C i rca d i a n N e u ro b i o l o g y a n d

t h e Ph y s i o l o g i c R e g u l a t i o n
o f S l e e p a n d Wa k e f u l n e s s
a,b, c
William J. Schwartz, MD *, Elizabeth B. Klerman, MD, PhD

KEYWORDS
 Circadian rhythms  Suprachiasmatic nucleus  Sleep  Homeostasis  Physiology

KEY POINTS
 Precise, robust, and flexible circadian rhythmicity is the product of a network of tissue and
organ clocks, including a master clock in the suprachiasmatic nucleus of the
hypothalamus.
 The timing, duration, and content of sleep are regulated by nonlinear interactions between
circadian and sleep homeostatic processes and the environmental and behavioral vari-
ables that affect them.
 Changes in the time course, strength, or alignment of circadian and homeostatic drives,
either environmentally or pathologically, can begin to account for disordered sleep
rhythms and suggest rational therapeutic approaches.

INTRODUCTION

The Earth’s 24-hour axial rotation has forced most organisms to evolve and adapt to a
dramatic recurring cycle of day and night. Endogenous circadian timekeeping sys-
tems are one of evolution’s responses to this challenge; these timekeeping systems
allow living things to recognize local environmental time and to measure time’s pas-
sage. The result is an ability to anticipate periodic daily events, orchestrate internal
temporal programs of behavioral activities and metabolic (and other physiologic)

Disclosure Statement: This article is funded by NIH Grant numbers: P01-AG009975; R21-
HD086392; K24-HL105664; R01-GM105018; R01-HL128538.
Funded by: NIHHYB Grant number(s): NIA P01-AG009975; NICHD R21-HD086392; NIH K24-
HL105664; R01-GM-105018; R01-HL128538 NIHMS-ID: 1524361
a
Department of Neurology, Dell Medical School, The University of Texas at Austin, Health
Discovery Building 5.702, 1601 Trinity Street Building B, Austin, TX 78712, USA; b Department
of Integrative Biology, College of Natural Sciences, The University of Texas at Austin, Austin, TX,
USA; c Division of Sleep and Circadian Disorders, Department of Medicine, Brigham and
Women’s Hospital, 221 Longwood Avenue, Boston, MA 02115, USA
* Corresponding authors. Department of Neurology, Dell Medical School, The University of
Texas at Austin, Health Discovery Building 5.702, 1601 Trinity Street Building B, Austin, TX
78712.
E-mail addresses: [email protected] (W.J.S.); [email protected]
(E.B.K.)

Neurol Clin 37 (2019) 475–486

https://doi.org/10.1016/j.ncl.2019.03.001 neurologic.theclinics.com
0733-8619/19/ª 2019 Elsevier Inc. All rights reserved.
476 Schwartz & Klerman

functions, and flexibly set the order and scheduling of such activities and functions to
(presumably) optimize fitness in the natural world.1,2 This system is vital to adjusting
the timing and duration of bouts of rest and activity, eating, and other functions to
the ecological niche. The system also lies at the core of various mechanisms for suc-
cessful adaptation to the seasons3,4; it enables tracking the changing day length to
activate winter strategies to conserve energy, change coloration to avoid detection
by predators, time reproductive events, and/or escape by migrating thousands of
miles away (using the sun as a time-compensated compass for navigating across
the latitudes). The circadian clock is indeed a clock for all seasons.

THE FORMAL LANGUAGE OF CIRCADIAN CLOCKS AND RHYTHMS

Circadian rhythms are innate, persisting with an approximately 24-hour oscillation in

