C134 1945-1960 PDF

134
SLEEP DISTURBANCES ASSOCIATED

WITH NEUROPSYCHIATRIC DISEASE
ERIC A. NOFZINGER
MATCHERI KESHAVAN
Several observations suggest important links between sleep data that may guide interpretations of human electrophysi-
and mental disorders. From a psychological perspective, it ology and functional neuroanatomy.
has long been thought that trouble sleeping is related to As a guide, this chapter attempts to first distill aspects
a troubled mind, an intrusion of mental activity into the of preclinical work that may inform our understanding of
quiescence of sleep. Historically, this vein has fueled much findings in electrophysiology and functional neuroanatomy
of the interest in the relationship between sleep and mental in clinical populations. We then review the subjective, EEG,
disorders until the discovery of electrophysiologically de- and brain imaging work in the major mental disorders that
fined sleep stages, particularly rapid eye movement (REM) help to define pathophysiology through a sleep window into
sleep in the 1950s. Even after that time and into the 1960s brain function. Disorders that are highlighted are the major
and 1970s, this basic psychological tenet guided interest in mood disorders, schizophrenia, and degenerative disorders
uncovering relationships between REM sleep and mental of aging given the extensive sleep research conducted in
disorders. Over time, interest shifted to defining the neuro- each of these areas.
biology of mental disorders. In this service, EEG sleep stag-
ing became a tool to be used in either diagnosis or validation
of the biological nature of mental disorders. A second, non- NEUROBIOLOGY OF HEALTHY SLEEP
psychiatric, line of investigation during this time concerned
Overview of the Sleep/Wake Cycle
itself with the physiology of normal sleep. This led to a
significant expansion in our understanding of the brain Sleep can be conceptualized as a motivated behavior, some-
mechanisms that govern basic sleep/wake, or behavioral thing the organism ‘‘needs’’ to do in order to survive and
state, regulation. Few attempts, however, have been made for which there is a pressure to perform. Sleep propensity is
to bridge these diverging lines of investigation despite the lowest shortly after awakening, increases in mid-afternoon,
complimentary information derived from each area. In part, plateaus across the evening, is greatest during the night and
this may be related to the vastly divergent levels of observa- declines across sleep. Sleep has been subclassified polysom-
tion of the brain capable in preclinical and human studies. nographically into NREM and REM sleep states based on
With the advent of advances in preclinical work defining the three measures: (1) electroencephalography (EEG), (2) elec-
brain mechanisms underlying electrophysiologic oscillations tromyography (EMG), and (3) electrooculography (EOG).
measured at the brain surface and in human brain imaging With sleep onset, the EEG frequency slows, the amplitude
work defining correlates of underlying functional neuroana- increases and the EMG decreases. This sleep is classified as
tomy, these divergent fields are beginning to communicate NREM stages 1, 2, 3, or 4, which are distinguished by
in meaningful ways to enrich the discoveries made in each increasing amounts of low-frequency, high-amplitude EEG
domain. Evidence of this evolution comes in the form of activity, also known as ‘‘delta’’ activity. Delta sleep decreases
chapters in this volume devoted to neuropsychopharmacol- across the night. REM sleep follows the first NREM period
ogy by neuroscientists such as Allan Hobson and Emmanuel and is characterized by low-amplitude, mixed high-fre-
Minot and the inclusion in this clinical chapter of preclinical quency EEG, the occurrence of intermittent REMs, skeletal
muscle atonia, and irregular cardiac and respiratory events.
Across the night, brain function oscillates between the glo-
bally distinct states of NREM and REM sleep about three
Eric A. Nofzinger and Matcheri Keshavan: Department of Psychiatry, or four times, approximately every 90 minutes. Across suc-
University of Pittsburgh School of Medicine, Pittsburgh, Pennsylvania. cessive sleep cycles within a night, stages 3 and 4 sleep de-
1946 Neuropsychopharmacology: The Fifth Generation of Progress
crease and then disappear, whereas REM sleep and lighter At the other end of the oscillatory spectrum are the fast
NREM sleep stages increase. oscillations in the beta and gamma frequencies (roughly ⬎
20 Hz) that have been associated with increased vigilance
and are most prevalent during waking and REM sleep (3,
Historical Development of Sleep
4). These fast rhythms are present in cortical and thalamic
Neurobiology
neurons and depend on neuronal depolarization characteris-
The current understanding that sleep and wakefulness are tic of brain activation, a state in which neurons are maxi-
both active brain states that are generated and maintained mally ready to respond to either external (in waking) or
from within the brain has its origins in the pioneering work internal (in REM sleep) stimuli. These fast synchronous
of Berger (1930), Economo (1929), Bremer (1935), Moru- rhythms may serve a purpose of binding aspects of a stimu-
zzi and Magoun (1949), and Jouvet (1962). Prior to the lus into a global representation. Importantly, these fast
work of these investigators, the state of sleep was thought rhythms as well as the slow rhythms are dependent on the
to represent an inactive period for the brain. Berger first level of excitability in local intracortical circuits as may be
described electrophysiologic patterns of brain electrical ac- modulated by ascending activating modulatory systems such
tivity that correlated with changes in behavioral state. Other as the brainstem monoaminergic and cholinergic nuclei (5).
investigators discovered that the brain itself was responsible Disturbances in these fast frequency oscillations in neuro-
for generating its own intrinsic electrical activity and for psychiatric disorders may reflect pathology at one or more
switching between behavioral states. Clinicoanatomic clues of these levels modulating these electrophysiologic rhythms.
came in the discovery by Economo (1929) that lesions to the
posterior hypothalamus in encephalitis lethargica produced Mechanisms Underlying Behavioral State
somnolence or coma, suggesting a role for the posterior
Changes: Core Structures Related to
hypothalamus in maintaining wakefulness. Moruzzi and
Arousal
Magoun defined the concept of a nonspecific ascending
reticular activating system (ARAS) responsible for cortical From the electrophysiologic preclinical literature, it has be-
activation following their discovery that electrical stimula- come apparent that the set point for electrical oscillations
tion of the brainstem reticular formation suppressed high- in widespread thalamocortical circuits can be modulated by
amplitude EEG waves in the cortex. Finally, following the global ascending activating influences. Identification of the
discovery of the behavioral state of REM sleep by Aserinsky structures that are important in providing this input, there-
and Kleitman (1), the French neurophysiologist Michel fore, may provide clues as to the brain mechanisms leading
Jouvet (2) localized the generation of this sleep state to struc- to altered electrophysiologic activity in diverse neuropsy-
tures in the pontobulbar brainstem using rostropontine chiatric disorders. Arousal and the maintenance of an
transections in cats. From these studies it is now recognized aroused state is an active process requiring the integrated
that behavioral state is intrinsically regulated and generated activity of a series of arousal systems shown diagrammati-
by the brain itself and manifested by electrical activity in cally in Fig. 134.1. (See Robbins and Everitt, 1996, for
the brain that can be recorded by EEG at the scalp. review and Moore and colleagues, 2001, for a discussion
An extensive preclinical literature has developed that de- of the relationship between the Orexin neurons and these
scribes the brain mechanisms of the distinct electrophysio- arousal systems) (6,7).
logic oscillations that characterize the various behavioral The central brainstem arousal system is the ascending
states of waking, NREM, and REM sleep. Several features reticular activating system (ARAS) (8,9). The ARAS
are important with respect to localizing the brain structures projects into a series of specific brainstem systems including
underlying slow wave sleep. First, the electrical oscillations the pontine cholinergic nuclei, midbrain raphe nuclei, and
observed at the macroscopic level are the end result of elec- the locus ceruleus, and into a series of forebrain structures
trical oscillations involving widespread thalamocortical involved in arousal. These include the midline and medial
neurons that are synchronized in a global fashion. Second, thalamus with widespread cortical projections and the
widespread changes in these oscillations can result from amygdala, which has interconnections with isocortex and
state-dependent changes in modulatory systems such as the other areas involved in arousal, particularly hypothalamus
brainstem, hypothalamus, and basal forebrain. Third, slow and ventral striatum. The amygdala is particularly involved
oscillations in the 1 to 4 Hz delta range have both cortical with autonomic regulation and the emotional component
and thalamic components. This line of research emphasizes of arousal. There are two important components of the basal
that rhythmic oscillations are the end result, therefore, of forebrain–ventral striatum system (10). One is the choliner-
integrated corticothalamic circuits and modulatory struc- gic neurons of the medial septum–nucleus of the diagonal
tures. Changes in delta sleep found in diverse neuropsychia- band–nucleus basalis complex that innervates the entire
tric disorders, therefore, may result from functional changes forebrain. The second is the nucleus accumbens–ventral
at one or more levels including the cortex, thalamus, and striatum complex that is involved in transmitting the arous-
modulatory structures. ing aspects of reinforcing stimuli. Until recently, the impor-
Chapter 134: Sleep Disturbances and Neuropsychiatric Disease 1947
FIGURE 134.1. Arousal systems in the

