work together to control sleep. Sleep need motivates the creation of sleep opportunity, which provides a context for sleep ability to produce sleep—thus satisfying sleep need and reinforcing the methods used to create sleep opportunity in the first place.

Conceptualizing sleep as a physiological process

Sleep involves a progression of neurophysiological changes in the brain. These changes are grouped (somewhat artificially) into stages based on scoring convention. To explore these changes, this section will briefly describe wakefulness and then proceed to give a description of each of the sleep stages.

Wakefulness

During wake, the brain is engaged in numerous activities, many of which are unrelated to one another. For example, someone might be watching a TV show while listening for the sound of a car in the rain and thinking about whether there is enough food in the refrigerator for lunch tomorrow. The aggregated electrical activity produced by these processes can be observed using electroencephalography (EEG), which would show a high-frequency, low-amplitude signal traveling across the surface of the brain (Fig. 1.2). To understand what this means, imagine a crowd cheering in a sports stadium. Everyone is shouting at different times. This means that while someone is always cheering (high frequency), it is impossible to discern individual words because everyone’s words are drowning each other out (low amplitude). Coming back to the brain, a high frequency signal means many different processes are present in the circuitry of the brain, but the timing of these processes is scattered. Because this activity is widespread, it is hard to resolve any one particular process, resulting in a low amplitude signal. The beta frequency is the classic frequency of active wake and ranges from 12 to 30 Hz (cycles per second). When subjects lay down and close their eyes, electrical activity generally slows to an alpha frequency (8.5–12 Hz), which indicates that the person is awake but not necessarily attending to their surroundings (Fig. 1.2). Along with electrical activity, wakefulness is characterized by high levels of arousal neurotransmitters, such as dopamine, noradrenaline, and serotonin. There is also increased autonomic activity; heart rate, respiratory rate, and blood pressure are constantly responding and adapting to changes in the body and the environment throughout the day.

Fig. 1.2 Example traces of the stages of sleep. Blue box highlights a sleep spindle. Green box highlights a k-complex. Notice how in REM sleep the electrical activity is similar to Wake or Stage 1, but there is a complete lack of motor activity in the electromyogram (EMG). Images taken from an anonymous human recording.

NREM sleep: General overview

The first half of the sleep cycle can be divided into three distinct stages of nonrapid eye movement (NREM) sleep, aptly named Stage 1, Stage 2, and Stage 3. Electrical activity throughout the brain decreases in frequency and increases in amplitude at each progressive stage. This reflects a reduction in overall neural activity, but an increasing coordination among neurons (i.e., enhanced oscillation). Recalling the example from above, imagine if the crowd slowly coordinated their cheering. At first, the sound would still be noisy and unintelligible. However, once everyone followed along, a wave of quiet alternating with cheering would emerge. The number of cheers would decrease (lower frequency), but the volume and clarity of each cheer would steadily increase (higher amplitude). By the end of NREM the electrical activity of the brain is tightly synchronized, leading to lower frequency, higher amplitude oscillating waves known as slow waves, which cycle about once every second. The final stage of NREM is commonly referred to as slow wave sleep because of the dominance of these waves in the EEG record.

In addition to shifts in brain activity, NREM sleep is accompanied by a global decrease in wake-related neurotransmitters and impaired perception of external stimuli. In fact, most sensory inputs are specifically filtered out by the thalamus to protect sleep. Indices of autonomic activity such as heart rate, respiratory rate, temperature, and blood flow to the brain are reduced as one advances from one NREM stage to the next. Motor activity is markedly reduced, but not completely absent.

NREM: Stage dissection

The Rechstschaffen and Kales scoring criteria divide NREM into four stages [1], while the more current American Academy of Sleep Medicine (AASM) criteria combine the last two NREM stages [2] since the distinction is not viewed as clinically relevant [3]. Here, in our stage dissection of NREM, we observe the AASM criteria.

Stage 1 is characterized by approximately 50% alpha activity (waves of brain activity cycling at 8–12 Hz) and the emergence of theta waves (4–7 Hz) in the EEG trace. Nonelectrophysiological markers can also include slow-rolling eye movements, unusual visual sensations that take the form of clouds or flares of light (phosphenes), and hypnagogic myoclonia, which are brief jerking movements. The arousal thresholds for waking during Stage 1 are selective, as the brain determines if there is something worth attending to or whether it can commit to extended sleep. Here, for example, one’s name spoken softly can awaken a person whereas a similar sounding word spoken at the same intensity might not. Stage 1 accounts for 5% of total sleep time and, in healthy sleep, is the entry point for NREM Stage 2.

Stage 2 is characterized by the absence of slow-rolling eye movements, mixed frequency neurophysiological activity, and the presence two major transient electrical phenomena: k-complexes and sleep spindles. K-complexes are large-amplitude rapidly fluctuating bursts of brain activity, while spindles are 12–15 Hz oscillating signals lasting 0.5–2 s (Fig. 1.2). While these phenomena are theorized to support memory consolidation and/or filter sensory input, their true functions remain unknown. Stage 2 sleep comprises 45%–55% of total sleep time and is viewed as a bridge between light (Stage 1) and deep (Stage 3) NREM sleep.

Stage 3 sleep is often referred to as slow wave sleep (SWS) owing to the prominence of high amplitude, low frequency delta oscillations recurring at ~ 1 Hz. The amount of time an individual spends in SWS positively correlates with lack of sleep, such that SWS is elevated during the first sleep cycle after a prolonged period of wakefulness. SWS is thought to discharge sleep pressure that has accumulated throughout the day because the amount of time spent in this stage decreases dramatically as the night progresses. SWS tends to coincide with the timing of peak growth hormone secretion, hinting at a role for this sleep stage in nightly maintenance and repair of the body. Additionally, oscillations that appear during SWS may function as a broad conduit for the repeated activation of memory centers of the brain to support memory-strengthening. The end of Stage 3 NREM sleep is usually followed by entry into REM sleep.