organisms and across generations in constant environmental conditions (eg, the
rhythms free-run in constant darkness), and are accurate and precise, even in the
face of changing temperatures. Like all rhythmic phenomena, their properties can
be described in mathematical terms, such as period, phase, and amplitude (Fig. 1).
Circadian rhythmicity represents the overt manifestation of an internal pacemaker
that functions as a clock through the daily resetting of its approximately 24-hour oscil-
lation by environmental 24-hour cues, thus adopting a stable phase relationship to the
environment (ie, entrainment). Of the possible multitude of entraining cues (zeitgebers),
light is the most powerful and the best studied. Photic phase response curves rigor-
ously quantify the resetting responses of rhythms to light pulses presented across
the circadian cycle, revealing how the interaction of an endogenous oscillation with a
rhythm of light responsiveness can lead to precise and accurate entrainment to the nat-
ural light:dark (day:night) cycle and shift it to a new phase (eg, after travel across time
zones). For entrainment, the clock is reset to match its intrinsic period to 24 hours; light
administered just before dawn advances its rhythm, whereas exposure to an identical
light stimulus (in duration, wavelength, and intensity) after dusk delays its rhythm. So, to
entrain the clock of an animal with an endogenous free-running period longer than
24 hours (as is typical for humans), daily morning light acts to increase the clock’s speed
and advance its phase. Importantly, a stimulus also can affect the expression of a
rhythm without entraining its underlying clock; such a mechanism has been termed,
masking, because it bypasses or acts downstream of the clock and thus masks the
clock’s true state. For example, studies of the core body temperature (CBT) rhythm
in humans (which is significantly influenced by the circadian clock) have revealed
that masking factors, such as light at night, activity levels, postural changes, meal
times, and sleep, may alter its value significantly5; therefore, determining the endoge-
nous circadian rhythm in CBT requires specialized conditions in which these masking
factors are minimized or spread evenly over 24 hours (discussed later).
Although historically the search for the clock focused on its self-sustained rhythmicity,
it is the phase of entrainment (eg, the timing of sleep onset, temperature minimum, or
locomotor activity peak in relation to the light:dark cycle) that matters for survival in
the wild.6 The clock’s free-running period and phase of entrainment are systematically
related to each other, such that faster clocks (shorter period length) tend to lead to
earlier phase of entrainment; but entrainment phase also is affected by other factors,
including the intensity of light and prior exposure to different day lengths. The entrain-
ment phase is sometimes referred to as chronotype, or the temporal phenotype of an in-
dividual or a population (as in morning larks and night owls), and has provided a window
for understanding how genetic determinants and ecological constraints (including those
imposed by own urban living and societal mores) shape rhythmic behaviors.7
 Physiologic Regulation of Circadian Rhythms and Sleep 477

Fig. 1. Terminology and displays. (A) A schematic circadian rhythm, on the left, entrained to
a 12-hour:12-hour light:dark cycle and, on the right, free-running under constant darkness.
White bar, light; black bar, dark. Under entrainment, zeitgeber time (ZT ) 0 represents the
time of lights on (dawn) and ZT 12 the time of lights off (dusk); during free-running, circa-
dian time (CT) 0 represents subjective dawn and CT 12 subjective dusk. T, period length of
the zeitgeber; t, period length of the free-running rhythm; 4, phase, a defined, stable
cycle-to-cycle reference point within the cycle of the rhythm (eg, 41, the minimum value);
J, phase angle of entrainment, the time difference (phase relationship) between a defined
phase of the rhythm and an external phase-reference (eg, zt 0). (B) A schematic circadian
rhythm in actogram format. Data (eg, locomotor activity) are plotted horizontally from
left to right over the course of 24-hour periods, with succeeding days stacked vertically
from top to bottom. LD, light:dark cycle; DD, constant darkness. (C) A schematic circular
plot of the phases of individual rhythms from 2 populations, from 0 to 24 hours (or some-
times graphed from 0 to 360 ). The solid radial arrow represents the mean phase of each
population and its length a measure of the scatter within each population. This format
for plotting angular data illustrates that a mean of 15 for the red population is not a larger
value than a mean of 3 for the blue population but rather a difference in their relative
phases.

A NETWORK FOR INTERNAL TIME

Although daily rhythms—the opening and closing of leaves in concert with the alterna-
tion of day and night, for example—have been known since antiquity, the mechanistic
understanding of daily and annual timing has blossomed over the past 50 years, now
encompassing details at the molecular, cellular, tissue, organismal, and even societal
levels. The basic oscillatory mechanism is intracellular, focused on a suite of clock
genes that function within negative autoregulatory feedback loops, rhythmically
repressing the transcription of their own mRNAs8 (see Allada, this issue). Post-
transcriptional and post-translational regulatory mechanisms, currently under active
investigation by many researchers, also clearly contribute to fundamental rhythm
478 Schwartz & Klerman

properties. Clock transcription factors do not operate solely within the clock’s feed-
back loops; they regulate the transcription of downstream clock-controlled genes in
a wide array of the cell’s metabolic and other pathways. This linkage begins to suggest
how the motion of a molecular clock can be translated into a 24-hour program of
biochemical events.
The identification of clock genes enabled the construction of transgenic mice
bearing bioluminescent reporters (eg, using clock gene promoter elements to drive
rhythmic transcription of the gene encoding the enzyme luciferase), contributing to
the discovery that cells, tissues, and organs throughout the brain and body express
circadian oscillators.9 In mammals, the suprachiasmatic nucleus (SCN) of the anterior
hypothalamus (discussed later) orchestrates this network for internal time, in which the
multiplicity of peripheral clocks exhibit defined but permutable phase relationships to
each other and the environmental light:dark cycle. This coordination is mediated by
the interplay of several coupling factors that carry the SCN’s output signals to subsid-
iary clocks—including rhythms of behavior (eg, rest and activity, feeding, and fasting),
CBT, hormone levels (eg, cortisol, melatonin), and neural activity—in various combina-
tions and strengths10 (Fig. 2). This system can ensure homeostasis over time; as an
example, the hepatic clock’s proper timing of compensatory glycogenolysis and
glucose export enables maintenance of plasma glucose levels through nighttime fast-
ing.11,12 By tuning coupling strengths and oscillator speeds, the network should be
capable of adaptively realigning the relative phasing of its components under chang-
ing internal and external conditions. Such short-term circumstances might include
food availability at an unexpected time as well as social interactions; note that during
development, the phase of the hepatic clock in rat pups reverses around the time of
weaning, corresponding to the switch from diurnal nursing to nocturnal feeding.13
On the other hand, pathologic misalignment—between the environment and behavior,
between behavior and the SCN, between the SCN and peripheral clocks, or between
peripheral clocks—might be both a consequence and a cause of at least some of the
symptoms of aging and disease (see Zee, this issue).