human brain (see text for description). The
transmitters associated with each system are
abbreviated. ACH, acetylcholine; DA, dopa-
mine; GABA, ␥ aminobutyric acid; GLU, gluta-
mate; HIST, histamine; HYP, hypocretin; NA,
noradrenaline; 5HT, serotonin.
tance of the hypothalamus was not fully recognized, but pontine reticular formation underlie the phasic and tonic
this is rapidly changing (10,11). First, we now appreciate components of REM sleep (9). A reciprocal interaction hy-
that an important component of the circadian control of pothesis (9) claims that these cholinergic nuclei become dis-
behavioral state is the maintenance of arousal by the circa- inhibited during the entry into REM sleep by the removal
dian pacemaker, the suprachiasmatic nucleus (12). Second, of tonic inhibition from noradrenergic and serotonergic nu-
there are extensive hypothalamic projections to isocortex, clei as these monoaminergic nuclei slow or become silent
predominantly from posterior hypothalamus. These include in the transition from NREM to REM sleep (15). Modifica-
a newly discovered projection from a group of neurons that tions of this model now account for the influence of addi-
produce a novel peptide hypocretin. This projection is of tional brainstem neurotransmitter systems such as GABAer-
particular interest because the hypocretin neurons project gic, nitroxergic, glutamatergic, glycinergic, histaminergic,
not only over the entire isocortex but also to all of the adenosinergic, and dopaminergic; various peptide systems
arousal systems noted in Fig. 134.1, including extraordinar- such as galanin, Orexin, vasoactive intestinal polypeptide,
ily dense projections to locus ceruleus, raphe nuclei, pontine and nerve growth factor; and hormonal influences such as
cholinergics, midline thalamus, nucleus basalis, and amyg- growth hormone-releasing hormone, prolactin, and cortico-
dala. (See ref. 7 for review.) The hypocretin projection has tropin-releasing factor (9,11,16–18). Brainstem reticular
become of particular interest because a hypocretin gene nuclei include a dorsal pathway innervating the thalamus
knockout produces a narcolepsy-like syndrome in mice (13) and a ventral pathway innervating the basal forebrain that
and hypocretin is below detectable levels in CSF from narco- thereby mediates widespread cortical arousal indicative of
leptics in comparison to controls (14). the REM sleep state. Human brain imaging studies of REM
sleep show that the ventral pathway predominates during
human REM sleep in activating anterior paralimbic struc-
Mechanisms Underlying Behavioral State
tures (19,20).
Changes: Core Structures Involved in
REM Sleep
Mechanisms Underlying Behavioral State
The nature of dreaming has led to extensive discussions
Changes: Modulatory Structures
about the relationships between REM sleep and mental dis-
orders (9). The finding of EEG sleep staging abnormalities In large part, these preclinical studies have focused on the
in REM sleep in certain mental disorders has also fueled primary nuclei that may be generative centers for one behav-
discussions about the relationships between the biology of ioral state or another. Less information is available regarding
REM sleep and the pathophysiology of mental disorders; structures that modulate activity in these centers and that
therefore, a brief review of brain structures that may have may play a role in the abnormal modulation of behavioral
overlapping roles in REM sleep regulation and the patho- states in various mental disorders (8,9). Recent work, for
physiology of mental disorders may guide theoretical models example, suggests that the amygdala has significant ana-
of behavioral state regulation in diverse mental disorders tomic connections and recently established modulatory ef-
(see Chapters 128 to 133). Evidence from a variety of ap- fects on the brainstem centers involved in REM sleep pro-
proaches suggests that the laterodorsal and pedunculopon- duction (21,22). Similarly, other forebrain structures such
tine tegmental cholinergic nuclei (LDT and PPT) in the as the hypothalamus (13) and basal forebrain (10,17) are
known to have both anatomic and functional relationships of limbic and paralimbic structures including the amygdala,
with brainstem centers thought to play a role in behavioral ventral striatum, anterior cingulate, and medial prefrontal
state regulation in addition to the primary roles they each cortex (19,20,23,29,30). This pattern of activation is super-
play in cortical arousal. Although it is possible that certain imposed on brainstem activation known to play a role in
disease states are associated with pathologic changes in dis- REM sleep generation on the basis of preclinical work.
crete brain structures that generate discrete behavioral states,
it is more likely (e.g., in the case of depression) that there
are pathologic functions in brain structures that have modu- Functional Brain Changes Associated with Sleep
lating effects on these core brain structures in producing Deprivation
the behavioral state changes. Several studies have discovered that sleep deprivation is asso-
ciated with a global reduction in metabolism with some
Mechanisms Underlying Behavioral State preference for the prefrontal cortex. This reduction is mag-
Changes: Evidence from Human Sleep nified with successive nights of sleep deprivation. These
Imaging Studies changes have recently been shown to play a role in the
cognitive alterations associated with sleep deprivation (31).
The advent of brain imaging methods that can be used
to study the functional neuroanatomy of human sleep has
recently created a venue for linking preclinical work with
human sleep electrophysiology. Findings from these new THEORETICAL MODELS OF THE FUNCTION
studies have enriched our understanding of the brain struc- OF SLEEP
tures that are preferentially active across behavioral states Homeostatic Function
and that may play a role in behavioral state regulation as
well as functional roles for discrete behavioral states. Several models of sleep/wake regulation attempt to define
parameters that may influence the probability at any point
Global Changes in Brain Function across in the day that sleep may occur. One such model is the
Behavioral States two-process model described by Borbely (32). One process,
called process S, describes a homeostatic sleep process. Pro-
In terms of global, or whole brain changes, from waking to cess S is thought to be dependent on the amount of prior
NREM sleep, there are reductions in measures of cortical wakefulness and is reflected by the amount of EEG slow
blood flow or metabolism (19). During REM sleep, global wave (.5 to 4.5 Hz EEG) activity. As sleep deprivation in-
blood flow or metabolism ranges from 10% below to 41% creases, for example, process S increases and amplifies sleep
above levels obtained during wakefulness (23,24). A recent propensity. The second process, process C, varies through-
report cited a positive correlation between waking global out the day in relation to a sinusoidal circadian phase across
and regional cerebral blood flow and slow wave sleep mea- a 24-hour day. The intensity of this sleep propensity is unre-
sures from the subsequent night of sleep (25). The authors lated to the amount of prior wakefulness. The sleep parame-
interpreted these findings as reflecting an energy conserva- ter most affected by this process is process C; nonsleep corre-
tion or restorative role for slow wave sleep. lates of this process include core body temperature, plasma
melatonin, and plasma cortisol levels. The regulatory struc-
Relative Regional Changes during NREM Sleep ture for this process is the suprachiasmatic nucleus.
In terms of regional relative changes during NREM sleep,
blood flow has been shown to negatively correlate with the Sleep and Neuronal Plasticity
presence of NREM sleep in the anterior cingulate (19,26,
27), pontine reticular formation (19,26,27), thalamus (19, Several theorists have conceptualized distinct roles for the
26–28), basal forebrain/hypothalamus (19,26), amygdala diverse sleep/wake states based on emerging knowledge re-
(26), and orbitofrontal cortex (19,26,27). These changes garding forebrain function in information processing. Buz-
are consistent with preclinical studies showing reductions saki (33) emphasized a two-stage model of waking hippo-
in brainstem, basal forebrain, and hypothalamus sources of campal memory trace formation and suggested parallels for
ascending activation. Declining function in the amygdala two stages of memory processing during NREM and REM
suggests the possibility that this structure modulates activity sleep. Karni and associates and Wilson and McNaughton
in ascending activating structures. have demonstrated direct evidence in support of memory
processing during sleep for REM and NREM sleep, respec-
tively.
Relative Regional Changes during REM Sleep
Although the general states of neuronal activation have
In terms of regional relative changes during REM sleep, this been described for NREM and REM sleep, little is known
state has been reliably associated with the selective activation within these states about the relevance of this neuronal acti-
vation. Llinas and Pare (34) suggest that wakefulness and asleep at night. Subjective sleep quality has been measured
REM sleep are fundamentally equivalent states of activation with a validated instrument called the Pittsburgh Sleep
characterized by intrinsic oscillatory thalamo-cortical loops Quality Index (PSQI) (37). In one study, subjective sleep
differing only in the degree to which external stimuli are quality was rated worse by patients with major depression
capable of modulating the global brain state. They suggest than by patients presenting with a chief complaint of insom-
that these oscillations serve the purpose of generating an nia or other sleep disorder patients. Whereas insomnia char-
internal representation of the world guided by innate predis- acterizes the melancholia of middle age and elderly unipolar
positions of the brain to categorize and integrate the sensory depression, younger patients and bipolar depressed patients
world in certain ways. During REM sleep, these innate tem- often describe, atypically, difficulty getting up in the morn-
plates, which have been molded by experience, may be used ing and hypersomnia during the daytime. This subjective
to recreate world-analogues. During wakefulness, these tem- hypersomnia, however, does not translate into an increased
plates are modulated by sensory events. Winson (35) sug- physiologic tendency to fall asleep when measured objec-
gested that, in humans, psychological experience essential tively by the Multiple Sleep Latency Test (MSLT).
to survival is integrated and further consolidated during
REM sleep. He based this hypothesis on: (a) the presence
EEG Findings
of hippocampal theta during REM sleep; (b) the sole occur-
rence of hippocampal theta during species-specific survival- An extensive literature describes the changes in EEG sleep
dependent behavior in nonprimates; and (c) the role of hip- in patients with depression (38,39). Measures derived from
pocampal theta in the induction of long-term potentiation the EEG sleep recordings that have been found to differ
(LTP), a model for the synaptic modulation underlying cer- between healthy and depressed subjects include measures
tain types of memory. Kavanau (36), reviewing data on of sleep continuity, measures of visually scored EEG sleep
synaptic events related to the formation of enhanced synap- stages, and automated measures of characteristics of the
tic efficacy in the context of evolutionary biology, suggests EEG waveform across the sleep period such as period ampli-
that REM sleep may perform the service of repetitive activa- tude or EEG spectral power measures.