Description of REM sleep

Rapid eye movement (REM) sleep represents a categorical shift in sleep-related brain activity and forms the latter half of the sleep cycle. While most neurotransmitters drop to low levels, acetylcholine levels match or exceed those produced during wake. The surge in acetylcholine creates patterns of electrical activity in the sleeping brain that approximate the high frequency, low amplitude patterns usually seen in alert individuals. Despite this increase in overall excitability, there is a paradoxical loss of muscle movement (i.e., sleep paralysis). The only exceptions are eye muscles, which show the rapid, jerking movements for which the stage is named, and the diaphragm, which remains functional but contracts erratically. The loss of muscle tone leads to further narrowing of the upper airway, which can trigger snoring. Blood pressure and heart rate are also destabilized in REM sleep, in some cases leading to sympathetic storms of phasic arousal. At the same time, temperature regulation is impaired due to the loss of the ability to shiver.

Perhaps the most dramatic change associated with REM is in the content of dreams. In NREM, dreams are often more grounded, logical, and procedural, lacking any real visual or sensory detail. REM dreams, by contrast, are an absolute free-for-all of sensory experience, visual content, and emotions that can rapidly morph in content and affect with little reasoning. REM dreams may support memory consolidation processes, particularly by linking disparate concepts or connecting new ideas to old ones. They may also support emotional processing of difficult events (e.g., divorce, bereavement). This is based on functional imaging studies which show elevated activity in limbic regions during REM sleep and increases in emotional regulation inventories after subjects awake [4, 5]. In recent years, connections between REM sleep and emotion regulation have been most manifest in individuals suffering from posttraumatic stress disorder (PTSD).

Although the brain generates motor commands during dreams, the descending motor neurons are inhibited in the brainstem to prevent execution of these commands. When this process is disturbed (as in several neurological and psychiatric disorders), subjects will act out their dreams, often posing a danger to themselves or their bed partners. REM sleep accounts for approximately 25% of sleep and occurs in 4–6 episodes distributed across the night.

Moving through the sleep stages

Although the stages of sleep are presented in a particular order here, it should be noted that progression is not always linear. While all subjects start in Stage 1, they may proceed rapidly through Stage 2 to 3, or they may backtrack to earlier stages before proceeding to REM. Conversely, while it is typical to return from REM to NREM Stage 1 or 2, it is possible to return to any NREM stage following a REM episode. An example diagram of sleep stages, known as a hypnogram, is presented in Fig. 1.3.

Fig. 1.3 A hypnogram, which tracks the amount of time spent in each sleep stage across the night.

Sleep and circadian rhythms

Across cultures, geography, seasons, and age, humans tend to sleep at night and wake up in the morning. This phenomenon is so ubiquitous that it often escapes scrutiny, but consider if it were a different biological function. What if, for example, humans only used the restroom at certain times of day? It seems banal that sleep should occur at night, but it is actually a remarkable feat of biology that humans (and many other animals) consolidate this large set of biological processes to a particular stretch of the day.

Generally speaking, there are two factors that ensure sleep occurs at night. The first is sleep propensity, or the drive for sleep. Physical and mental fatigue that accumulates during the day increases sleep propensity, and by nighttime the elevated sleep propensity drives humans to engage in sleep. The other factor is the circadian system. The molecular machinery of the circadian system is found within each cell of the body, comprised of an interlocking set of signaling proteins that produce a ~ 24-h rhythm of cellular functions that can be further adjusted by cues in the environment. This machinery ensures that functions such as digestion and immune system maintenance are optimized at specific points in the 24-h solar day. For example, when the clock signals that it is biological night, the body responds by shifting neurobiological activities to favor sleep.

These two factors are formally referred to as the Two Process Model of Sleep (Fig. 1.4). In the morning, sleep propensity is low, but increases over the course of wakefulness. Conversely, the circadian drive for wakefulness increases during the morning, peaks during the midday, and then drops at night. While there is a short, early evening peak that sustains wakefulness after sunset (referred to as the wake-maintenance zone), sleep propensity eventually exceeds circadian wakefulness and sleep onset occurs. This point is referred to as the sleep gate. In humans, the peak of sleep propensity and the trough of wakefulness occur at night, which is why humans tend to sleep at that time.

Fig. 1.4 The two process model of sleep.

So how exactly do these two forces work to generate the sleep gate? Sleep propensity is not well understood, but current theories focus on the buildup of certain substances (such as adenosine) that signal increasing levels of fatigue. This explains why caffeine, which opposes rising adenosine levels, is an effective stimulant.

The circadian system is largely underpinned by a part of the brain called the suprachiasmatic nucleus of the hypothalamus (SCN). Because of its circuit connections with the eye and ability to track sunrise and sunset, the SCN can operate as the master pacemaker that synchronizes all the miniature cellular clocks of the body to the light schedule set by the Earth’s rotation (like a conductor of a symphony orchestra). In the absence of external photic cues, the SCN can still produce an endogenous rhythm that approximates the lengths of day and night. However, this rhythm is imprecise and follows a schedule that—depending on the person—is slightly longer or shorter than 24 h. Without any means of correction, the endogenous rhythm set by the SCN would slowly drift away from the solar day’s 24-h cycle (resulting in non-24-h circadian rhythm disorder; Fig. 1.5). Fortunately, the SCN can use the light information it receives from the eye on a daily basis to adjust for the difference in timing, a process known as entrainment.

Fig. 1.5 Circadian rhythms of wake and sleep. Each bar represents a day. Wake is presented as yellow , while sleep is presented as black . Both typical and abnormal circadian rhythms are presented.

Disruptions to the circadian system can manifest as difficulties with sleep. One example is jet lag. When an individual rapidly changes time-zones, the external light/dark cues of the new destination become misaligned with the endogenous rhythm of night and day, which is still operating on the previous light schedule. Depending on the direction of the shift, an individual may awaken hours before dawn in the new location (phase advance), or take hours to fall asleep after night has fallen (phase delay). Fortunately again, after several sleep/wake intervals, the SCN will re-entrain the body to the new light/dark cues, thus normalizing sleep. Examples of typical and atypical circadian rhythms are presented in Fig. 1.5.

Basic sleep physiology

There are no specific parts of the brain that act as monolithic sleep or wake centers. Rather, the neurobiological states of sleep and arousal are achieved via coordinated interactions between multiple brain regions. This section will highlight brain regions, chemical signals, and physiological processes that coordinate sleep and wake.

The brainstem

The brainstem is the most evolutionarily conserved structure within brain. As such, it is the control center for the autonomic nervous system, which regulates basic life-sustaining activities such as heart rate, blood pressure, and respiration. Regarding sleep and wake, the brainstem produces wake-promoting neuromodulators such as serotonin, norepinephrine, and dopamine that set the general volume of brain activity. The brainstem regions that produce these chemicals are collectively referred to as the ascending activating system because these regions project to and activate higher order brain areas located in the cerebral cortex.