THE SUPRACHIASMATIC NUCLEUS AS A MAMMALIAN MASTER CLOCK

The SCN is a bilaterally paired hypothalamic nucleus straddling the midline, bordering
the third ventricle above the optic chiasm.14 Although present in diverse species, it has
been studied most intensively in rodents. Its metabolic and neuronal spike activities
exhibit circadian rhythms in vivo and in vitro; in vivo, there is a close congruence be-
tween the rhythm of electrical activity and the temporal profile of behavioral rest and
activity, with feedback effects of locomotion and sleep shaping the waveform of SCN
firing rate.15,16 Behavioral, physiologic, and hormonal circadian rhythms depend on
the integrity of the SCN; ablation of the nucleus results in circadian arrhythmicity
without disrupting homeostatic mechanisms, whereas neural grafts of fetal SCN tissue
re-establish behavioral rhythmicity in arrhythmic SCN-lesioned recipients, with the
rhythms restored by the transplants exhibiting properties characteristic of the circa-
dian pacemakers of the donors rather than those of the hosts.17
Photic entrainment of SCN rhythmicity relies on a monosynaptic retinal input that
primarily codes for luminance, involves signaling by glutamate and pituitary adenylate
cyclase-activating peptide, and reflects the activity of a class of melanopsin-
expressing, intrinsically photoreceptive retinal ganglion cells (ipRGCs); other ipRGCs
seem to subserve light-modulatory effects on aspects of sleep and mood.18 SCN neu-
ral outputs are mostly intrahypothalamic but with close connections to key targets.
Via the ventral subparaventricular zone, these include polysynaptic pathways to
 Physiologic Regulation of Circadian Rhythms and Sleep 479

Fig. 2. A network for internal time. Simplified cartoon of the mammalian circadian time-
keeping system, with the SCN entrained by the light:dark cycle and peripheral body clocks
entrained by SCN output signals, including rhythmic behaviors, CBT, hormone levels, and
nervous system activity; not shown are other clocks within the nervous system itself. These
subsidiary clocks and rhythms also can respond to inputs downstream of the SCN (eg, a
shifted cycle of food availability on the hepatic clock). Of possible feedbacks, 2 exemplary
ones are illustrated (ie, locomotor activity itself affecting CBT; adrenal output itself acting
as a coupling signal).

sleep-active and wake-active cell groups, suggesting a circadian influence on both

behavioral states that is time dependent.19,20 The phases of intrinsic SCN rhythms
are similar in diurnally active and nocturnally active species, suggesting that the
differing chronotypes of these species require inversion of SCN output signals or their
reception.21 Another important output is via the sympathetic system and its innerva-
tion of the pineal gland, which drives the nocturnal synthesis and secretion of mela-
tonin. Both diurnal and nocturnal species display high melatonin levels during the
night (dark phase), with its duration inversely related to the length of the day (light
phase).3,4 There also is evidence implicating SCN output via diffusible substances,
for example, transforming growth factor a22 and prokineticin 2.23
The circadian clock in the SCN is itself a complex coupled network of heterogeneous
neuronal (and glial) oscillators.24,25 Anatomic modules have been defined by neuro-
transmitters and other biomarkers, with functional consequences following their
480 Schwartz & Klerman

specific genetic manipulation: a retino-recipient region (referred to as core) that in-

cludes neurons expressing vasoactive intestinal polypeptide and gastrin-releasing
peptide and an arginine vasopressin neuronal shell and other identified cell groups
that overlap these subdivisions (eg, expressing neuromedin S or D1 dopamine recep-
tor). Somehow these anatomic units are dynamically coupled to yield emergent prop-
erties of the SCN clock that are critical to its function. Circuit-level mechanisms enable
clock precision (generation of a stable ensemble period from sloppy cellular oscilla-
tions), robustness (persistence in the face of genetic perturbations), and plasticity
(cellular-phase dispersion in response to changing day length). One such mechanism
involves the interaction of SCN molecular and electrical activities. The intracellular mo-
lecular clock governs electrical activity through an array of ion channels that switch
resting membrane potential from daytime depolarization to nighttime hyperpolarization
and sustain the level of neuronal firing; in turn, electrical activity regulates calcium-
dependent transcription and thus the phase and amplitude of the molecular clock. Un-
der investigation are additional mechanisms (eg, state-dependent actions of vasoac-
tive intestinal polypeptide and g-aminobutyric acid, small-world network topology) to
account for the complex spatiotemporal activity patterns exhibited by SCN tissue.