tions of synapses in neural circuits that underlie essential The changes in subjective sleep complaints are paralleled
adaptive behaviors. This ‘‘dynamic stabilization’’ is thought by EEG measures of sleep. These include increases in sleep
to ensure the efficacy of circuits that otherwise may suffer latency and decreases in sleep continuity. In terms of EEG
from a disuse atrophy. Although clearly these hypotheses sleep stages or ‘‘sleep architecture,’’ depressed patients often
require validation, the underlying premises are consistent show reduced stage 3 and 4 NREM sleep (also known as
with psychological theories that REM sleep and dreaming ‘‘slow wave sleep’’ because of the presence of slow EEG
may play a role in affective adaptation; in the integration delta activity during these stages). Several changes in REM
of recent, remote, and perhaps phyletic memory; maintain- sleep also have been noted. These include an increase in the
ing or facilitating behavioral repertoires underlying crucial amount of REM sleep, shortening of the time to onset of
behavior; and the genetic programming of inherited be- the first REM period of the night, shortened REM latency,
havior. and increase in the frequency of eye movements within a
REM period.
In terms of quantitative EEG changes in sleep, many
(40,41) but not all studies have reported reductions in the
SLEEP IN MAJOR NEUROPSYCHIATRIC
amplitude or a reduction in the number of low frequency
DISEASES
(0 to 4 Hz) delta waves during sleep in depressed patients.
Increased high frequency EEG activity has also been re-
We turn now to characterizing the sleep disturbances in the
ported in depressed patients, including alpha (40) and beta.
major mental disorders. The vast majority of relevant data
Importantly, sex differences have been found in these abnor-
come from clinical and EEG sleep reports. Early findings
malities. Depressed women appear to have relative preserva-
from functional brain imaging studies are also reviewed.
tion of delta sleep in relation to depressed men, despite
elevations in higher frequency EEG activity in both groups
(42).
DEPRESSION Several observations regarding sleep disruption in de-
pressed patients suggest that there may be a timing abnor-
Subjective Findings
mality in the evolution of sleep across the night in depressed
The majority of patients with mood disorders describe diffi- patients (38,41). A number of factors can influence the sleep
culty falling asleep, staying asleep, and returning to sleep EEG findings characteristic of depression. For instance,
after early morning awakenings. Clinically, they report a sleep continuity deteriorates, slow wave sleep decreases, and
paradoxic state of physical daytime fatigue, yet with persis- REM latency shortens with age, even in healthy subjects;
tent mental activity that makes it difficult for them to fall however, age-related changes are more pronounced in pa-
tients with depression (43). By contrast, sleep EEG mea- sleep would be shifted toward states of cortical activation,
sures are generally less abnormal in adolescents and pre- that is, either waking or REM sleep, and away from NREM
pubertal children with depression, and only appear sleep, which has been associated with the homeostatic func-
consistently in those adolescents who are hospitalized and/ tion of sleep.
or suicidal (44).
Other studies have shown that patients with psychotic
Sleep Neuropharmacologic Findings
depression have particularly severe EEG sleep disturbances
and very short REM sleep latencies; patients with recurrent Each of the major neurotransmitter systems shown to mod-
depression have more severe REM sleep disturbances than ulate the ascending activation of the cortex, that is, the
patients in their first episode; and sleep continuity and REM cholinergic, noradrenergic, and serotonergic systems, have
sleep disturbances are more prominent early in the depres- been implicated in the pathophysiology of mood disorders.
sive episode than later (45). Some studies suggest that pa- The role of additional brainstem neurotransmitter systems
tients with dysthymia and mania (surprisingly) have EEG such as GABAergic, nitroxergic, glutamatergic, glycinergic,
sleep disturbances very similar to those observed in major histaminergic, adenosinergic, dopaminergic, and various
depression. peptide systems such as galanin, Orexin, vasoactive intes-
Stressful life events also interact with EEG sleep. For tinal polypeptide, and nerve growth factor in the sleep dis-
instance, individuals who have severe stressful events pre- turbances in depression remain to be defined. Nearly all
ceding the onset of depression are less likely to have reduced effective antidepressant medications show a pronounced in-
REM latency than patients without such a stressor. Among hibition of REM sleep including a prolongation of the first
older depressed patients, poor sleep is associated with REM cycle and a reduction in the overall percent of REM
shorter episode duration, older age, greater medical burden, sleep. (Exceptions include nefazodone [48] and bupropion,
the presence of life stressors, and a lower level of perceived which do not suppress REM sleep [49].) Enhanced cholin-
social support (45). ergic function concurrent with reduced monoaminergic
EEG sleep findings help to inform our understanding tone in the central nervous system has been proposed as a
of the neurobiology of longitudinal course and treatment pharmacologic model for depression. In an exaggerated
outcome in depression. Although severely reduced REM sense, the state of REM sleep mimics this formulation, that
latencies, phasic REM measures, and sleep continuity dis- is, a cholinergically driven state with reduced firing of nora-
turbances generally move toward control values after remis- drenergic and serotonergic neurons. Cholinergic agents such
sion of depression, most sleep measures show high correla- as the muscarinic agonist RS 86, arecoline, physostigmine,
tions across the course of an episode. Reduced REM latency and scopolamine produce exaggerated REM sleep effects in
is associated with increased response rates to pharmacother- depressed patients in comparison with patients with eating
apy (46) but not psychotherapy. Depressed patients with disorders, personality disorders, anxiety disorders, and
abnormal sleep profiles (reduced REM latency, increased healthy controls (50). These studies suggest that there may
REM density, and poor sleep continuity) are significantly be a supersensitivity of the cholinergic system driving REM
less likely to respond to cognitive behavior therapy and in- sleep in mood disorders patients, although an alternative
terpersonal therapy than patients with a ‘‘normal’’ profile. plausible hypothesis is that there may be reduced monoami-
Other studies have indicated that reduced REM latency and nergic (5-HT and/or NE) inhibition of the brainstem cho-
decreased delta EEG activity are associated with increased linergic nuclei in mood disorders patients. Cholinergic acti-
likelihood or decreased time until recurrence of depression vation may also play a role in the hyperactivity in the HPA
in patients treated with medications or psychotherapy (47). axis and in the blunting of growth hormone secretion noted
in depressed patients across the night, given the influence
of cholinergic drugs on HPA activity and GH release.
Sleep Neuroendocrine Findings
Selective serotonin reuptake inhibitors are known to have
Cortisol secretion has been linked with the circadian cycle prominent REM suppressing activity, most notably early in
and growth hormone with slow wave sleep processes. In the night when enhances in REM sleep are most often seen
depressed subjects, studies have shown increases in cortisol in mood disorders patients (39). A tryptophan-free diet,
secretion rates, a flattening of the circadian rhythms in corti- which depletes central serotonin activity, is noted to de-
sol, and elevated cortisol nadir. Secretion of growth hor- crease REM latency in healthy controls and in depressed
mone on the other hand, is reduced in the first half of the patients (51) and ipsapirone, a 5-HT1a agonist, is noted
night in depressed subjects both in the acute phase and to prolong REM latency in both normal controls and in
following remission from depression. Evidence such as this depressed patients (52). Anatomically, 5-HT1a receptors
has been used to support hypotheses that a balance between have been conceptualized as the limbic receptors given their
the oppositional actions of CRF and GHRH may be shifted high densities in the hippocampus, septum, amygdala, and
toward CRF in major depression. Given the activating qual- cortical paralimbic structures. The action in these structures
ities of CRF and its direct or indirect inhibition of GHRH, has been shown to be largely inhibitory (hyperpolarizing).
Given the importance of limbic and paralimbic structures gray) during REM sleep. Finally, in contrast to controls,
in REM sleep modulation, the influence of SSRI medica- depressed subjects activate left sensorimotor cortex, left infe-
tions may be mediated by these limbic receptors. Impor- rior temporal cortex, left uncal gyrus and amygdala, and
tantly, in the brainstem LDT, a locus of cholinergic cells left subicular complex during REM sleep. These findings
identified in the generation of REM sleep, bursting cholin- suggest that depressed patients demonstrate uniquely differ-
ergic neurons are inhibited by the action of 5-HT on 5- ent patterns of activation from waking to REM sleep than
HT1a receptors. Finally, the effects of the 5-HT1a-antago- do healthy controls. In the context of neuroscience models
nist pindolol on EEG sleep in healthy subjects was studied relating forebrain function during REM sleep to attention,
and noted to reduce REM sleep. This was interpreted as motivation, emotion, and memory, these results suggest that
supportive of a reduction in raphe serotonergic autoregula- prior REM sleep abnormalities in mood disorders patients,
tion, resulting in increased serotonergic input to pontine therefore, likely reflect alterations in limbic and paralimbic
cholinergic centers and inhibiting REM sleep. forebrain function related to depression (Fig. 134.2).
Functional neuroimaging of NREM sleep in depressed
subjects would be expected to provide evidence regarding
Functional Neuroimaging Findings the functioning of homeostatic mechanisms in mood disor-
Given the selective activation of limbic and paralimbic ders patients since this is a time of nonselective nonactiva-
structures during REM sleep in healthy subjects, the study tion of the cortex in which the buildup of a sleep dependent
of the functional neuroanatomy during REM sleep in de- process, process S, is discharged and during which growth
pressed patients may provide clues as to alterations in limbic hormone secretion occurs. Ho and Gillin and colleagues
and paralimbic function related to the pathophysiology of (53) demonstrated that whole brain and regional cerebral
depression. In contrast to healthy controls (4), depressed glucose metabolism was elevated during the first NREM
patients fail to activate anterior paralimbic structures (sub- period of the night for depressed men in relation to healthy
genual and pregenual anterior cingulate and medial prefron- men. These findings are supportive of a deficiency of ho-
tal cortices) from waking to REM sleep. In contrast to meostatic mechanisms in mood disorders patients, perhaps
healthy controls, depressed subjects show large activations secondary to cortical hyperarousal. Clark and associates (25)
in the dorsal tectum (superior colliculus and periaqueductal reported that reductions in delta sleep in depressed patients
FIGURE 134.2. Healthy and depressed subjects’ changes in relative glucose metabolism from
waking to rapid eye movement (REM) sleep. The top two images are midsagittal sections showing