The hypothalamus

The hypothalamus supports three major processes associated with sleep. First, it houses the SCN, and thus maintains circadian timekeeping. Second, the hypothalamus regulates the autonomic nervous system, particularly with regard to temperature. Third, it augments the wake-promoting neuromodulators of the brainstem with two additional chemicals: histamine and orexin (also known as hypocretin). Hypothalamic production of histamine and orexin drives wakefulness during the day, while low concentrations of histamine and orexin at night facilitate drowsiness and a tendency to sleep.

The thalamus

The thalamus is a collection of nuclei that serves as the gateway for information related to touch, taste, sight, and sound to travel to and between areas of the cerebral cortex. Although historically seen as a simple relay station, the thalamus is now understood to perform an extensive filtering function. During sleep, the thalamus blocks most sensory information from reaching the cortex. Ambient noise, whispers, and low light are all eliminated, allowing sleep to occur without having to consciously process what is going on in the environment.

Cerebrum

The cerebrum includes a variety of cortical and subcortical structures, such as the somatosensory and motor cortices, basal ganglia, and hippocampus. Sensory processing, motor commands, language, memory, and emotion all occur in or involve elements of the cerebrum, which exhibits the vast majority of neural activity in the brain. The cerebrum does not drive a specific element of sleep or wake, but the activity of billions of cortical neurons plays a large role in whether a person is awake or asleep.

Neuromodulators

As mentioned above, neuromodulators play a major role in sleep and wakefulness. Listed below are six major neuromodulators known to influence sleep.

●Dopamine: Produced by the substantia nigra and the ventral tegmental area of the brainstem, dopamine promotes wakefulness. Some other functions of dopamine include stimulation of the basal ganglia to promote voluntary movement, and stimulation of the nucleus accumbens as part of the pleasure and reward systems.

●Histamine: Produced by the tuberomammillary nucleus of the hypothalamus, histamine promotes wakefulness. This is why antihistamines such as diphenhydramine (Benadryl) cause drowsiness; they are able to enter the brain and block the wakefulness promoting effect of histamine. Second generation antihistamines, such as cetirizine (Zyrtec), do not cause drowsiness because they do not cross the blood brain barrier.

●Norepinephrine: the precursor to epinephrine (adrenalin), norepinephrine is produced by the locus coeruleus in the brainstem. Norepinephrine acts at the same receptors as epinephrine to stimulate wakefulness, although at a much reduced half-life.

●Acetylcholine: Although acetylcholine is often used as a neurotransmitter, it is also produced by the basal forebrain and multiple regions of the brainstem to act as a neuromodulator. Acetylcholine promotes wakefulness, but also supports REM sleep.

●Serotonin: Produced by the dorsal raphe nucleus of the brainstem, serotonin has more than 14 receptor subtypes, many of which differ in their activity. In general, increased serotonin levels tend to promote wakefulness and inhibit REM.

●Orexin: Produced by the lateral hypothalamus, orexin is a major wake-promoting agent in the brain. Its absence, most likely due to autoimmune destruction, is the chief cause of the sleep disorder narcolepsy. In addition, orexin enhances the activity of brainstem neuromodulators such as noradrenaline.

The autonomic nervous system

The autonomic nervous system is responsible for the subconscious regulation of heart rate, respiration, blood pressure, temperature, and other vital pieces of physiology. This system is divided into two opposing branches: the sympathetic (fight or flight) and parasympathetic (rest and digest) branches. During wakefulness, the sympathetic and parasympathetic branches are constantly adapting to environmental stimuli and emotional/mental processes. During NREM sleep, however, the sympathetic branch is largely quiescent while the parasympathetic branch remains active. This results in a progressive decrease in heart rate, temperature, and blood pressure which promotes and maintains NREM sleep moving through stages 1–3.

Quantifying sleep

Sleep incorporates a range of biological and behavioral activities which cannot be captured by a single measurement. Instead, sleep is quantified within two broad domains: sleep continuity and sleep architecture.

Sleep continuity encapsulates the timeline of how a person sleeps. Total sleep time, sleep onset latency (amount of time it takes to fall asleep), and the number and duration of awakenings in the night are all measures of sleep continuity. Sleep continuity also includes sleep efficiency, which is defined as the total sleep time divided by the time in bed. In other words, sleep efficiency is a ratio of sleep ability (sleep time) to sleep opportunity (time in bed). Sleep continuity variables are typically self-reported in a sleep diary or by using an activity monitor (discussed below). This information can help identify sleep patterns over time or diagnose specific sleep disturbances, such as insomnia or circadian rhythm disorders.

The second domain of sleep quantification is sleep architecture, which measures the electrophysiological changes throughout a sleep episode. Sleep architecture quantifies each stage of sleep, and the progression through each stage. For example, measures of sleep architecture would capture if someone enters REM very quickly after sleep onset, a key symptom in the diagnosis of narcolepsy. The gold- standard for measuring sleep architecture is polysomnography (PSG).

Capturing both the psychological and physiological elements of sleep requires subjective and objective measurements. Subjective assessments, such as questionnaires or sleep diaries, rely on self-report data. While subject perceptions may bias the results, these are the only measures that capture the subjective experience of sleep. Objective measures, such as actigraphy or PSG, replace subject perceptions with independently observable data, such as changes in brain waves or body movement. Although researchers and clinicians tend to prefer objective hard data to self-reports, subjective data should be seen as complementary to objective data, not subordinate. For example, suppose an objective measurement captures 8 h of sleep, but a patient only reports 30 min in between tossing and turning. It is tempting to think one of the measurements is wrong, but the reality is that something unusual is happening that neither measure adequately captures. In fact, this phenomenon is referred to as paradoxical insomnia, and is both poorly understood and difficult to treat.

Subjective measures

The simplest measure of sleep is a single question: How much do you normally sleep? This question varies in form, sometimes asking about weekday versus weekend sleep, sleeping alone or with a partner, and sleep before and after a child. This basic question is widely used in epidemiological studies and large datasets, such as the National Health and Nutrition Examination Survey. However, this question offers the most limited insight into a person’s sleep. First, it is subject to recall bias, in that subjects can recall things differently than what actually happened. The second problem is resolution, since asking about sleep in the last month or year will lead subjects to average across many nights based only on what they can remember. This reduces temporal precision and increases recall bias. However, this question may be the only way to acquire historical information about sleep, such as when a clinician is seeking to understand the course of a sleep disorder. When patients report decreasing sleep durations and increasing sleep onset latency over the course of a month, objective measures may not be necessary to initiate treatment for early insomnia.