ON SLEEP AND WAKEFULNESS

The most obvious circadian-regulated output is sleep and wake timing. Sleep is a
complex multisite and multifunction behavior. Overtly, it involves changes in con-
sciousness, responsiveness, and posture. Within the organism, it causes changes in
all physiologic systems. Unlike coma and anesthesia, sleep is unique in its ability to,
without external pharmacologic manipulation, rapidly and reversibly change to a fully
functioning wake state.26 Considerable progress is being made in elucidating the neu-
ral circuitry underlying the regulation of sleep:wake states and their transitions.27,28
Sleep is required for normal physiology; death occurs if there is long-term sleep
deprivation.29 The multiple changes that occur during sleep are important for the entire
body, not just for the brain. Surprisingly, the function of sleep is unknown. As Dr Alan
Rechtschaffen famously said, “If sleep doesn’t serve an absolutely vital function, it is
the greatest mistake evolution ever made.”30 During the time animals sleep, they are
not procreating, feeding, or defending themselves—3 important behaviors for sustain-
ing the species. Therefore, the fact that sleep or sleep-like behavior has persisted in
every animal species suggests that it is required for normal physiologic functioning.
There are probably multiple functions of—or, more properly, functions that occur dur-
ing—sleep, suggesting that sleep is the platform during which multiple physiologic
processes occur. Sleep is not just required for feeling rested; metabolic, immunologic,
cognitive, and other physiologic work occur during sleep. Each of these physiologic
processes probably has its own time course of buildup and decay, so enough sleep
for one function may be different than for another. Therefore, completing these func-
tions would be expected to be part of the regulation of timing or duration of sleep.
Even sleep is not uniform; it is composed of rapid eye movement (REM) sleep and
non-REM (NREM) sleep, which are differentially regulated (discussed later). More
studies are needed to determine the time course of buildup and decay of each of these
processes, because the timing and duration of sleep would also be expected to be
regulated to increase the chance of survival.

CIRCADIAN AND HOMEOSTATIC REGULATION OF SLEEP TIMING AND CONTENT

The two-process model31 is the dominant framework for understanding the timing and
content of sleep behavior at the level of the whole organism and specific brain areas.
 Physiologic Regulation of Circadian Rhythms and Sleep 481

The 2 processes are circadian rhythmicity (process C) and sleep homeostasis (pro-
cess S). Sleep homeostasis increases during wake (and possibly REM sleep) and de-
creases during sleep (or possibly just NREM sleep). Accepted markers of sleep
homeostasis across 1 or 2 sleep and wake episodes (eg, wake episodes lasting
4–60 hours and daytime and/or nighttime sleep episodes) are slow wave sleep
(SWS) (N3 under current American Academy of Sleep Medicine criteria) or slow
wave activity (SWA) (approximately 0.5–4.5 Hz during sleep and/or approximately
12.25–25.0 Hz32 during wake in the electroencephalogram [EEG]); the putative
biochemical agent is adenosine. All these markers have levels that increase with
wake duration and decrease during sleep. Consistent with this is that caffeine, an
adenosine receptor antagonist, decreases sleepiness. The buildup and decay of pro-
cess S are modeled mathematically as saturating exponentials—consistent with data
that lost sleep is not recovered minute-by-minute and, therefore, an intensity factor
(eg, increased SWA) must be involved. For a longer time course of sleep homeostasis,
as seen with chronic sleep restriction (ie, multiple nights with insufficient sleep), SWS
and SWA are not appropriate markers and the time course of buildup and recovery is
different than for sleep deprivation (ie, 1 continuous wake episode)33; a hypothesized
biochemical change is in adenosine receptor concentrations.34