areas of the brain that have greater relative glucose metabolism during REM sleep than during
waking. These areas include anteriorly located paralimbic structures such as the ventral striatum,
anterior cingulate cortex, and medial prefrontal cortex. The lower two images show the same
comparisons, but in depressed patients. Of note is the absence of an increase in relative metabolism
in any of these anteriorly located paralimbic structures in depressed patients from waking to REM
sleep. See color version of figure.
were associated with reductions in afternoon waking relative Depression and Alcoholism
and global blood flow. This suggests that the elevations
Significant comorbidity exists between depression and alco-
in glucose metabolism during NREM sleep in depressed
holism. EEG sleep studies have been used as a tool to explore
patients are not related to a waking hypermetabolic state.
the relationships between the neurobiology of the two disor-
Studies across waking and NREM sleep are needed in order ders. Clark and colleagues (56) found that alcoholism played
to clarify this notion. a more significant role in sleep disruption than depression.
Nofzinger and associates (54) sought to clarify the neuro- Increased REM density in primary alcoholics with and with-
biological basis of variations in one aspect of central nervous out a lifetime diagnosis of secondary depression predicted
system ‘‘arousal’’ in depression by characterizing the func- 3-month alcohol-related relapse rates. Drummond and co-
tional neuroanatomic correlates of beta EEG power density workers (57) examined the relationships among sleep, natu-
during NREM sleep. First, nine healthy (N ⳱ 9) subjects ral course, and relapse in abstinent pure primary alcoholic
underwent concurrent EEG sleep studies and [18F]2-fluoro- patients. They found that the sleep was short, fragmented,
2-deoxy-D-glucose ([18F]FDG) positron emission tomogra- and shallow early in abstinence with some incomplete im-
phy (PET) scans during their first NREM period of sleep provements over the first year of abstinence. Further, the
in order to generate hypotheses about specific brain struc- presence of sleep disruption at 5 months was shown to pre-
tures that show a relationship between increased beta power dict relapse by 14 months.
and increased relative glucose metabolism. Second, brain
structures identified in the healthy subjects were then used
as a priori regions of interest in similar analyses from identi- SCHIZOPHRENIA
cal studies in 12 depressed subjects. Statistical parametric
mapping was used to identify the relationship between beta REM Sleep
power and relative regional cerebral glucose metabolism Early sleep EEG studies sought to test the intriguing hy-
(rCMRglu) during NREM sleep. Regions that demon- pothesis that schizophrenia is a spillover of the dream state
strated significant correlations between beta power and rela- into wakefulness. No evidence has accrued to support this
tive cerebral glucose metabolism in both the healthy and prediction; however, subtle alterations in architecture of
depressed subjects included the ventromedial prefrontal cor- REM sleep may occur. REM latency was found decreased
tex and the right lateral inferior occipital cortex. During a in seven of the 10 studies (58). It has been proposed that
baseline night of sleep, depressed patients demonstrated a this may result from a deficit in SWS in the first NREM
trend toward greater beta power in relation to a separate period leading to a passive advance, or early onset of the first
age- and gender-matched healthy control group. In both REM period. An alternative explanation is ‘‘REM pressure.’’
healthy and depressed subjects, beta power negatively corre- However, studies of the amounts of REM sleep have been
lated with subjective sleep quality. Finally, in the depressed conflicting with increases, decreases, as well as no change
group, there was a trend for beta power to correlate with being found (58,59). Studies examining treatment naive
an indirect measure of absolute whole brain metabolism schizophrenia patients show no increases in REM sleep (60,
during NREM sleep. This study demonstrated a similar 61); the increases in REM sleep observed in previously
treated subjects may reflect effects of medication withdrawal
relationship between electrophysiologic arousal and glucose
and/or changes related to the acute psychotic state (61). It
metabolism in the ventromedial prefrontal cortex in de-
is unlikely that the observed decreases in REM latency in
pressed and healthy subjects. Given the increased electro-
some schizophrenia patients result from primary abnormali-
physiologic arousal in some depressed patients and the ties in REM sleep.
known anatomic relations between the ventromedial pre-
frontal cortex and brain activating structures, this study
raises the possibility that the ventromedial prefrontal cortex Slow Wave Sleep
plays a significant role in mediating one aspect of dysfunc- Slow wave sleep is of particular interest to schizophrenia
tional arousal found in more severely aroused depressed pa- because of the implication of the prefrontal cortex in this
tients. disorder (62) and in generation of SWS (63). Several studies
Wu and associates (55) characterized the functional neu- have shown a reduction of SWS in schizophrenic patients;
roanatomic changes following sleep deprivation therapy in SWS deficits have been seen in acute, chronic, as well as
depressed patients. They found that depressed patients who remitted states; and in never-medicated, neuroleptic-
demonstrated high pretreatment relative glucose metabolic treated, as well as unmedicated patients (58); however, not
rates in the medial prefrontal cortex were more likely to all studies show these deficits. Studies that fail to find differ-
respond to sleep deprivation. Further, a reduction in relative ences in SWS have generally used conventional visual scor-
metabolism in this region was found following sleep depri- ing. Three studies that have quantified sleep EEG param-
vation. eters have consistently shown reductions in SWS. Ganguli
and associates (60) observed no change in visually scored sleep abnormalities in schizophrenia, in order to elucidate
SWS, but instead, a significant reduction in delta wave their significance for pathophysiology. Stage 4 does not ap-
counts in drug-naive schizophrenic patients, suggesting that pear to improve while other sleep stages change following
visually scored SWS may not be sensitive enough, and auto- 3 to 4 weeks of conventional antipsychotic treatment (69).
mated counts may be a better marker of SWS deficiency in In a longitudinal study, alterations of SWS appeared to be
schizophrenia. Other groups have described similar reduc- stable when polysomnographic studies were repeated at 1
tions in delta counts. SWS deficits have been recently dem- year, but the REM sleep parameters appeared to change.
onstrated in early course schizophrenia using sensitive ap- These observations suggest that SWS deficits in schizophre-
proaches such as spectral analysis (64). nia might be trait-related (70). Consistent with this view,
delta sleep abnormalities have been found to correlate with
negative symptoms (60) and with impaired outcome at 1
Sleep Deprivation Studies
and at 2 years (71).
Sleep deprivation provides a naturalistic, physiologic chal- An association between attentional impairment and SWS
lenge for dynamic manipulation of sleep processes, and can deficits also has been reported in early studies (72). The
help clarify the primary nature of sleep abnormalities. An thalamus plays a crucial role in attention and gating of infor-
intriguing reduction in REM rebound following REM sleep mation because it is the major relay station receiving input
deprivation has been described in several studies in acute from the reticular activating system and limbic and cortical
schizophrenia, but a normal or exaggerated REM rebound association areas. A defect in this structure could explain
in remitted schizophrenic patients (58). This rebound fail- much of the psychopathology of schizophrenia and altera-
ure in acute schizophrenia has been attributed to a possible tions in SWS.
‘‘leakage’’ of phasic REM events from REM sleep into
NREM sleep, although no systematic investigation has sup-
Relation Between Sleep Findings and
ported this hypothesis.
Neurobiology in Schizophrenia
Sleep deprivation studies can also help clarify whether
SWS abnormalities in schizophrenia are secondary to pa- Some clues to the possible pathophysiologic significance of
thology in neuronal circuits in this disorder, or whether SWS deficits in schizophrenia may derive from the studies
they reflect primary homeostatic disruption in sleep pro- of the ontogeny of sleep during normal adolescence, in the
cesses. There is evidence (65) that following total sleep dep- context of a neurodevelopmental framework for schizophre-
rivation, recovery of Stage 4 sleep is diminished in schizo- nia. Converging evidence suggests a substantial reorganiza-
phrenia. SWS deprivation is known to consistently cause tion of human brain function during adolescence; a marked
impaired attention, prolonged reaction time, verbal learn- decline in synaptic density is seen in the postmortem pre-
ing, and vigilance similar to what is seen in frontal lobe frontal cortex during adolescence. Pronounced reductions
dysfunction and schizophrenia (66). A defect in SWS recov- in cortical gray matter volume and regional cerebral metabo-
ery might be consistent with impairments in critical cogni- lism also have been seen during adolescence. Polysomno-
tive processes such as psychomotor vigilance observed in graphic studies show major changes in EEG sleep patterns
schizophrenia. Such a defect might also suggest impairment in adolescence; the amounts of SWS decrease (73) across
in trophotropic or restorative processes in schizophrenia. the age span from childhood to late adolescence. The time
courses for maturational changes in SWS, cortical metabolic
rate, and synaptic density during the postnatal phase of
Relation between Sleep Abnormalities
human development are strikingly similar; Feinberg (74)
and Clinical Measures in Schizophrenia
suggested that cortical neurons, after birth, go through a
The neurobiological correlates of the psychopathological di- phase of initial overproduction and subsequent regression
mensions are critical for our understanding of the patho- (pruning) of neural elements. Thus, the maturational pro-
physiology of psychiatric disorders. Research during the past cesses in sleep EEG, cortical synaptic density and regional
decade has focused increasingly on the positive and negative cerebral metabolism might reflect a common underlying
syndromes, a conceptual distinction of particular impor- biological change (i.e., a large-scale synaptic elimination).
tance to pathophysiology of schizophrenia. A number of It is instructive to examine the polysomnographic abnor-
studies have examined the association between REM sleep malities in schizophrenia in relation to the brain matura-
parameters and clinical parameters. Tandon and associates tional parameters discussed in the preceding. In addition to
(61) reported an inverse association between REM latency SWS deficits, consistent alterations in structure and func-
and negative symptoms. No association has been seen be- tion of cortical and subcortical brain regions have been ob-
tween sleep abnormalities and depressive symptoms (61), served in schizophrenia. Cross sectional studies of the corre-
but two studies have shown that increased REM sleep may lations between such alterations and sleep are valuable to
correlate with suicidal behavior in schizophrenia (67,68). better understand the pathophysiologic substrate of schizo-
It is important to understand the longitudinal nature of phrenia.
Altered Neuroanatomy Altered Physiology

