Motor Control and Dyscontrol in Sleep

Abstract

This chapter deals with motor functions and dysfunctions in sleep. A variety of motor disorders may appear during sleep interfering with sleep. After a brief summary of motor control in wakefulness the chapter addresses how this control changes during different stages of sleep and how its dyscontrol may arise and affect an individual. We classify motor disorders in sleep into two broad categories: Diurnal movement disorders persisting in sleep and those occurring exclusively in sleep. Following a brief description of these entities we conclude the chapter by addressing approach to these motor disorders in sleep and outlining principles of treatment.

Some portions of this chapter are partly modified from Chap. 28, “Motor Functions and Dysfunctions of Sleep,” in the third edition of this book (2009), authored by Hening WA, Allen RP, Walters AS and Chokroverty S.

Introduction

Motor control in humans is the end result of a complex, intricate interplay between several central and peripheral loci, including the cerebral cortex, basal ganglia, brain stem motor centers, cerebellum, spinal cord, and the peripheral neuromuscular system [1–3]. The delicate balance of excitatory and inhibitory influences, created through the coordination of these systems, determines the occurrence of volitional movements and the suppression of unwanted ones. As an individual progresses from drowsiness to light sleep, slow-wave sleep (SWS), and rapid eye movement (REM) sleep, several additional factors come into play to further modulate motor activity. In recent years, there has been growing interest and a lot of research into normal and abnormal motor control in sleep. Disorders of motor control that occur in sleep may be those that are present during the day (diurnal movement disorders) but that impact sleep, either directly through their motor effects or indirectly through a variety of other mechanisms. Alternately, some motor disturbances are exclusive to sleep or the sleep–wake transition period. Motor disturbances in this latter group are generally classified as sleep disorders; the third edition of the International Classification of Sleep Disorders (ICSD-3) includes them in a specific category of sleep disorders called Sleep-Related Movement Disorders [4], and the American Academy of Sleep Medicine has elaborated on rules for recognizing them during polysomnography (PSG) [5].

This chapter discusses how motor control in sleep differs from that in wakefulness, classifies and describes disorders of motor control that intrude into sleep, including recommended treatment options, and discusses the investigative techniques employed in the evaluation of motor disorders in sleep. Non-motor symptoms in hypokinetic and hyperkinetic movement disorders that impact sleep are also briefly discussed. The main purpose of this chapter is to provide an overview of the various motor disturbances that impact sleep and to demonstrate a usable approach to the diagnosis and management of these disorders. This chapter should be useful to sleep medicine practitioners who deal with patients presenting with abnormal movements in sleep; however, readers are referred to several excellent volumes that discuss movement disorders in greater depth [6, 7].

Motor Control in Wakefulness and Sleep

The motor system is organized in a functional hierarchy at three levels: the forebrain, brain stem, and spinal cord (Fig. 39.1a, b). These three levels are under the control of two subcortical systems, the cerebellum and basal ganglia, and receive sensory inputs. The spinal cord is the lowest level in the scheme of the motor organization and contains neuronal circuits mediating reflexes and rhythmic movements. The motor neurons and interneurons of these circuits receive inputs both segmentally and supraspinally from higher centers. All motor commands ultimately converge either directly or indirectly (through the descending brain stem pathways) on the motor neurons in the anterior horn cells of the spinal cord, which Sherrington called the “final common pathway” for motor actions [1]. In the next higher level at the brain stem, there are two descending systems: the medial descending systems consisting of the reticulospinal, vestibulospinal, and tectospinal tracts controlling posture and the lateral descending system consisting mainly of the rubrospinal system, controlling more distal limb muscles (Fig. 39.2a, b). The highest level, located in the cerebral cortex, controls spinal motor neurons directly through the corticospinal tracts and the brain stem motor neurons through the corticobulbar pathways. The cerebral cortex also controls spinal motor neurons indirectly through its influence on the descending brain stem systems. The major subcortical inputs to the motor cortex originate from the cerebellum and basal ganglia controlling the cerebral cortex through their projections to the thalamic nuclei.