PROTOCOLS FOR SEPARATING CIRCADIAN AND SLEEP:WAKE INFLUENCES

A difficulty in determining the relative effect of circadian rhythmicity and sleep homeo-
static factors is that circadian timing (phase) and sleep homeostatic buildup and decay
covary: on a 24-hour schedule with sleep at night and wake during the day, both circa-
dian phase and the length of time awake (during which sleep homeostasis increases)
or asleep (during which sleep homeostasis decreases) advance at the same rate. For
example, suppose circadian time 0 is set as 5:00 AM, wake time at 7:00 AM, and sleep
onset time at 11:00 PM. Then wake time is at circadian time 2; 5 hours after wake time is
circadian time 7; and sleep onset time (after 16 hours awake) is circadian time 18. It
can never be learned what happens when circadian time is 18 but wake duration is
2 hours, such as might happen with someone working the night shift or experiencing
jet lag. An additional difficulty in determining circadian phase or sleep:wake influence
on a variable is that sleep and wake are associated with multiple other differences be-
sides the sleep or wake state; these include differences in physical and mental activity
levels, fasting/feeding, social interactions, lighting conditions, and posture. Therefore,
sleep or wake per se may not be the only determining factor influencing the variable(s)
of interest.
For these reasons, several types of protocols have been developed to separate the
circadian and homeostatic influences on sleep:wake physiology. In two of these proto-
cols, subjects are not allowed to choose their sleep and wake times, change lighting
conditions (which usually are very dim to minimize alerting and phase-shifting effects
of light), or know clock time. One protocol is the constant routine, in which the subjects
remain awake for 40 hours or longer in constant posture and dim light conditions and
eat multiple small meals; this enables study of wake at all circadian phases and espe-
cially at 2 different sleep homeostatic pressures (24 hours apart) for each circadian
phase and under conditions in which masking (defined previously) is minimized or
distributed equally across all circadian phases. This protocol, however, induces sleep
deprivation; it, therefore, is of great use for studying hormones, performance, and car-
diovascular, metabolic, immune, and other physiologic functions but of limited use for
studying the independent effects of circadian rhythmicity and sleep homeostasis on
sleep. The other protocol is a forced desynchrony protocol in which subjects live on
482 Schwartz & Klerman

non–24-hour days. Studies have been conducted with days as short as 90 minutes35
and as long as 42.85 hours.33 Within each of these days, there usually is a 2:1 wake:-
sleep (and light:dark) ratio, although other ratios can be used if chronic sleep restriction
is being studied.36 Under such schedules, the circadian pacemaker cannot entrain to
the wake:sleep or light:dark cycle, leading to desynchrony between circadian and sleep
homeostatic processes. These forced desynchrony protocols create multiple evenly
distributed combinations of length of time awake or asleep at each circadian phase,
so that separate effects of circadian rhythmicity and sleep homeostasis can be studied.
Such intensive inpatient studies have documented the findings (described later) about
the relative circadian and homeostatic influences on sleep timing and duration.
A third protocol in which individuals sometimes exhibit spontaneous desynchrony be-
tween circadian and sleep:wake cycles occurs when they are allowed to self-select their
wake:sleep and light:dark schedules.37 Under these conditions, the observed cycle
length usually is longer than the endogenous circadian pacemaker’s period of approx-
imately 24.2 hours,38 and there is no longer the opportunity to study all possible timing
combinations of circadian and homeostatic factors. The observed circadian period also
is approximately 25 hours instead of approximately 24.2 hours; this apparent discrep-
ancy can be explained by the nonuniform (across the day) light exposure,39 such that
there is more light exposure during the circadian phases that cause phase delays.
Nevertheless, spontaneous desynchrony protocols have also yielded much useful infor-
mation on the effects of circadian rhythmicity and sleep homeostasis on sleep.40

EFFECTS OF SELF-SELECTED BEHAVIORS ON CIRCADIAN RHYTHMS AND SLEEP

There also are direct and indirect effects of behaviors that may affect both circadian
rhythmicity and sleep homeostasis (Fig. 3). Light during the normal dark times can

Fig. 3. The regulation of sleep and wake. (A) Inter-relationships of factors affecting multiple
aspects of sleep and wake. (B). Diagrams of changes in time of (left) two-process model of
sleep regulation. S, process S, the sleep homeostatic factor; C, process C, the circadian
rhythm process; gray areas, time of sleep. When an individual self-selects sleep onset (1 of
the 2 vertical lines), the level of homeostatic drive is at the process S level (upper line); ho-
meostatic drive then declines in an exponential manner until it reaches the process C level.
Then the person awakens and the level of homeostatic drive begins to rise again (not
shown). (Right) Sleep durations associated with different self-selected sleep onsets.
 Physiologic Regulation of Circadian Rhythms and Sleep 483