The next level of subjective measurements of sleep are validated questionnaires, such as the Insomnia Severity Index. Although questionnaires are both subjective and retrospective, they are usually standardized to capture specific data or screen for specific disorders. A wide variety of questionnaires exist, and a few are listed in Table 1.1.

Table 1.1

The final subjective measurement is the sleep diary which has been used for decades in research and clinical settings. Subjects report on different sleep continuity variables shortly after waking up. Unlike questionnaires, a sleep diary is considered a prospective measurement of sleep. However, if too much time passes between the sleep episode and the recording date, recall bias affects the accuracy of the data. Sleep diaries are easy to use (the subject can complete it on paper or electronically) and can be easily collected for any length of time. For these reasons, sleep diaries are an effective tool for longitudinal assessments of sleep and sleep/wake timing.

Objective measures

Objective measures replace subjective perceptions with independent, observable phenomena. The most ubiquitous form of objective sleep measurement is actigraphy. An activity monitoring device, usually a wrist-worn device, uses an accelerometer to detect and measure bodily motion. An activity threshold is set and any activity level below the threshold is classified as either rest or sleep. In addition to devices used by clinicians and researchers, there are consumer smart devices, such as phones and watches, that utilize actigraphy. However, the algorithms used to assess sleep efficiency and stages are proprietary and can vary between companies. It is important to choose a company and device that has been validated in many populations and against PSG data and other sleep metrics when conducting a study or for sleep assessment. An activity monitoring device can also include a light sensor. A light sensor allows the clinician to also measure changes in natural and artificial light the patient is exposed to. This is useful for capturing an individual’s photoperiod (i.e., the period of light exposure), which can help determine if light exposure is related to the individual’s sleep. For example, blue light has a detrimental effect on sleep, and so limiting blue light late in the evening may be a treatment strategy for some patients.

Like sleep diaries, activity monitors provide day-to-day measures of sleep continuity. Additionally, multiple weeks of actigraphy data can be used to evaluate sleep/wake cycles and related circadian rhythms. Additional photopic data can show whether light exposure is affecting a person’s circadian rhythm. For example, repeated blue light exposure late at night may shift sleep onset to later in the evening, resulting in a delayed circadian phase. The coupling of actigraphy to light exposure allows the comparison of sleep/wake behavior with external phototopic cues, which are helpful in assessing the synchronization of sleep/wake cycles with light/dark cycles.

The other objective measure of sleep is polysomnography (PSG), also known as a sleep study. During a sleep study, subjects spend 1–2 nights in the sleep laboratory wearing sensors that measure brain activity, eye movements, muscle movements, heart activity, respiratory activity, and blood oxygen levels. Additional sensors may be placed on the legs to measure periodic limb movements, which can occur naturally or as part of sleep movement disorders.

PSG measures the electrical and physiological changes that occur during sleep and is currently the only way to determine sleep stages. The primary clinical utility of PSG, however, is for the diagnosis and treatment of sleep apnea. Sleep apnea is a condition where patients cease breathing during sleep, often due to upper airway collapse. The PSG captures these events as a decrease in both nasal airflow and blood oxygenation, and the number of events per hour is used as a measure of the severity of the sleep apnea. In some cases, physicians will order a split-night study, in which sleep apnea is measured in the first part of the night, and then positive airway pressure therapy is initiated to control the apneas in the later half.

Conclusion

Without sleep, a wide variety of systems such as cell division, metabolism, neurological functions, and mental and emotional health would all be greatly impaired. Diseases that affect sleep are life-altering, and if left untreated, can decrease quality of life and increase risk of death. It is also important to synchronize sleep to our external world, a job done exceedingly well by the biological clock. The rhythmicity and predictability of sleep highlights where and how disruptions are occurring in various conditions and disorders. Understanding sleep at the fundamental level is critical in understanding the clinical significance sleep has on overall health.

References

[1] Kales A., Rechtschaffen A. A manual of standardized terminology, techniques and scoring system for sleep stages of human subjects. Bethesda, MD: U.S. National Institute of Neurological Diseases and Blindness, Neurological Information Network; . 1968;vol. 57.

[2] Berry R.B., et al. AASM scoring manual updates for 2017 (version 2.4). J Clin Sleep Med. 2017;13(5):665–666.

[3] Moser D., et al. Sleep classification according to AASM and Rechtschaffen & Kales: effects on sleep scoring parameters. Sleep. 2009;32(2):139–149.

[4] Rothbaum B.O., Mellman T.A. Dreams and exposure therapy in PTSD. J Trauma Stress. 2001;14(3):481–490.

[5] Desseilles M., et al. Cognitive and emotional processes during dreaming: a neuroimaging view. Conscious Cogn. 2011;20(4):998–1008.

[6] Bastien C.H., Vallieres A., Morin C.M. Validation of the insomnia severity index as an outcome measure for insomnia research. Sleep Med. 2001;2(4):297–307.

[7] Johns M.W. Reliability and factor analysis of the Epworth sleepiness scale. Sleep. 1992;15(4):376–381.

[8] Kaida K., et al. Validation of the Karolinska sleepiness scale against performance and EEG variables. Clin Neurophysiol. 2006;117(7):1574–1581.

[9] Buysse D.J., et al. The Pittsburgh sleep quality index: a new instrument for psychiatric practice and research. Psychiatry Res. 1989;28(2):193–213.

[10] Klingman K.J., Jungquist C.R., Perlis M.L. Questionnaires that screen for multiple sleep disorders. Sleep Med Rev. 2017;32:37–44.

[11] Chung F., et al. STOP questionnaire: a tool to screen patients for obstructive sleep apnea. Anesthesiology. 2008;108(5):812–821.

[12] Horne J.A., Ostberg O. A self-assessment questionnaire to determine morningness-eveningness in human circadian rhythms. Int J Chronobiol. 1976;4(2):97–110.