Reductions in cortical gray matter are seen in schizophrenia, Some evidence, albeit modest, for decreased frontal lobe
perhaps more prominently in the frontal and temporal cor- metabolism has been documented in schizophrenia using a
tex as well as in thalamic volume (75). SWS generation variety of techniques, including PET, single photon emis-
appears to be regulated by a complex neural system involv- sion computed tomography (SPECT), 31P MRS, and
ing the anterior brain regions and the thalamus. The rela- xenon 133 inhalation technique. It may be instructive to
tionship between alterations in these brain structures and examine SWS deficits in the context of such physiologic
SWS, therefore, is interesting. SWS tends to be inversely alterations. An association has been demonstrated between
correlated with anterior horn ratio, a measure of frontal lobe SWS deficits and reduced frontal lobe membrane phospho-
size (76), and positively correlated with lateral ventricular lipid metabolism as examined by 31P MRS (85).
volume (77). These observations may reflect reductions in Decreased synaptic density, postulated to underlie schiz-
subcortical structures such as the thalamus, which forms a ophrenia, could result in reduced SWS by decreased mem-
substantial part of the ventricular boundaries. A correlation brane surface (fewer dendrites per neuron), causing a smaller
between thalamic volume and SWS deficits may be pre- voltage response to the synchronizing stimulus. Recent stud-
dicted, and is consistent with the former’s critical role in ies in cats using single cell recordings (86) have shown that
generation of SWS. slower (⬍1 Hz) synchronized oscillations originate mainly
in the neocortex, whereas delta waves (1 to 4 Hz) arise
primarily from activity of thalamocortical neurons. A finer
Altered Neurochemistry analysis of these oscillations may clarify the nature of patho-
What are the neurochemical mechanisms involved in SWS physiology in schizophrenia. Preliminary analysis of this
deficits? SWS may result from several neurochemical pro- question using period amplitude analyses suggested more
cesses that produce neural inhibition, functional differentia- prominent deficits in the ⬍1 Hz range in schizophrenia,
tion, and EEG synchrony. Activation of the cholinergic sys- pointing to a thalamocortical dysfunction (64). This finding
tem facilitates arousal and hastens REM sleep. Cholinergic deserves further study and replication.
hyperfunction postulated to underlie schizophrenia, there-
fore, could account for SWS and REM latency reduction
in schizophrenia sleep (78). Interestingly, schizophrenic pa- Effects of Antipsychotic Drugs on Sleep
tients show supersensitive REM sleep induction with the
cholinergic agonist R5 86 (79), suggesting cholinergic hy- Studies of the acute effects of neuroleptics have consistently
perfunction. Nicotinic receptors may also be involved; sen- shown improvements in sleep continuity, as reflected by
sory gating deficits as evidenced by P50 event-related poten- reduced sleep latencies, improved sleep time, greater sleep
tials, which are possibly related to central nicotinergic efficiency, and prolongation of REM latency (58); however,
system alterations, are reversed following sleep in schizo- changes in SWS have been less consistent. Studies that have
phrenia (80). Alternatively, serotonergic abnormalities may examined the sedative effect of conventional neuroleptics
also be involved, as indicated by an inverse correlation be- have reported either no effects or modest increases in SWS.
tween serotonin metabolites in the CSF and SWS in schizo- A recent study showed a robust increase in SWS with olan-
phrenia (81). Disturbances in catecholaminergic mecha- zapine following acute administration (87). On the other
nisms may also underlie SWS deficits in schizophrenia. hand clozapine increases stage 2 sleep, but may actually
Norepinephrine, which is presumed to be hyperfunctional decrease stage 4 sleep (88). These studies have frequently
in schizophrenia, is inhibitory to REM; therefore, it may be used small sample sizes; few studies have examined sleep
argued that cholinergic and monoaminergic abnormalities variables in relation to acute versus long-term treatment
mediate the constellation of reduced REM latency and SWS with neuroleptics in a longitudinal design.
deficit without increases in REM sleep amounts in schizo- Attempts to examine polysomnographic characteristics
phrenia (78). of schizophrenia have to consider potential effects of neuro-
The possible relation between hormonal substances and leptic discontinuation on sleep EEG. Neylan and associates
delta sleep has also received some attention. Van Kammen (89) reported significant worsening of REM and non-REM
and associates (82) showed an association between reduc- sleep in a series of schizophrenic patients undergoing con-
tion in delta sleep induction peptide, a putative endogenous trolled neuroleptic discontinuation. Patients experiencing
sleep modulator, and SWS decrements. Adenosine, an relapse has larger impairments in sleep. The effects of neuro-
amino acid neuromodulator has drawn increasing interest leptic discontinuation continued to worsen from 2 to 4
in recent years as a possible endogenous sleep-promoting weeks of a neuroleptic-free condition, and did not correlate
agent, as it tends to accumulate during waking hours (83). with clinical change (90). These findings highlight the im-
Adenosine agonists have been proposed as possible thera- portance of controlling for medication state in investigation
peutic agents in schizophrenia (84). of EEG sleep in schizophrenia.
Future Directions significant losses or in acute depressive episodes. In a study