induce sleep in rodents, and dark during the normal light times can induce wake. In
addition, as described previously, self-selected light:dark timing may affect alertness
and circadian rhythms, which then affect timing of sleep (which occurs in the dark).
Self-selected light:dark timing differentially affects early lark versus late owl chrono-
types and their circadian phase. When larks and owls can choose their own lighting
schedules, circadian phase is later in owls than larks; but when all are exposed to
the same light:dark schedule, circadian phase is approximately the same in all.41
Caffeine can directly affect the circadian system42 as well as increase alertness, which
then includes increased light that can directly affect the circadian clock.
Self-selected light:dark schedules may be involved in some circadian rhythm sleep
disorders. For example, individuals who do not choose regular light:dark schedules or
live mostly under indoor dim light conditions, blind people, or individuals with relatively
long endogenous periods who stay up late at night or for long durations (possibly
because of endogenous induced or pharmaceutically induced low sleep homeostatic
pressure buildup rate) and have light exposure that promotes circadian delays may
develop non–24-hour sleep:wake disorder (see Abbott, this issue). Or, individuals
with reduced light sensitivity or preferential late-night light exposure may develop
delayed sleep phase syndrome (see Wyatt and Culnan, this issue). A reduced rate
of rise of sleep homeostatic pressure (as postulated to occur with aging) would affect
the relative amplitudes of circadian and homeostatic influences on sleep timing, and,
together with a relative circadian phase advance, may be a reason for early morning
awakening complaints of many older people.

MARKERS OF CIRCADIAN RHYTHMS IN HUMANS

Circadian rhythms generated by the SCN heavily influence the timing of sleep onset
and offset and of the timing of REM sleep within a sleep episode. The SCN’s tight
regulation of melatonin rhythmicity (discussed previously) makes melatonin a valuable
marker of circadian timing if melatonin is collected under dim light conditions, because
ocular light exposure suppresses melatonin. Another marker of circadian timing and
amplitude is CBT; however, given the circadian phase–dependent masking effects
of sleep:wake and activity on CBT, it cannot be used as a marker of circadian phase
and amplitude except under highly controlled inpatient conditions. Melatonin (usually
dim light melatonin onset [DLMO]) and CBT (usually minimum of fit sinusoidal curve
[CBTmin]) markers are tightly linked in both phase (ie, DLMO occurring approximately
6 hours before CBTmin43) and amplitude.38,44,45 Unexpectedly, even though melatonin
levels are high when individuals are usually tired, it is not effective as a hypnotic46
when taken for insomnia and/or at an individual’s habitual sleep time. On the other
hand, melatonin or its agonists can phase shift the circadian pacemaker of sighted
and blind people47,48 and through that mechanism affect sleep. That sleep itself can
modulate pacemaker function already has been noted earlier in this review16; sleep re-
striction also can decrease the effectiveness of phase-shifting light stimuli,49 an inter-
action that may be important for some circadian rhythm sleep disorders.

CIRCADIAN AND HOMEOSTATIC EFFECTS ON SLEEP

Paradoxically, in humans, the circadian drive for sleep is strongest near the end of the
habitual sleep episode, perhaps to consolidate sleep as the homeostatic drive de-
creases during the sleep episode, whereas the circadian drive for wake is strongest
a few hours before habitual sleep during the wake maintenance zone, perhaps to
consolidate wake as sleep homeostatic drive builds up over an approximately 16-
hour wake episode.50 REM sleep is most likely to occur during the end of the habitual
484 Schwartz & Klerman

sleep episode, both because that is the time the circadian drive promotes REM sleep
and because the NREM-REM sleep competition across the night has decreased
NREM sleep pressure as sleep homeostasis declines.
The timing and content of sleep, including sleep latency and wake within a sleep
episode, are regulated by a nonlinear interaction between these circadian and sleep
homeostatic processes: the magnitude of the circadian influence depends on how
long an individual has been awake or asleep. When a person has just awoken, there
is little circadian variation in sleep latency; similarly, when a person has just fallen
asleep, there is little circadian variation in wake within a sleep episode. If the person
has been awake for a long time, however, then sleep latency depends on circadian
phase (large-amplitude circadian influence); and if the person has been asleep for a
long time, then the amount of wake within the sleep episode depends on circadian
phase. Therefore, when sleep homeostasis and circadian phase are aligned, there is
relatively short sleep latency and consolidated sleep without significant wake within
the sleep episodes.50 The two, however, are misaligned in jet lag and shift work; during
these conditions, the circadian system is promoting sleep when the individual desires
to be awake and promoting wake when the individual desires to be asleep. The
nonlinear combination of sleep homeostasis and circadian phase also affects some
EEG frequencies51 and cortical and subcortical activity in the brain as measured by
functional MRI.52

CODA

Circadian neurobiology and sleep:wake regulation affect virtually all physiologic sys-
tems, including each other. Self-selected behaviors also may affect these processes,
leading to cycles that may cause diseases or disorders (eg, jet lag, shift work, and/or
other circadian rhythm sleep disorders) and influence multiple physiologic, metabolic,
immunologic, and mental functions. Understanding the underlying physiology is
crucial for developing and testing appropriate educational, behavioral, pharmaceu-
tical, and other interventions to lessen these adverse consequences.