Chapter 2

Epidemiology of insufficient sleep and poor sleep quality

Michael A. Grandner    Sleep and Health Research Program, Department of Psychiatry, University of Arizona College of Medicine, Tucson, AZ, United States

Abstract

Insufficient sleep duration and poor sleep quality have emerged as key behavioral risk factors for cardiometabolic disease risk, daytime functioning deficits, and other adverse outcomes, including mortality. Understanding the degree to which these exist in the population and how this prevalence differs based on demographic and other characteristics can provide context for the scope of the problem as it affects the population. Insufficient sleep is common, affecting about 1/3 of the general population. In addition, sleep complaints and symptoms – including nonrestorative sleep, difficulty initiating or maintaining sleep, daytime sleepiness, and snoring – are also common in the general population. Further, factors such as age, sex, race/ethnicity, and socioeconomic status can impact the prevalence of sleep problems.

Keywords

Sleep duration; Sleep quality; Epidemiology; Insomnia; Population

Sleep at the population level

Sleep is a universal human phenomenon and impacts every person, every day (whether or not they actually get to sleep). For this reason, population-level estimates of sleep are important. However, they may be difficult to obtain. Since an individual is unconscious while they are sleeping (and for the time surrounding sleep onset and awakening), accurate assessment of the population burden of sleep disturbance may be difficult. Methods typically exist on a continuum whereby increased generalizability is compromised by reduced precision. For example, most population-level estimates are based on a retrospective self-report, which lacks precision. More precise measures, such as polysomnography and even actigraphy, have been thus far impractical for truly large and population-level assessments. Still, several tentative conclusions about the population can be drawn regarding sleep health.

Defining insufficient sleep

There has been a general lack of consensus on the definition of what constitutes insufficient sleep in the general population since at least 1964 [1], when Hammond published the finding that habitual short and long sleep duration were associated with increased mortality rates. Since that time, there has been considerable debate regarding how sleep insufficiency should be defined. Laboratory studies where sleep is manipulated in an experimental protocol are preferred by some (because of their precision) and population-based studies where individuals are observed relative to habitual sleep behaviors are preferred by others (because of their generalizability).

Regarding the former, information about the physiologic and health consequences of sleep duration often come from studies that employ total sleep deprivation (defined as an experimental manipulation where an individual is kept awake for at least an entire sleep period) and partial sleep deprivation (defined as an experimental manipulation where an individual’s sleep period is restricted over a period of days). This is also sometimes called sleep restriction. Sometimes, partial sleep deprivation can be characterized as chronic partial sleep deprivation (defined as partial sleep deprivation over a period of weeks). All of these experimental manipulations can be useful to discern physiologic effects of changes in sleep duration, but they are generally poor approximations of real-world sleep. As such, total sleep deprivation, partial sleep deprivation/sleep restriction, and chronic partial sleep deprivation sacrifice generalizability for precision [2–4].

Other studies use population-based studies of sleep. These studies can characterize habitual sleep duration (defined as typical perceived sleep duration experienced in real-world settings), often categorized as short sleep duration, normal/normative sleep duration, and long sleep duration based on cutoffs that often vary by study. These studies may also model sleep loss (reduction in sleep duration over time). They may also capture aspects of sleep continuity, including total sleep time (calculated sleep duration based on time in bed, subtracting sleep latency and wake time after sleep onset). These parameters may be assessed retrospectively (e.g., through surveys and questionnaires) or prospectively; prospective assessments can be subjective (e.g., sleep diary) or objective (e.g., actigraphy). These studies often sacrifice precision for generalizability [2–4].

But what is insufficient sleep? Often, terms such as sleep deprivation, sleep loss, short sleep, and others are used interchangeably. Also, insufficient sleep is sometimes used interchangeably with concepts such as sleep deficiency (insufficient sleep duration or inadequate sleep quality), poor sleep quality, and even insomnia despite these concepts being misapplied to insufficient sleep [2–4].

With this in mind, defining insufficient sleep has been problematic, since all of these concepts have appropriated the label of insufficient sleep. For the purposes of this chapter, insufficient sleep will refer to sleep duration that is likely too brief to meet physiologic needs. Also, this chapter focuses on habitual sleep duration in the population and thus experimental terms such as sleep deprivation are not appropriate. Even at the population level, there is disagreement regarding how much sleep is insufficient. Various studies use cutoffs of 4, 5, 6, or 7 h as representing insufficient sleep.

Recently, a consensus panel was convened by the American Academy of Sleep Medicine and Sleep Research Society to determine the recommended amount of sleep for a healthy adult. This panel recommended that 7 or more hours was recommended [5, 6]. In a follow-up manuscript, the panel members discussed in detail how this was reached, pointing out that the consensus was most clear that 6 h or less was likely insufficient and less clear for sleep durations between 6 and 7 h [5, 6]. This finding was echoed in similar consensus statements issues by the National Sleep Foundation [7, 8], the American Thoracic Society [9], and the American Heart Association [10]. Therefore, for the purposes of this chapter, insufficient sleep will generally refer to habitual sleep duration of 6 h or less.

Prevalence of insufficient sleep

In order to estimate the prevalence of insufficient sleep in the population, data sources that assess habitual sleep duration in large samples that are representative of the general population. Existing work in this area is limited, as most studies that investigate sleep in such samples do so without using well-validated assessments of sleep. It is important to note that most population estimates of habitual sleep duration are based on subjective, retrospective self-report, which presents biases in assessing sleep [11, 12]. These estimates may better reflect time in bed than actual physiologic sleep and should be interpreted with appropriate caution.

Insufficient sleep in the population

Estimates of the prevalence of insufficient sleep have used the Behavioral Risk Factor Surveillance System (BRFSS) in the United States. The BRFSS is an annual telephone survey of hundreds of thousands of US adults, conducted by the Centers for Disease Control and Prevention (CDC) (http://www.cdc.gov/brfss). It is state-based, with population-weighted samples representing each strata of age, sex, race/ethnicity, and geographic region. Sleep duration in the BRFSS is assessed with the item, On average, how many hours of sleep do you get in a 24-h period? Responses are coded in whole numbers. Liu and colleagues reported population-weighted prevalence estimates for sleep duration around a cutoff of 7 h (based on the consensus statement [13] from the 2014 BRFSS (N = 444,306). Overall, the age-adjusted estimated prevalence of insufficient sleep (≤ 6 h) was reported to be 35.1% of the US population. Grandner and colleagues [14] reported prevalence estimates also using the 2014 BRFSS. Estimated prevalence by hour was calculated, such that the estimated prevalence by hour of sleep duration was 1.12% for ≤ 3 h, 3.19% for 4 h, 7.75% for 5 h, 23.55% for 6 h, 28.72% for 7 h, 27.64% for 8 h, 4.42% for 9 h, 2.35% for 10 h, and 1.27% for ≥ 11 h. See Fig. 2.1 for a graphical representation of these data.