of elderly volunteers who had lost a spouse, Reynolds and
New knowledge on brain mechanisms of sleep is likely to
colleagues were able to distinguish EEG sleep changes dis-
open new avenues to investigate the pathophysiology of
criminating subjects who did from those who did not de-
schizophrenia. First, functional brain imaging studies sug-
velop an episode of major depression. Bereaved subjects
gest distinct patterns of regional brain activation in SWS
with depression had significantly lower sleep efficiency,
and REM sleep; such studies can provide clues to the patho-
more early morning awakening, shorter REM latency,
physiology of schizophrenia, especially used in conjunction
greater REM percent, and lower rates of delta-wave genera-
with physiologic perturbation paradigms such as sleep dep-
tion in the first NREM period, as compared to nonde-
rivation (31). Second, sleep architecture changes dramati-
pressed bereaved volunteers. These findings are similar to
cally during development; sleep studies during development
those of elderly patients with recurrent unipolar depression.
in health and disease can shed considerable light on develop-
In a subsequent longitudinal study of bereaved elders who
mentally mediated neuropsychiatric disorders (91). Finally,
did not become clinically depressed, only increases in phasic
sleep changes are often the earliest signs of disturbance, and
REM activity and density (compared to normal controls)
may even represent trait-related vulnerability markers for
were observed throughout the first 2 years of bereavement
psychiatric disorders; sleep studies of individuals at risk for
(95). These findings are similar to those of Cartwright for
schizophrenia are likely to be fruitful (92).
depressed versus nondepressed divorcing women, suggest-
ing that short REM latency and slow wave sleep are corre-
lates of depression during stressful life events, whereas in-
GERIATRIC DISORDERS creased REM activity may correlate with successful recovery
from stress in the absence of major depression. Given the
Late Life Depression
selective activation of limbic and paralimbic structures dur-
Assessments of the timing and quality of NREM and REM ing REM sleep, that is, structures related to affective adapta-
sleep cycles through EEG sleep studies have proven to be tion, the increase in REM activity in the successful resolu-
particularly useful indicators of homeostatic and adaptive tion of stressful events may relate to the activation of brain
physiologic processes during successful and pathologic aging structures mediating adaptation to these stressful events. A
in humans. In a study of healthy ‘‘old old’’ and ‘‘young failure to terminate this normal adaptive response, similar
old,’’ Reynolds and colleagues (93) reported: (a) a small age- to the failure of counterregulatory neuroendocrine events
dependent decrease in slow-wave sleep, in contrast to the in the generalized stress response, appears to result in a shift
stability of REM sleep measures from ‘‘young old’’ to ‘‘old of central neurophysiologic events favoring cortical activa-
old’’; (b) much better preservation of slow wave sleep among tion. This results in shifts away from homeostatic NREM
aging women than men, particularly in the first NREM sleep states, and toward more activated waking and REM
period of the night, but no sex-related differences in REM sleep cortical states. In the EEG these are manifested by
sleep measures; (c) greater stability of sleep maintenance reductions in intrinsic, nonactivated thalamocortical syn-
among aging men than women; and (d) longer REM sleep chronous EEG waves of NREM sleep, and an early entry
latencies among aging women than men. In comparison into desynchronized REM sleep. Subjectively, this may re-
with 20 year olds, 80 year olds have significant reductions late to depressed patients complaints of both increased fa-
in both REM sleep percent and latency, as well as a slower tigue, or anergy (loss of homeostatic mechanisms) and in-
recovery from the effects of acute sleep loss. Elderly people creased internal ‘‘arousal’’ (shifting toward a more
show greater rigidity in sleep patterns, with less intersubject continuous on-line state of readiness to process salient
and intrasubject variability in habitual sleep times compared stimuli).
to the young (94); therefore, although some loss of REM
and slow wave NREM sleep characterize the aging process,
Alzheimer’s Disease
successful aging is associated with a relative stability of sleep
states over the later years. One explanation for these findings Disturbances in sleep commonly accompany Alzheimer’s
may be some mild losses in both homeostatic as well as disease (AD) (96). These disturbances are a significant cause
adaptive physiologic mechanisms that result in NREM and of distress for caregivers, often leading to institutionalization
REM sleep decrements, respectively. Alternatively, informa- of these patients (97). The changes in sleep often parallel
tion processing related to both homeostatic and adaptive the changes in cognitive function in demented patients (98).
behavior may be more stable and efficient in the elderly, Also, daytime agitation has been associated with sleep qual-
especially in those for whom the aging process has been ity at night (99). A large-scale community based study of
bridged successfully. AD patients reported that sleeping more than usual and
EEG sleep studies have also provided insights into the early morning awakenings were the most common sleep
pathophysiology of disorders in which affective adaptation disturbances in noninstitutionalized patients. Nighttime
is significantly stressed in late-life, that is, in response to awakenings, however, were more disturbing to caregivers.
Nighttime awakenings were associated with male gender, sleep spindles (107,108). Reductions in REM latency have
and greater memory and functional declines. Three groups been observed. Increased muscular activity, contractions
of subjects were identified in association with nocturnal and periodic limb movements may prevent slow wave sleep
awakenings: (a) patients with only daytime inactivity; (b) and foster light fragmented sleep. Disorganized respiration
patients with fearfulness, fidgeting, and occasional sadness; also is found (109).
and (c) patients with multiple behavioral problems includ-
ing frequent episodes of sadness, fearfulness, inactivity, fid-
geting, and hallucinations. FUTURE DIRECTIONS
In terms of sleep laboratory-based evaluations, sleep con-
tinuity disturbances in these patients include decreased sleep As one of the earlier tools available to psychiatric research
efficiencies, increased lighter stage 1 NREM sleep, and an for discovering the biological basis of mental disorders, EEG
increased frequency of arousals and awakenings. Sleep archi- sleep recordings have been used extensively to characterize
tecture abnormalities include decreases in stages 3 and 4 alterations in brain function across diverse mental disorders.
NREM sleep and some reports of decreases in REM sleep This literature now forms the background for new research
(100). Loss of sleep spindling and K complexes have also that can uncover the brain mechanisms that underlie these
been noted in dementia. Sleep apnea has been observed in descriptive changes. Tools available for such purposes in-
33% to 53% of patients with probable AD (101). It is clude refinements in electrophysiologic recordings using au-
unclear if there is an increased prevalence of sleep apnea, tomated EEG and the concurrent use of electrical recordings
however, in AD patients in relation to age- and gender- of cognitive processes such as evoked responses to character-
matched controls. Nocturnal behavioral disruptions, or ize changes in information processing during sleep in rela-
‘‘sundowning’’ are reported commonly in the clinical man- tion to mental disorders. Extensive advances in functional
agement of AD patients, although specific diagnostic criteria neuroimaging that are currently available, as well as in devel-
for a ‘‘sundowning’’ episode have been difficult to define opment, promise to provide us with dynamic images of
(96,102). Despite extensive clinical research in this area, the brain function as it transitions throughout the sleep/wake
pathophysiology of sundowning, including its relationship cycle. In this manner, the functional neuroanatomic basis
with brain mechanisms that control sleep/wake and circa- of the electrophysiologic abnormalities can be determined
dian regulation remain unclear. Overall, the literature on and interventions designed targeting not only specific neu-
sleep in AD suggest that the primary defect in this disease rotransmitter systems but systems that are specific to a dis-
is the more general neurodegenerative changes that lead to crete brain region that is responsible for the sleep/wake dis-
the profound cognitive and functional declines of this disturbance. The discovery of the genetic basis of narcolepsy
ease and that the sleep changes are secondary manifestations brings the dream of uncovering the genetic basis of sleep
of the disorder. If sleep is viewed as generated by core sleep disruption into closer view. Application of the advances in
systems that then require an intact neural structure through- cognitive neuroscience to sleep also promise to define the
out the rest of the brain for expression of behavioral states, importance of sleep in shaping cognitive function as well as
then the sleep changes in AD are most likely related to the role of pathological sleep in disrupting waking cognitive
end-organ failure as opposed to pathology in key sleep or activity.
circadian systems themselves.
Parkinson’s Disease ACKNOWLEDGMENTS