REFERENCES

1. Dunlap JC, Loros JJ, DeCoursey PJ, editors. Chronobiology: biological time-
keeping. Sunderland (MA): Sinauer Associates; 2009. p. 382.
2. Roenneberg T, Merrow M. The circadian clock and human health. Curr Biol 2016;
26:R432–43.
3. Paul MJ, Zucker I, Schwartz WJ. Tracking the seasons: the internal calendars of
animals. Philos Trans R Soc Lond B Biol Sci 2008;363:341–61.
4. Wood S, Loudon A. Clocks for all seasons: unwinding the roles and mechanisms
of circadian and interval timers in the hypothalamus and pituitary. J Endocrinol
2014;222:R39–59.
5. Wever RA. Internal interactions within the human circadian system: the masking
effect. Experientia 1985;41:332–42.
6. Helm B, Visser ME, Schwartz WJ, et al. Two sides of a coin: ecological and chro-
nobiological perspectives of timing in the wild. Philos Trans R Soc Lond B Biol Sci
2017;372:20160246.
7. Roenneberg T, Wirz-Justice A, Merrow M. Life between clocks: daily temporal
patterns of human chronotypes. J Biol Rhythms 2003;18:80–90.
8. Takahashi JS. Transcriptional architecture of the mammalian circadian clock. Nat
Rev Genet 2017;18:164–79.
 Physiologic Regulation of Circadian Rhythms and Sleep 485

9. Mohawk JA, Green CB, Takahashi JS. Central and peripheral circadian clocks in
mammals. Annu Rev Neurosci 2012;35:445–62.
10. Schibler U, Gotic I, Saini C, et al. Clock-talk: interactions between central and pe-
ripheral circadian oscillators in mammals. Cold Spring Harb Symp Quant Biol
2015;80:223–32.
11. Lamia KA, Storch K-F, Weitz CJ. Physiological significance of a peripheral tissue
circadian clock. Proc Natl Acad Sci U S A 2008;105:15172–7.
12. Qian J, Scheer FAJL. Circadian system and glucose metabolism: implications for
physiology and disease. Trends Endocrinol Metab 2016;27:282–93.
13. Yamazaki S, Yoshikawa T, Biscoe EW, et al. Ontogeny of circadian organization in
the rat. J Biol Rhythms 2009;24:55–63.
14. Weaver DR. The suprachiasmatic nucleus: a 25-year retrospective. J Biol
Rhythms 1998;13:100–12.
15. Houben T, Coomans CP, Meijer JH. Regulation of circadian and acute activity
levels by the murine suprachiasmatic nuclei. PLoS One 2014;9:e110172.
16. Deboer T, Vansteensel MJ, Détári L, et al. Sleep states alter activity of suprachi-
asmatic nucleus neurons. Nat Neurosci 2003;6:1086–90.
17. Ralph MR, Foster RG, Davis FC, et al. Transplanted suprachiasmatic nucleus de-
termines circadian period. Science 1990;247:975–8.
18. LeGates TA, Fernandez DC, Hattar S. Light as a central modulator of circadian
rhythms, sleep and affect. Nat Rev Neurosci 2014;15:443–54.
19. Mistlberger RE. Circadian regulation of sleep in mammals: role of the suprachias-
matic nucleus. Brain Res Rev 2005;49:429–54.
20. Saper CB, Scammell TE, Lu J. Hypothalamic regulation of sleep and circadian
rhythms. Nature 2005;437:1257–63.
21. Smale L, Lee T, Nunez AA. Mammalian diurnality: some facts and gaps. J Biol
Rhythms 2003;18:356–66.
22. Kramer A, Yang FC, Snodgrass P, et al. Regulation of daily locomotor activity and
sleep by hypothalamic EGF receptor signaling. Science 2001;294:2511–5.
23. Zhou QY, Cheng MY. Prokineticin 2 and circadian clock output. FEBS J 2005;272:
5703–9.
24. Welsh DK, Takahashi JS, Kay SA. Suprachiasmatic nucleus: cell autonomy and
network properties. Annu Rev Physiol 2010;72:551–77.
25. Hastings MH, Maywood ES, Brancaccio M. Generation of circadian rhythms in
the suprachiasmatic nucleus. Nat Rev Neurosci 2018;19:453–69.
26. Brown EN, Lydic R, Schiff ND. General anesthesia, sleep, and coma. N Engl J
Med 2010;363:2638–50.
27. Saper CB, Fuller PM. Wake-sleep circuitry: an overview. Curr Opin Neurobiol
2017;44:186–92.
28. Scammell TE, Arrigoni, Lipton JO. Neural circuitry of wakefulness and sleep.
Neuron 2017;93:747–65.
29. Everson CA. Functional consequences of sustained sleep deprivation in the rat.
Behav Brain Res 1995;69:43–54.
30. Available at: https://www.nytimes.com/2007/10/23/science/23quot.html. Ac-
cessed April 8, 2019.
31. Borbély AA. A two process model of sleep regulation. Hum Neurobiol 1982;1:
195–204.
32. Aeschbach D, Matthews JR, Postolache TT, et al. Dynamics of the human EEG
during prolonged wakefulness: evidence for frequency-specific circadian and
homeostatic influences. Neurosci Lett 1997;239:121–4.
486 Schwartz & Klerman