Fig. 2.1 Distribution of sleep duration in the US Population using 2014 BRFSS. Data from Grandner MA, Seixas A, Shetty S, Shenoy S. Sleep duration and diabetes risk: population trends and potential mechanisms. Curr Diab Rep 2016;16(11):106. PubMed PMID: 27664039.

Other prevalence estimates have also been calculated using the National Health and Nutrition Examination Survey (NHANES). The NHANES is a survey that is also conducted by the CDC that includes a nationally-representative sample (http://www.cdc.gov/nchs/nhanes). The sample size is much smaller than the BRFSS, though reliability of data may be improved since surveys were administered in person rather than over the phone. Similar to the BRFSS, NHANES assesses sleep duration by whole number hour (no partial hours). Unlike the BRFSS, though, NHANES assesses sleep duration with the item, How much sleep do you usually get at night on weekdays or workdays? Thus, this item may capture modal nighttime sleep, rather than 24-h sleep, which may include naps. Using the 2007–2008 wave of NHANES, Grandner and colleagues calculated prevalence estimates for sleep duration by category, with 4.96% reporting ≤ 4 h, 32.16% reporting 5–6 h, 55.68% reporting 7–8 h, and 7.20% reporting ≥ 9 h [15]. Thus, insufficient sleep (≤ 6 h) was reported by 37.12% of the US population. The higher estimate relative to BRFSS may be explained by the wording of the item, which does not include naps or weekends. See Fig. 2.1 for an illustration of these values.

Lower estimates of short sleep duration are reported by Basner and colleagues using data from the American Time Use Survey (ATUS) [16]. The ATUS is conducted annually by the US Bureau of Labor Statistics and assigns activity codes to each 15-min increment of the 24-h day in a representative sample of US adults (http://www.bls.gov/tus). Because ATUS does not distinguish time in bed from time asleep, values will generally overestimate sleep and understate insufficient sleep [16]. Using ATUS from 2003 to 2011 (N = 124,517), the estimated prevalence of insufficient sleep (≤ 6 h) was 10.6%, compared to 78.4% for 6–11 h and 11.0% for ≥ 11 h.

Thus, estimates for insufficient sleep (≤ 6 h) from relatively recent, nationally representative surveys, are 10.6% from ATUS, 35.1% from BRFSS, and 37.12% from NHANES. These may vary as a result of the survey item asked, as well as other factors including the years included and sampling methodologies. Although other studies have examined large samples using more well-validated measures, none of these studies are nationally-representative and thus cannot be used to develop population prevalence estimates.

Rather than assess insufficient sleep relative to a benchmark (sleep hours), an alternative approach would be to ask individuals how often they perceive their sleep to be insufficient. The 2008 BRFSS asked, During the past 30 days, for about how many days have you felt that you did not get enough rest or sleep? Based on this variable, Mcknight-Eily and colleagues [17] reported prevalence estimates based on responses to this variable. They estimate that 30.7% of the population reports 0/30 days of insufficient sleep, with 1–13 days reported by 41.3% of the population, 14–29 days reported by 16.8% of the population, and 30/30 days reported by 11.1% of the population. Based on these estimates, 27.9% of the US population reports perceived sleep insufficiency at least 2 weeks out of the month. Interestingly, this estimate is similar to the ~ 1/3 of the population who experience insufficient sleep based on sleep duration, though the overlap between these groups is only moderate [18].

Insufficient sleep by age

Based on BRFSS data, Liu and colleagues [19] provided age-based prevalence estimates for insufficient sleep (≤ 6 h). They reported estimated of 32.2% for those age 18–24, 37.9% for 25–34, 38.3% for 35–44, 37.3% for 45–64, and 26.3% for those 65 or older (see Fig. 2.2). Of note, the lowest rate of insufficient sleep was seen among the oldest adults. This is consistent with other studies that showed that perceived insufficient sleep declines with age [20], as does self-reported sleep disturbance [21–23]. This is in contrast to more objective sleep disturbances, which are well-characterized to increase in older adults [24–26]. There are a number of potential reasons for this, including retirement offering greater sleep opportunity and differing expectations regarding sleep [27].

Fig. 2.2 Insufficient sleep (6 h or less) by age.

Similar prevalence estimates of sleep duration by age in NHANES were reported by Grandner and colleagues [15]. Among teenagers aged 16–17, prevalence of sleep duration was 0.63% for ≤ 4 h, 19.38% for 5–6 h, 62.47% for 7–8 h, and 17.52% for ≥ 9 h. For younger adults aged 18–30, prevalence was 4.83% for ≤ 4 h, 31.02% for 5–6 h, 54.44% for 7–8 h, and 9.81% for ≥ 9 h. For adults aged 30–50, prevalence was 5.86% for ≤ 4 h, 33.61% for 5–6 h, 55.49% for 7–8 h, and 5.03% for ≥ 9 h. For adults aged 50–65, prevalence was 4.95% for ≤ 4 h, 35.41% for 5–6 h, 56.04% for 7–8 h, and 3.61% for ≥ 9 h. For older adults 65 and older, prevalence was 4.17% for ≤ 4 h, 28.31% for 5–6 h, 55.58% for 7–8 h, and 11.94% for ≥ 9 h. Thus, prevalence of insufficient sleep (≤ 6 h) was reported to be 20.01% for those aged 16–17, 35.85% for those aged 18–30, 39.47% for adults 30–50, 40.36% for adults age 50–65, and 32.48% for older adults over 65. Again, prevalence of insufficient sleep is highest in working age adults.

Using the ATUS data, Basner and colleagues [16] found that, compared to 15–24 year olds, increased likelihood of insufficient sleep (≤ 6 h) was seen in those aged 25–34 (OR = 1.38; 95% CI = 1.18;1.61), 35–44 (OR = 1.40; 95% CI = 1.22;1.62), 45–54 (OR = 1.68; 95% CI = 1.44;1.94), and 55–64 (OR = 1.41; 95% CI = 1.18;1.68), but not those 65 or older. Similarly, shortest sleep durations were seen in working age adults.