Light, fragmented sleep occurs frequently in Parkinson’s
disease (PD) patients. Sleep problems have been reported This work was supported in part by MH30915, MH52247,
in as many as 74% to 96% of patients (103). Complaints MH37869, MH00295, MH01414, MH45203, and the
include frequent awakenings, early awakening, nocturnal Theodore and Vada Stanley Foundation. The authors thank
craps, pains, nightmares, vivid dreams, visual hallucinations, Robert Y. Moore for his contributions. EAN has received
vocalizations, somnambulisms impaired motor function honoraria for speaking from Glaxo-Wellcome, Inc. MK has
during sleep, myoclonic jerks, excessive daytime sleepiness, received research support from Pfizer Inc., Janssen Pharma-
REM sleep behavior disorder, and sleep-related violence ceutical Research Foundation, and Eli Lilly Pharmaceutical.
leading to injury (103–106). These changes may result from
the disease itself, or to complication from treatment with
dopaminergic agents (103–105,107). Additionally, depres- REFERENCES
sion is common in PD and the sleep disruption may in part
1. Aserinsky E, Kleitman N. Regularly occurring periods of eye
be related to this comorbid disorder (103). motility, and concomitant phenomena, during sleep. Science
Sleep architecture abnormalities include increased awak- 1953;118:273–274.
enings, reductions in stages 3 and 4 sleep, REM sleep, and 2. Jouvet M. Recherches sur les structures nerveuses et les meca-
nismes responsables des differentes phases du sommeil physio- 23. Buchsbaum MS, Gillin JC, Wu J, et al. Regional cerebral glu-
logique. Arch Italiennes Biol 1962;100:125–206. cose metabolic rate in human sleep assessed by positron emission
3. Steriade M, Amzica F, Contreras D. Synchronization of fast tomography. Life Sci 1989;45:1349–1356.
(30 to 40 Hz) spontaneous cortical rhythms during brain activa- 24. Maquet P, Franck G. Cerebral glucose metabolism during the
tion. J Neurosci 1996;16:392–417. sleep-wake cycle in man as measured by positron emission to-
4. Nofzinger EA, Nichols TE, Meltzer CC, et al. Changes in fore- mography. In: Horne J. Sleep ‘88. New York: Gustav Fischer,
brain function from waking to REM sleep in depression: prelim- 1989:76–78.
inary analyses of [18F] FDG PET studies. Psychiatry Res Neu- 25. Clark C, Dupont R, Lehr P, et al. Is there a relationship between
roimag 1999;91:59–78. delta sleep at night and afternoon cerebral blood flow, assessed
5. Destexhe A, Contreras D, Steriade M. Cortically-induced co- by HMPAO-SPECT in depressed patients and normal control
herence of a thalamic-generated oscillation. Neuroscience 1999; subjects? Preliminary data. Psychiatry Res 1998;84:89–99.
92:427–443. 26. Maquet P, Degueldre C, Delfiore G, et al. Functional neuro-
6. Robbins TW, Everitt BJ. Arousal systems and attention. In: anatomy of human slow wave sleep. J Neurosci 1997;17:
Gazzaniga M. The cognitive neurosciences. Cambridge: MIT 2807–2812.
Press, 1996:703–720. 27. Hofle N, Paus T, Reutens D, et al. Regional cerebral blood
7. Moore RY, Abrahamson EA, van den Pol A. The hypocretin flow changes as a function of delta and spindle activity during
neuron system: an arousal system in the human brain. Arch slow wave sleep in humans. J Neurosci 1997;15:4800–4808.
Italiennes Biol 2001;139:195–205. 28. Andersson JL, Onoe H, Hetta J, et al. Brain networks affected
8. Steriade M, McCarley RW. Brainstem control of wakefulness and by synchronized sleep visualized by positron emission tomogra-
sleep. New York: Plenum Press, 1990. phy. J Cereb Blood Flow Metab 1998;18:701–715.
9. Hobson JA, Pace-Schott E, Stickgold R. Dreaming and the 29. Lydic R, Baghdoyan HA, Hibbard L, et al. Regional brain glu-
brain: towards a cognitive neuroscience of conscious states. cose metabolism is altered during rapid eye movement sleep in
Behav Brain Sci 2000;23:793–842. the cat: a preliminary study. J Comp Neurol 1991;304:517–529.
10. Szymusiak R. Magnocellular nuclei of the basal forebrain: sub- 30. Maquet P, Peters JM, Aerts J, et al. Functional neuroanatomy of
strates of sleep and arousal regulation. Sleep 1995;18:478–500. human rapid-eye-movement sleep and dreaming. Nature 1996;
11. Shiromani PJ, Scammell T, Sherin JE, et al. Hypothalamic regu- 383:163–166.
lation of sleep. In: Lydic R, Baghdoyan HA. Handbook of behav- 31. Drummond SPA, Brown GC, Gillin JC, et al. Altered brain
ioral state control: cellular and molecular mechanisms. Boca Raton, response to verbal learning following sleep deprivation. Nature
FL: CRC Press, 1999:311–326. 2000;43:655–657.
12. Edgar DM. Sleep-wake circadian rhythms and aging: potential 32. Borbély AA. A two process model of sleep regulation. Hum
etiologies and relevance to age-related changes in integrated Neurobiol 1982;1:195–204.
physiological systems. Neurobiol Aging 1994;15:499–501. 33. Buzsaki G. Commentary: two-stage model of memory trace
13. Chemelli RM, Willie JT, Sinton CM, et al. Narcolepsy in orexin formation: a role for ‘‘noisy’’ brain states. Neuroscience 1989;
knockout mice: molecular genetics of sleep regulation [see com- 31:551–70.
ments]. Cell 1999;98:437–451. 34. Llinas RR, Pare D. Commentary: of dreaming and wakefulness.
14. Nishino S, Ripley B, Overeem S, et al. Hypocretin (Orexin) Neuroscience 1991;44:521–535.
deficiency in human narcolepsy [letter]. Lancet 2000;355: 35. Winson J. The biology and function of REM sleep. Curr Opin
39–40. Neurobiol 1993;3:243–248.
15. McCarley RW, Hobson JA. Neuronal excitability modulation 36. Kavanau JL. Sleep and dynamic stabilization of neural circuitry:
over the sleep cycle: a structural and mathematical model. Sci- a review and synthesis. Behav Brain Res 1994;63:111–126.
ence 1975;189:58–60. 37. Buysse DJ, Reynolds CF, Monk TH, et al. The Pittsburgh Sleep
16. Krueger JM, Obal F, Fang J. Humoral regulation of physiologi- Quality Index (PSQI): a new instrument for psychiatric research
cal sleep: cytokines and GHRH. J Sleep Res 1999;8:53–59. and practice. Psychiatry Res 1989;28:193–213.
17. Semba K. The mesopontine cholinergic system: a dual role in 38. Armitage R, Hoffmann RF, Rush AJ. Biological rhythm dis-
REM sleep and wakefulness. In: Lydic R, Baghdoyan HA. turbance in depression: temporal coherence of ultradian sleep
Handbook of behavioral state control: cellular and molecular mech- EEG rhythms. Psychol Med 1999;29:1435–1448.
anisms. Boca Raton, FL: CRC Press, 1999:161–180. 39. Nofzinger EA, Buysse DJ, Reynolds CF, et al. Sleep disorders
18. Luppi P-H, Peyron C, Rampon C, et al. Inhibitory mechanisms related to another mental disorder (nonsubstance/primary): a
in the dorsal raphe nucleus and locus ceruleus during sleep. In: DSM-IV literature review [Review]. J Clin Psychiatry 1993;54:
Lydic R, Baghdoyan HA. Handbook of behavioral state control: 244–255; discussion 256–259.
cellular and molecular mechanisms. Boca Raton, FL: CRC Press, 40. Borbély AA, Tobler I, Loepfe M, et al. All-night spectral analysis
1999:195–212. of the sleep EEG in untreated depressives and normal controls.
19. Braun AR, Balkin TJ, Wesenten NJ, et al. Regional cerebral Psychiatry Res 1984;12:27–33.
blood flow throughout the sleep-wake cycle. An H2(15)O PET 41. Kupfer DJ, Frank E, McEachran AB, et al. Delta sleep ratio:
study. Brain 1997;120:1173–1197. a biological correlate of early recurrence in unipolar affective
20. Nofzinger EA, Mintun MA, Wiseman MB, et al. Forebrain disorder. Arch Gen Psychiatry 1990;47:1100–1105.
activation in REM sleep: an FDG PET study. Brain Res 1997; 42. Armitage R, Hoffmann R, Trivedi M, et al. Sleep macro- and
770:192–201. microarchitecture in depression: age and gender effects. Sleep
21. Calvo JM, Simon-Arceo K. Cholinergic enhancement of REM Res 1997;26:283.
sleep from sites in the pons and amygdala. In: Lydic R, Bag- 43. Lauer CJ, Riemann D, Wiegand M, et al. From early to late
hdoyan HA. Handbook of behavioral state control: cellular and adulthood changes in EEG sleep of depressed patients and
molecular mechanisms. Boca Raton, FL: CRC Press, 1999: healthy volunteers. Biol Psychiatry 1991;29:979–993.
391–406. 44. Dahl RE, Ryan ND, Birmaher B, et al. EEG sleep measures in
22. Sanford LD, Tejani-Butt SM, Ross RJ, et al. Amygdaloid con- prepubertal depression. Psychiatry Res 1991;38:201–214.
trol of alerting and behavioral arousal in rats: involvement of 45. Dew MA, Reynolds CF, Buysse DJ, et al. Electroencephalo-
serotonergic mechanisms. Arch Italiennes Biol 1995;134:81–99. graphic sleep profiles during depression. Effects of episode dura-
tion and other clinical and psychosocial factors in older adults. sleep deprivation on slow wave recovery in schizophrenia. Sleep
Arch Gen Psychiatry 1996;53:148–156. Res 1993;22:143.
46. Rush AJ, Giles DE, Jarrett RB, et al. Reduced REM latency 66. Horne J. Neuroscience. Images of lost sleep [News; Comment].
predicts response to tricyclic medication in depressed outpa- Nature 2000;403:605–606.
tients. Biol Psychiatry 1989;26:61–72. 67. Keshavan MS, Reynolds CF, Montrose D, et al. Sleep and suici-
47. Giles DE, Jarrett RB, Roffwarg HP, et al. Reduced rapid eye dality in psychotic patients. Acta Psychiatr Scand 1994;89:
movement latency: a predictor of recurrence in depression. Neu- 122–125.
ropsychopharmacology 1987;1:33–39. 68. Lewis C, Tandon R, Shipley JR, et al. Biological predictors
48. Sharpley AL, Walsh AE, Cowen PJ. Nefazodone—a novel anti- of suicidality in schizophrenia. Acta Psychiatr Scand 1996;94:
depressant—may increase REM sleep. Biol Psychiatry 1992;31: 416–420.
1070–1073. 69. Maixner S, Tandon R, Eiser A, et al. Effects of antipsychotic
49. Nofzinger EA, Reynolds CF, Thase ME, et al. REM sleep en- treatment on polysomnographic measures in schizophrenia: a
hancement by bupropion in depressed men. Am J Psychiatry replication and extension [Report]. Am J Psychiatry 1997;155:
1995;152(2):274–276. 1600–1602.