33. Cohen DA, Wang W, Wyatt JK, et al. Uncovering residual effects of chronic sleep
loss on human performance. Sci Transl Med 2010;2:14ra13.
34. Phillips AJK, Klerman EB, Butler JP. Modeling the adenosine system as a modu-
lator of cognitive performance and sleep patterns during sleep restriction and re-
covery. PLoS Comput Biol 2017;12:e1005759.
35. Carskadon MA, Dement WC. Sleep studies on a 90-minute day. Electroencepha-
logr Clin Neurophysiol 1975;39:145–55.
36. McHill AW, Hull JT, Wang W, et al. Chronic sleep curtailment, even without
extended (>16 h) wakefulness, degrades human vigilance performance. Proc
Natl Acad Sci U S A 2018;115:6070–5.
37. Wever RA. The circadian system of man: results of experiments under temporal
isolation. New York: Springer-Verlag; 1979. p. 276.
38. Duffy JF, Cain SW, Chang AM, et al. Sex difference in the near-24-hour intrinsic
period of the human circadian timing system. Proc Natl Acad Sci U S A 2011;
108:15602–8.
39. Klerman EB, Dijk DJ, Kronauer RE, et al. Simulations of light effects on the human
circadian pacemaker: implications for assessment of intrinsic period. Am J Phys-
iol 1996;270:R271–82.
40. Strogatz SH, Kronauer RE, Czeisler CA. Circadian regulation dominates homeostat-
ic control of sleep length and prior wake length in humans. Sleep 1986;9:353–64.
41. Wright KP Jr, McHill AW, Birks BR, et al. Entrainment of the human circadian clock
to the natural light-dark cycle. Curr Biol 2013;23:1554–8.
42. Burke TM, Markwald RR, McHill AW, et al. Effects of caffeine on the human circa-
dian clock in vivo and in vitro. Sci Transl Med 2015;7:305ra146.
43. Dahl K, Avery DH, Lewy AJ, et al. Dim light melatonin onset and circadian tem-
perature during a constant routine in hypersomnia winter depression. Acta Psy-
chiatr Scand 1993;88:60–6.
44. Shanahan TL, Czeisler CA. Light exposure induces equivalent phase shifts of the
endogenous circadian rhythms of circulating plasma melatonin and core body
temperature in men. J Clin Endocrinol Metab 1991;73:227–35.
45. Shanahan TL, Kronauer RE, Duffy JF, et al. Melatonin rhythms observed
throughout a three-cycle bright-light stimulus designed to reset the human circa-
dian pacemaker. J Biol Rhythms 1999;14:237–53.
46. Sateia MJ, Buysse DJ, Krystal AD, et al. Clinical practice guideline for the phar-
macologic treatment of chronic insomnia in adults: an American Academy of
Sleep Medicine clinical practice guideline. J Clin Sleep Med 2017;13:307–49.
47. Rajaratnam SM, Polymeropoulos MH, Fisher DM, et al. Melatonin agonist tasimel-
teon (VEC162) for transient insomnia after sleep-time shift: two randomised
controlled multicenter trials. Lancet 2008;373:482–91.
48. Lockley SW, Skene DJ, James K, et al. Melatonin administration can entrain the
free running circadian system of blind subjects. J Endocrinol 2000;164:R1–6.
49. Burgess HJ. Partial sleep deprivation reduces phase advances to light in hu-
mans. J Biol Rhythms 2010;25:460–8.
50. Dijk DJ, Czeisler CA. Paradoxical timing of the circadian rhythm of sleep propen-
sity serves to consolidate sleep and wakefulness in humans. Neurosci Lett 1994;
166:63–8.
51. Dijk DJ, Czeisler CA. Contribution of the circadian pacemaker and the sleep ho-
meostat to sleep propensity, sleep structure, electroencephalographic slow
waves, and sleep spindle activity in humans. J Neurosci 1995;15:3526–38.
52. Muto V, Jaspar M, Meyer C, et al. Local modulation of human brain responses by
circadian rhythmicity and sleep debt. Science 2016;353:687–90.