Using self-reported insufficiency from the BRFSS, Mcknight-Eily and colleagues [17] report that the prevalence of self-reported insufficient sleep at least 14 of the past 30 days was reported by 31.3% of 18–24 year olds. Estimated prevalence was 34.2% for 35–34 year olds, 32.1% for 35–44 year olds, 27.2% for 45–64 year olds, and 15.0% for those 65 or older.

Insufficient sleep by sex

Several studies have examined sex relative to insufficient sleep. Liu and colleagues reports that based on the 2014 BRFSS data, insufficient sleep (≤ 6 h) is reported by 35.4% of men and 34.8% of women [19]. Using data from the 2007–2008 NHANES, Whinnery and colleagues report no sex differences in likelihood of insufficient sleep (though they report that women are 35% less likely to report long sleep duration after adjusting for covariates) [28]. Using NHIS data, Krueger and Friedman report that men are 7% less likely to report ≤ 5 vs 7 h of sleep [29]. Basner and colleagues report that men are more likely to report insufficient sleep (OR = 1.27; 95% CI = 1.20; 1.35) [16]. McKnight-Eily reports that self-reported insufficient sleep at least 14 out of the past 30 days was reported by 25.5% of men and 30.4% of women [17]. Taken together, sex differences in insufficient sleep are likely small and difficult to observe. This is in contrast to self-reported sleep disturbances, which are much more prevalent in women [30–32].

Insufficient sleep by race/ethnicity

Many studies have documented differences in sleep duration by race/ethnicity. In general, racial/ethnic minorities are more likely to experience insufficient sleep duration. Actigraphic studies have shown that racial/ethnic minorities demonstrate a sleep duration between 40 and 60 min less than non-Hispanic White counterparts [33–35].

More data are available from survey studies that included larger numbers of people but lack the precision of objective measurements. For example, data from the NHIS has shown that sleep duration of 6 h or less was more prevalent among Blacks/African-Americans, non-Mexican Hispanics/Latinos, and Asians/Others, compared to non-Hispanic Whites [36, 37]. Longitudinal analysis of NHIS data suggests that Black-White differences in insufficient sleep have persisted, relatively unchanged since 1977 [38, 39]. See Fig. 2.3 for an illustration of this.

Fig. 2.3 Black-white differences in 7–8 h sleep in the US Population in NHIS. Data from Jean-Louis G, Grandner MA, Youngstedt SD, Williams NJ, Zizi F, Sarpong DF, Ogedegbe GG. Differential increase in prevalence estimates of inadequate sleep among black and white Americans. BMC Public Health 2015;15:1185. PubMed PMID: 26611643; PMCID: PMC4661980; Jean-Louis G, Youngstedt S, Grandner M, Williams NJ, Sarpong D, Zizi F, Ogedegbe G. Unequal burden of sleep-related obesity among black and white Americans. Sleep Health. 2015;1(3):169–176. PubMed PMID: 26937487; PMCID: PMC4770938.

Other population-level studies have found similar patterns. For example, Stamatakis showed in the Alameda County study that African-Americans were about twice as likely to report short sleep duration [40]. Using NHANES data, Whinnery and colleagues showed that Blacks/African-Americans are about 2.5 times as likely to sleep < 5 h and about twice as likely to sleep 5–6 h, compared to non-Hispanic Whites. Non-Mexican Hispanics/Latinos were about 2.7 times as likely to sleep < 5 h and Asians/Others were about four times as likely to sleep < 5 h and about twice as likely to sleep 5–6 h. Mexican-Americans were the only minority group not more likely to report insufficient sleep [28].

Insufficient sleep by socioeconomic status

Perhaps due to environmental stressors, those of lower socioeconomic status are more likely to experience insufficient sleep. Kruger and Friedman used NHIS data to compute mean family income according to sleep duration [29]. They found that the highest mean income was reported among 7-h sleepers ($48,065), with the lowest income levels in those sleeping 5 h or less ($36,819) or 9 h or more ($34,883). Stamatakis evaluated likelihood of insufficient sleep relative to income quintile [40]. This study reported that compared to the highest income quintile, short sleep duration (6 h or less) was increasingly reported in the fourth (3% more likely), third (11% more likely), second (29% more likely), and first quintile (54% more likely). Using BRFSS data, days of perceived insufficient sleep decreased at higher levels of household income [20].

Using NHANES data, Whinnery and colleagues examined several socioeconomic indices relative to sleep duration [28]. Compared to those with family income over $75,000, increased likelihood of < 5 h of sleep (P < 0.05) was observed for all categories, including <$20,000 (OR = 5.5), $20,000–$25,000 (OR = 2.9), $25,000–$35,000 (OR = 4.1), $35,000–$45,000 (OR = 2.4), $45,000–$55,000 (OR = 2.8), $55,000–$65,000 (OR = 2.4), and even $65,000–$75,000 (OR = 3.8). Increased likelihood of 5–6 h sleep relative to those earning over $75,000 was only seen in the lowest income group earning <$20,000 (OR = 1.3). Education level was another socioeconomic indicator that was associated with sleep duration in this sample. Those with less than a high school education were approximately four times as likely to report < 5 h of sleep, compared to college graduates. Similarly, those who completed some high school were more likely than college graduates to report < 5 (OR = 5.3) and 5–6 (OR = 1.7) hours of sleep, those who completed high school were more likely than college graduates to report < 5 (OR = 4.3) or 5–6 (OR = 1.6) hours, and those with some college were also more likely than college graduates to report < 5 (OR = 3.6) or 5–6 (OR = 1.6) hours of sleep [28]. Another socioeconomic indicator evaluated in this study was lack of access to healthcare, which was more common among those reporting < 5 h of sleep. Food insecurity—a measure of inability to financially provide healthy access to enough food—was also more common among those reporting < 5 and 5–6 h of sleep [28].

Insufficient sleep by geography

Insufficient sleep in the United States is differentially experienced across varying regions of the country. An analysis of self-reported perceived insufficient sleep using BRFSS data was reported [41]. Using a geospatial hotspot analysis, several key hotspots of insufficient sleep were identified in the United States, including parts of the southeast, parts of the Texas/Louisiana border, areas in the Midwest, and the largest hotspot in central Appalachia. Coldspots with abnormally low levels of insufficient sleep were seen in the northern Midwest (Wisconsin/Minnesota/Iowa), central Texas, central Virginia, and areas in along the West Coast. See Fig. 2.4 for a map of US counties relative to their proportion of insufficient sleep and Fig. 2.5 for a map of hotspots and coldspots.