Fig. 39.1

a Showing schematically the principal relationships between cerebellum and other components of the motor system. PMC Premotor cortex, SMA Supplementary motor area, MC Motor cortex, SC Somesthetic cortex, SPL Superior Parietal Lobule, DCN Deep Cerebellar Nuclei, VN Vestibular Nucleus, RN Red Nucleaus Magnocellular portion, PN Ponitine Nuclei, IO Inferior Olive, RF Reticular formation, VPLo The oral portion of the ventral posterolateral nucleus of the ventrolateral thalamus. b Schematic diagram to show the basal ganglia-thalamocortical motor circuit. PMC, SMA, MC, SC, SPL: As in figure a GPi Globus Pallidus, internal segment, SNr Substantia Nigra pars reticulata. VM Ventromedial group of brain stem descending pathways (see text). DL Dorsolateral group of brain stem descending pathways (see text). VLo/VLm The nucleus ventralis lateralis pars oralis and ventralis lateralis pars medialis of the thalamus

Fig. 39.2

Schematic diagram showing brain stem motor control. a Medial motor control pathway in the brain stem. b Lateral motor control pathway in the brain stem. (Modified from Ref. [8])

The prefrontal cortex is responsible for the planning and initiation of voluntary movement. The premotor cortex and the supplementary motor cortex are involved in the programming of voluntary movement. Other parts of the cerebral cortex (e.g., somatosensory and association cortices including those responsible for tactile sensations, vision, and hearing) send their projections to the motor cortex for the coordination of skilled movements. Two subcortical circuits, the thalamocorticostriate and dentatorubrothalamocortical circuits, play a significant role in controlling coordination, posture, and muscle tone.

Muscle Tone, Posture, and Reflexes

In addition to voluntary targeted activity, motor control includes the control of muscle tone and posture directed by central generators and spinal stretch reflexes [1]. Phasic muscle contractions producing movements occur on a background of constant muscle tone, which can be defined as a state of mild contraction due to the firing of a few motor units of the muscle (i.e., due to asynchronous sustained firing of motor neurons). Clinically, muscle tone is perceived as the resistance offered by the muscle to stretch as a result of passive movement (flexion or extension) of a joint. The resting muscle tone is produced by viscosity and elastic properties of the muscle. Muscle tone is related to posture, state of alertness, and the degree of muscle stretch. Reticulospinal and vestibulospinal tracts influence the alpha and, to an extent, also the gamma motor neurons to maintain muscle tone.

Movements can be divided into three categories: voluntary, rhythmic, and reflexive. Reflexes are involuntary movements elicited by peripheral stimuli. Voluntary movements are goal-directed. The circuits for rhythmic movements lie in the spinal cord and brain stem. Alterations of muscle tone and movements may be negative (atonia and paralysis) or positive (hypertonia and abnormal movements). Central pattern generators responsible for locomotion are located in the mesencephalic and spinal locomotor centers. Central pattern generators in the midbrain and pons drive spinal locomotor generators. The mesencephalic locomotor region is under the control of the cerebral and limbic cortex (Fig. 39.3). The nature, however, of pattern generators is uncertain because the exact connection of interneurons is not known.

Fig. 39.3

Schematic diagram showing brain stem motor control: Mesencephalc locomotor ergion (MLR) and medial reticular formation (MRF) activating spinal locomotor region after receiving inputs from cerebellum, visual and motor cortices. (Modifed from Fig. 36.10 in “Locomotion.” Ref. [8])

Motor Control in Wakefulness

In wakefulness, several circuits are involved in the generation and modulation of voluntary movement. These include the cortico-basal ganglionic-thalamo-cortical circuit, cortico-ponto-cerebello-thalamo-cortical circuit, descending brain stem motor pathways, brain stem, and spinal segmental circuits. All these circuits are influenced by peripheral afferent inputs. To summarize, the cerebellum participates in the initiation, timing, and coordination of the movements; the basal ganglia help in influencing the direction, force, and amplitude of the movements, as well as the internal generation and assembly of movements; and the cerebral cortex selects, plans, programs, and commands the movement. The corticospinal system then distributes the commands, and the segmental spinal motor apparatus drives the muscles to execute the movements.