50. Gillin JC, Sutton L, Ruiz C, et al. The cholinergic rapid eye 70. Keshavan MS, Reynolds CF, Miewald JM, et al. A longitudinal
movement test with arecoline in depression. Arch Gen Psychiatry study of EEG sleep in schizophrenia. Psychiatry Res 1996;59:
1991;48:264–270. 203–211.
51. Bhatti T, Gillin JC, Seifritz E, et al. Effects of a tryptophan- 71. Keshavan MS, Reynolds CF, Miewald J, et al. Slow-wave sleep
free amino acid drink challenge on normal human sleep electro- deficits and outcome in schizophrenia and schizoaffective disor-
encephalogram and mood. Biol Psychiatry 1998;43:52–59. der. Acta Psychiatr Scand 1995;91:289–292.
52. Gillin JC, Sohn JW, Stahl SM, et al. Ipsapirone, a 5-HT1A 72. Orzack MH, Hartman EL, Kornetsky C. The relationship be-
agonist, suppresses REM sleep equally in unmedicated depressed tween attention and slow wave sleep in schizophrenia. Psycho-
patients and normal controls. Neuropsychopharmacology 1996; pharmacol Bull 1977;13:59–61.
15:109–115. 73. Smith JR, Karacan I, Yang M. Ontongeny of delta activity
53. Ho AP, Gillin JC, Buchsbaum MS, et al. Brain glucose metabo- during human sleep. Electroenceph Clin Neurophysiol 1977;43:
lism during non-rapid eye movement sleep in major depression: 229.
a positron emission tomography study. Arch Gen Psychiatry 74. Feinberg I. Schizophrenia: caused by a default in programmed
1996;53:645–652. synaptic elimination during adolescence. J Psychiatric Res 1982;
54. Nofzinger EA, Price JC, Meltzer CC, et al. Towards a neurobi- 17:319–330.
ology of dysfunctional arousal in depression: the relationship 75. Andreasen NC, Arndt S, Swayze V, et al. Thalamic abnormali-
between beta EEG power and regional cerebral glucose metabo- ties in schizophrenia visualized through magnetic resonance
lism during NREM sleep. Psychiatry Res Neuroimag 2000;98: image averaging. Science 1994;266:294–298.
71–91. 76. Keshavan MS, Reynolds CF, Ganguli R, et al. Electroencephalo-
55. Wu J, Buchsbaum MS, Gillin JC, et al. Prediction of antidepres- graphic sleep and cerebral morphology in functional psychoses:
sant effects of sleep deprivation by metabolic rates in the ventral a preliminary study with computed tomography. Psychiatry Res
anterior cingulate and medial prefrontal cortex. Am J Psychiatry 1991;39:293–301.
1999;156:1149–1158. 77. Van Kammen DP, van Kammen WB, Peters J, et al. Decreased
56. Clark CP, Gillin JC, Golshan S, et al. Polysomnography and
slow-wave sleep and enlarged lateral ventricles in schizophrenia.
depressive symptoms in primary alcoholics with and without a
Neuropsychopharmacology 1988;1:265–271.
lifetime diagnosis of secondary depression and in patients with
78. Keshavan MS, Tandon R. Sleep abnormalities in schizophrenia:
primary major depression. J Affect Disord 1999;52:177–185.
pathophysiologic significance. Psychol Med 1993;23:831–835.
57. Drummond SP, Gillin JC, Smith TL, et al. The sleep of absti-
79. Riemann D. Cholinergic REM induction test: muscarinic su-
nent pure primary alcoholic patients: natural course and rela-
tionship to relapse. Alcohol Clin Exp Res 1998;22:1796–1802. persensitivity underlies polysomnographic findings in both
58. Zarcone VP, Benson KL. Sleep and schizophrenia. Principles and depression and schizophrenia. J Psychiatric Res 1994;28:
practice of sleep medicine. Philadelphia: WB Saunders, 1994: 195–210.
105–214. 80. Griffith JM, O’Neill JE, Petty F, et al. Nicotinic receptor desen-
59. Benca RM, Obermeyer WH, Thisted RA, et al. Sleep and psy- sitization and sensory gating deficits in schizophrenia. Biol Psy-
chiatric disorders: a meta-analysis. Arch Gen Psychiatry 1992; chiatry 1998;44:98–106.
49:651–668. 81. Benson KL, Faull KF, Zarcone VPJ. Evidence for the role of
60. Ganguli R, Reynolds CF, Kupfer DJ. Electroencephalographic serotonin in the regulation of slow wave sleep in schizophrenia.
sleep in young, never medicated schizophrenics. Arch Gen Psy- Sleep 1991;14:133–139.
chiatry 1987;44:36–44. 82. Van Kammen DP, Widerlov E, Neylan TC, et al. Delta sleep-
61. Tandon R, Shipley JE, Taylor S, et al. Electroencephalographic inducing-peptide-like immunoreactivity (DSIP-LI) and sleep in
sleep abnormalities in schizophrenia: relationship to positive/ schizophrenic volunteers. Sleep 1992;15:519–525.
negative symptoms and prior neuroleptic treatment. Arch Gen 83. Porkka-Heiskanen T, Strecker RE, Thakkar M, et al. Adeno-
Psychiatry 1992;49:185–194. sine: a mediator of the sleep-inducing effects of prolonged wake-
62. Keshavan MS, Anderson S, Pettegrew JW. Is schizophrenia due fulness. Science 1997;276:1265–1268.
to excessive synaptic pruning in prefrontal cortex? The Feinberg 84. Ferre S. Adenosine-dopamine interactions in the ventral stria-
hypothesis revisited. J Psychiatric Res 1994;28:239–265. tum. Implications for the treatment of schizophrenia. Psycho-
63. Werth E, Achermann P, Borbely AA. Fronto-occipital EEG pharmacol (Berl) 1997;133:107–120.
power gradients in human sleep. J Sleep Res 1997;6:102–112. 85. Keshavan MS, Pettegrew JW, Reynolds CFJ, et al. Slow wave
64. Keshavan MS, Reynolds CF, Miewald JM, et al. Delta sleep sleep deficits in schizophrenia: pathophysiologic significance.
deficits in schizophrenia: evidence from automated analyses of Psychiatry Res 1995;57:91–100.
sleep data. Arch Gen Psychiatry 1998;55:443–448. 86. Steriade M. Brain electrical activity and sensory processing dur-
65. Benson KL, Sullivan EV, Lim KO, et al. The effect of total ing waking and sleep states. In: Kryger MH, Roth T, Dement
WC, eds. Principles and practice of sleep medicine. Philadelphia: wakefulness and severity of dementia in an Alzheimer’s disease
WB Saunders, 1994:105–214. special care unit. J Gerontol, Series A, Biol Sci Med Sci 1995;50:
87. Salin-Pascual RJ, Herrera-Estrella M, Galicia-Polo L, et al. M303–M306.
Olanzapine acute administration in schizophrenic patients in- 99. Cohen-Mansfield J, Werner P, Freedman L. Sleep and agitation
creases delta sleep and sleep efficiency. Biol Psychiatry 1999;46: in agitated nursing home residents: an observational study. Sleep
141–143. 1995;18:674–680.
88. Hinze-Selch D, Mullington J, Orth A, et al. Effects of clozapine 100. Prinz PN, Peskind ER, Vitaliano PP, et al. Changes in the
on sleep: a longitudinal study. Biol Psychiatry 1997;42:260–266. sleep and waking EEGs of nondemented and demented elderly
89. Neylan TC, Van Kammen DP, Kelley ME, et al. Sleep in schizo- subjects. J Am Geriatr Soc 1982;30:86–93.
phrenic patients on and off haloperidol therapy: clinically stable 101. Ancoli-Israel S, Klauber MR, Butters N, et al. Dementia in
vs. relapsed patients. Arch Gen Psychiatry 1992;49:643–649. institutionalized elderly: relation to sleep apnea. J Am Geriatr
90. Nofzinger EA, vanKammen DP, Gilbertson MW, et al. Elec- Soc 1991;39:258–263.
troencephalographic sleep in clinically stable schizophrenic pa- 102. Cohen-Mansfield J, Marx MS, Rosenthal AS. Dementia and
tients: two- vs. six-weeks neuroleptic free. Biol Psychiatry 1993; agitation in nursing home residents: how are they related? Psy-
33:829–835. chol Aging 1990;5:3–8.
91. Dahl RE. The development and disorders of sleep. Adv Pediatr 103. Tandberg E, Larsen JP, Karlsen K. A community-based study of
1998;45:73–90.
sleep disorders in patients with Parkinson’s disease. Mov Disord
92. Lauer CJ, Schreiber W, Holsboer F, et al. In quest of identifying
1998;13:895–899.
vulnerability markers for psychiatric disorders by all-night polys-
104. Factor SA, McAlarney T, Sanchez-Ramos JR, et al. Sleep disor-
omnography. Arch Gen Psychiatry 1995;52:145–153.
93. Reynolds CF, Monk TH, Hoch CC, et al. Electroencephalo- ders and sleep effect in Parkinson’s disease. Mov Disord 1990;
graphic sleep in the healthy ‘‘old old’’: A comparison with the 5:280–285.
‘‘young old’’ in visually scored and automated measures. J Ger- 105. Stocchi F, Barbato L, Nordera G, et al. Sleep disorders in Parkin-
ontol 1991;46:M39–M46. son’s disease. J Neurol 1998;245:S15–S18.
94. Monk TH, Reynolds CF, Machen MA, et al. Daily social 106. Bernath O, Guilleminault C. Sleep-related violence, injury, and
rhythms in the elderly and their relation to objectively recorded REM sleep behavior disorder in PD. Neurology 1999;52:1924.
sleep. Sleep 1992;15:322–329. 107. Kales A, Ansel RD, Markham CH, et al. Sleep in patients with
95. Reynolds CF, Hoch CC, Buysse DJ, et al. Sleep after spousal Parkinson’s disease and normal subjects prior to and following
bereavement: a study of recovery from stress. Biol Psychiatry levodopa administration. Clin Pharmacol Therapeut 1971;12:
1993;34:791–797. 397–406.
96. Bliwise DL. Dementia. In: Kryger MH, Roth T, Dement WC, 108. Kostic VS, Susic V, Przedborski S, et al. Sleep EEG in depressed
eds. Principles and practice of sleep medicine, second ed. Philadel- and nondepressed patients with Parkinson’s disease. J Neuropsy-
phia: WB Saunders, 1994:790. chiatry Clin Neurosci 1991;3:176–179.
97. Chenier MC. Review and analysis of caregiver burden and nurs- 109. Apps MC, Sheaff PC, Ingram DA, et al. Respiration and sleep
ing home placement. Geriatr Nurs 1997;18:121–126. in Parkinson’s disease. J Neurol Neurosurg Psychiatry 1985;48:
98. Bliwise DL, Hughes M, McMahon PM, et al. Observed sleep/ 1240–1245.
Neuropsychopharmacology: The Fifth Generation of Progress. Edited by Kenneth L. Davis, Dennis Charney, Joseph T. Coyle, and
Charles Nemeroff. American College of Neuropsychopharmacology 䉷 2002.

C134 1945-1960 PDF

Uploaded by

Copyright:

Available Formats

C134 1945-1960 PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

C134 1945-1960 PDF

Uploaded by

Copyright:

Available Formats

134

SLEEP DISTURBANCES ASSOCIATED

FIGURE 134.1. Arousal systems in the

Altered Neuroanatomy Altered Physiology

Future Directions significant losses or in acute depressive episodes. In a study

Parkinson’s Disease ACKNOWLEDGMENTS

You might also like