Fig. 2.4 County-level insufficient sleep in the US. From Grandner MA, Smith TE, Jackson N, Jackson T, Burgard S, Branas C. Geographic distribution of insufficient sleep across the United States: a county-level hotspot analysis. Sleep Health 2015;1(3):158–165. PubMed PMID: 26989761; PMCID: 4790125.

Fig. 2.5 Hotspots and coldspots of insufficient sleep in the US. From Grandner MA, Smith TE, Jackson N, Jackson T, Burgard S, Branas C. Geographic distribution of insufficient sleep across the United States: a county-level hotspot analysis. Sleep Health 2015;1(3):158–165. PubMed PMID: 26989761; PMCID: 4790125.

Rather than examine statistical hotspots of perceived insufficient sleep, researchers at the CDC used BRFSS data to map prevalence of ≤ 6 h of sleep across the United States¹⁹. The US states with the highest prevalence were (in order) Hawaii (43.9%), Kentucky (39.7%), Maryland (38.9%), Alabama (38.8%), Georgia (38.7%) and Michigan (38.7%). The US states with the lowest prevalence were (in order) South Dakota (28.4%), Colorado (28.5%), Minnesota (29.2%), Nebraska (30.4%), and Idaho (30.6%).

Key limitations to population estimates of insufficient sleep

There are several key limitations in the existing literature on insufficient sleep epidemiology. First, there is a lack of clarity of gold-standard methods for estimating population levels of sleep duration. Most of these studies used single-item self-report measures from surveys, which are fraught with psychometric problems [2, 11, 42]. Not only do self-report measures tend to over-report sleep relative to physiologic recordings and likely better approximate time in bed than physiologic sleep, they may be subject to a number of other biases, demand characteristics, and social desirability. There still exists no nationally-representative dataset that estimates sleep duration based on gold-standard approaches, especially those that record physiologic sleep.

Second, the definition of insufficient sleep varies widely across studies, and most studies do not allow enough resolution to examine different cutoffs. Given recent consensus statements [5–7, 9, 10], a cutoff of 7 h seems reasonable, but there is yet no clear consensus on the range between 6 and 7 h, where many Americans fall regarding their typical sleep habits. Also, it is not clear whether a determination of insufficient sleep should be made on the basis of physiologic sleep or perceived sleep.

Third, definitions of insufficient sleep are based on nomothetic, population-level recommendations which don’t take into account individual differences in sleep need, sleep ability, and resilience to sleep loss. Also, these do not necessarily take into account sleep sufficiency relative to any particular outcome. Future work should consider these issues in order to take a more personalized/precision medicine view of sleep duration, as it relates to an individual and impacts on specific outcome measures, in a specific set of contexts.

Prevalence of poor sleep quality

Poor sleep quality, like insufficient sleep, has been variably defined. The National Sleep Foundation has recently attempted to develop a coherent conceptualization of sleep quality [43, 44]. In a consensus document, elements of sleep quality included sleep latency (amount of time to fall asleep), wake time after sleep onset (amount of time awake at night), and sleep efficiency (proportion of the time in bed spent sleeping). Thus, sleep quality was generally defined as good sleep continuity. Recognizing the limitations of this, the National Sleep Foundation has begun work on a tool to measure sleep satisfaction with is presented as another key element of overall sleep quality [45]. In addition to sleep-focused elements as indicators of sleep quality, perhaps daytime indicators can be useful as well. For example, daytime sleepiness is often an indicator of poor nighttime sleep [46, 47] and may also serve as an indicator of poor sleep quality.

Prevalence of sleep disorders

Poor sleep quality can refer to a relatively wide range of problems, including sleep disorders as well as sleep symptoms. The most common types of sleep disorders in the population are insomnia and sleep apnea. Although other chapters in this volume focus specifically on these issues at the population level, it is important to note that the population prevalence of acute insomnia is high (about 4% per month) [48, 49] and that although most of these resolve, approximately 10% of the population likely meets criteria for an insomnia disorder [50, 51].

Regarding sleep apnea, prevalence estimates need to account for sex and body mass index. Relatively recent estimates of the prevalence of sleep apnea estimate that among men age 30–49, rates are 7.0%, 18.3%, 44.6%, and 79.5% for those with BMI of < 25, 25–29.9, 30–39.9, and 40 or above, respectively. For men 50–70, the rates increase to 18.9%, 36.6%, 61.4%, and 82.8%, respectively. For women age 30–49, the rates of sleep apnea are lower, at 1.4$, 4.2%, 13.5%, and 43.0% for women with a BMI of < 25, 25–29.9, 30–39.9, and 40 or higher, respectively. As with men, these numbers are higher in women age 50–70, with 9.3%, 20.2%, 41.1%, and 67.9% with sleep apnea among those with BMI of < 25, 25–29.9, 30–39.9, and 40 or greater, respectively. This high prevalence of sleep apnea (Fig. 2.6) is particularly notable [52], especially since recent estimates suggest that approximately 85% of sleep apnea cases are never diagnosed, and up to half of diagnosed cases remain insufficiently treated [53].

Fig. 2.6 Estimated prevalence of sleep apnea by age group, sex, and BMI. From Peppard PE, Young T, Barnet JH, Palta M, Hagen EW, Hla KM. Increased prevalence of sleep-disordered breathing in adults. Am J Epidemiol. 2013. PubMed PMID: 23589584; PMCID: 3639722.

Regarding circadian rhythm sleep disorders, the prevalence of delayed sleep phase disorder is estimated to be about 0.2% of the general population but 7–16% of adolescents [54]. Prevalence of other circadian rhythm sleep disorders is largely unknown, though the prevalence of shift work disorder is estimated to be about 5–10% of the population, based on prevalence estimates of night shift work and the prevalence of the disorder among shift workers [55].

Prevalence of sleep complaints

Several studies have examined prevalence of sleep complaints in the general population. For example, Grandner and colleagues [21] found that the rate of general sleep disturbance in the US population was about 16% in men and 21% in women, and general daytime fatigue was 18% in men and 26% in women. However, this depended on age. Fig. 2.7 depicts the rates of these across age groups, illustrating a general decline in reports with age. Of note, in women, increased sleep duration and