Changes in Motor Control in Sleep

In the simplest sense, it can be said that sleep causes a progressive decrease in muscle tone and inhibits gross movements. However, motor control in sleep is far more complex than this simple statement suggests. As in wakefulness, organized movements during sleep are the result of intricate checks and balances at multiple levels of the central nervous system—cerebral cortex, basal ganglia, thalamus, brain stem, cerebellum, and spinal cord. Sleep in general is dominated by central inhibitory drive, but the excitatory mechanism intermittently breaks through the inhibitory phase in normal individuals giving rise to physiologic motor activities during sleep (e.g., gross body shifts and hypnic jerks). It is the breakdown of this delicate mechanism of organized movements during non-rapid eye movement (NREM) sleep and REM sleep, due to a number of inciting factors and diseases, that causes abnormal motor events in sleep. In particular, lower centers (e.g., mesencephalic locomotor region and spinal locomotor generators) are released from controls by forebrain mechanisms as a result of a variety of conditions causing abnormal jerks, shakes, and screams during sleep. It is notable that Pakhomov in 1947 first observed a reduction in muscle tone in finger flexors during sleep [1]. The discovery of REM sleep associated with phasic eye movements and desynchronized EEG in 1953 by Aserinsky and Kleitman [9] paved the way to intense research into motor control during sleep. In 1959 Jouvet and Michel [10] observed muscle atonia in cats during REM sleep. This was followed by Berger’s [11] observations of similar atonia in laryngeal muscles during human REM sleep, thus completing the description of REM sleep phenomena. Fundamental contributions to understanding motor control and muscle tone during sleep were subsequently made by Pompeiano [12], Chase and Morales [13], and Morrison [14] and Siegel [15]. To paraphrase Chase and Morales, “motor control landscape” in sleep is characterized by “storms of inhibition” coupled with brief “whirlwinds of excitation” directed toward the “final common pathway” [16, 17].

Modulation of Neuronal Firing

During wakefulness, most central neurons fire irregularly, at different frequencies in different brain regions. In NREM sleep, and especially during SWS, they fire more slowly compared to wakefulness (lower frequency) but tend to fire in bursts. During SWS, for example, thalamic cells fire slowly and are less responsive to afferent activity. These changes are related to the greater synchronization of surface electrical activity as measured by electroencephalography (EEG) during sleep. In REM sleep, by contrast, cellular activity is increased in many motor regions of the brain such as the primary motor cortex [18], thalamus, red nucleus, and cerebellum [19]. This increased firing in the motor centers of the brain is presumably balanced by the increased descending drive that occurs during REM sleep.

Motor Neuronal Modulation

More focused studies have examined motor neurons in the brain stem and spinal cord and traced backward some of the descending influences which modulate them depending on state. Recent studies have begun to establish the brain stem centers whose altered activity is related to the various stages of sleep and wake causing alterations in motor activity [20]. Particular information has been obtained about the control of REM sleep [21]. Neurons located near the border between the pons and midbrain such as the pedunculopontine nucleus (PPN) and laterodorsal tegmental (LDT) nuclei [22] appear to release acetylcholine into the more central reticular formation of the pons and medulla. These neurons can be divided into two classes: REM-on cells are selectively active during REM, while wake/some REM-on cells are also activated during wake. The REM-on cells are selectively inhibited by serotonin [23]. Various, as yet poorly defined, centers or cell groups in the reticular formation are then stimulated [24] to exert descending influences that act upon motor neurons. One current model suggests that cholinoceptive cells in a pontine inhibitory area project to the medulla where they release glutamate to stimulate inhibitory neurons in the medullary reticular formation [25]. Some of this modulation may also occur through suppression of the orexinergic system, which appears to cause arousal and increased motor activity in wake [26].

Reflex Modulation

Since much of motor behavior is generated, at least in part, by reflexes, studying reflexes can be quite relevant to examining changes of the motor system with state. The most commonly studied reflex has been the Achilles tendon reflex or its electrical counterpart, the H reflex. This reflex is diminished in NREM sleep, especially SWS, and then almost completely abolished in REM sleep, especially during rapid eye movements (REMs) [27]. Polysynaptic spinal reflexes are similarly depressed in NREM and REM sleep [28]. A somewhat related bulbar reflex, the response of the genioglossus muscle to negative pressure in the airway, is decreased in NREM sleep [29] and may be further reduced in REM sleep [30, 31]. This reduced response has significant consequences, because reduced reflex gain may contribute to airway collapse and respiratory difficulties in sleep, especially obstructive sleep apnea (OSA). It has been noted that some brain stem reflexes, such as vestibular reflexes and the blink reflex, show decreased gain in NREM sleep but may then recover partially in REM sleep [32–34]. This recovery of some brain stem reflexes during REM sleep parallels the relatively greater activity of the eye muscles, compared to trunk and limb muscles at that time, and reinforces the mixed picture of excitation and inhibition characteristic of REM. Even drowsiness, short of actual sleep, can attenuate some reflexes, such as the vestibulo-ocular reflex, which has two outputs: quick restorative jerks to head rotation and slower smooth compensatory eye deviations [35]. The more polysynaptic quick jerks are more easily suppressed by even modest drowsiness. While all these various changes in reflex gain indicate altered excitability, they do not indicate where in the reflex arc the changes occur.

The basis for much of the reduced reflex gain during sleep is most likely inhibition of motor output, rather than decrease of sensory response. In one supportive study, Morrison et al. [36] created pontine tegmental lesions in cats that caused REM sleep without atonia. They found that both orienting to tone stimuli and acoustic startle responses were evident in REM sleep in the lesioned cats, but rare or absent throughout sleep in intact cats. Since the same tones elicited brain stem generated ponto-geniculo-occipital (PGO) waves in both normal and lesioned cats, it seems likely that the block to the further responses of orienting and startle reflexes is on the motor side of the reflex arc. PGO waves were also identified in the human pons, recorded during placement of a pedunculopontine nucleus stimulator [37]. On the other hand, sensory transmission may itself be altered by sleep. Studies have indicated that in primary afferent neurons located both in the spinal cord [38] and in the brain stem [39] there is a significant, presynaptically mediated decrease in responsiveness during REM sleep but not NREM sleep.

Reflexes can also change their characteristic motor output in sleep [40] indicating that sleep is not merely a general change in activity levels but a rearranged organization of responsiveness. Sensory stimulation, which would cause motor neuron excitation in waking, can cause additional inhibitory potentials in sleep [41]. In addition, certain reflexes which would be abnormal during waking, such as the Babinski sign, may be elicited in sleep [42, 43].

Effect of NREM Sleep on Motor Control

Progressive muscle hypotonia is a cardinal feature of sleep. In NREM sleep, motor activity is less than in the waking or resting state. At the onset of NREM sleep, intracellular microelectrode recording of motor neurons by Chase et al. [44] clearly showed either no change in membrane potential or a slight hyperpolarization. This, as well as disfacilitation of brain stem motor neurons controlling muscle tone, likely explains the mild muscle hypotonia seen in NREM sleep. Postural shifts, which may signal stage changes (into or out of wake or REM), occur. There are also small flickering movements, called sleep myoclonus, which may cause no apparent movement and are associated with very brief, highly localized electromyography (EMG) potentials seen on PSG [45, 46]. In some cases, these movements may have a greater amplitude and be of increased frequency, at which point they are called excessive fragmentary myoclonus, (EFM, discussed later in this chapter) but the significance of these movements, if any, is unknown [47]. The frequency of all movements decreases with depth of sleep, being least in SWS [48–50]. Postural shifts rarely occur before entrance into SWS. A number of abnormal motor activities such as somnambulism or periodic limb movements of sleep (PLMS) occur predominantly during NREM sleep. In infants, who move more in sleep, most NREM movements are generalized, full body movements or jerks, while REM movements tend to be more focal and uncoordinated [51].

Effect of REM Sleep on Motor Control

REM sleep is dramatically different from NREM. During REM sleep there is tonic reduction in muscle tone, even below that of SWS, in the presence of a highly active forebrain (paralyzed body with activated brain) with inhibition of the mesencephalic locomotor region. This is a protective mechanism to prevent abnormal movements during REM sleep in the presence of highly active cerebral cortex and forebrain regions. However, bursts of small movements (“phasic twitching”), similar to those seen in NREM sleep but more clustered, occur in REM sleep in association with bursts of REMs. During REM there is a close balance between strong upper motor center excitation and inhibition at the level of the motor effecter. When the inhibitory influences break down, significant motor activity may be released. Infants lack this inhibition and have more movement during REM. Inhibition can also be disrupted by lesions in the brain stem of animals which destroy the inhibitory centers [52, 53] or, it is believed, in human sleep disorders such as REM Sleep Behavior disorder (RBD). The resulting movements may represent an “acting out” of dreams, which characteristically have a motoric component [54, 55].

The mechanism of muscle atonia during REM sleep includes an activation of a polysynaptic descending pathway from the perilocus coeruleus alpha region in the pons to the lateral tegmentoreticular tract, nucleus gigantocellularis, and magnocellularis in the medial medulla (the inhibitory area of Magoun and Rhines), ventral tegmentoreticular, and reticulospinal tracts to the alpha motor neurons, causing hyperpolarization and thus giving rise to muscle atonia (see Fig. 41.5) [see also Chap. 41]. Immunocytochemical techniques detected an increased number of C-Fos (a nuclear protein synthesized during neuronal activation) labeled cells in the inhibitory region of Magoun and Rhines [56] during REM sleep. A key element in the REM sleep-generating mechanism in the pons is the activation of GABAergic neurons located in a subgroup of pontine reticular formation, as well as GABAergic neurons in the ventrolateral periaqueductal gray. An activation of GABAergic neurons causes an activation or disinhibition of cholinergic neurons and inhibition of noradrenergic and serotonergic neurons in the pons. The cholinergic neurons, in turn, excite pontine glutamatergic neurons projecting to the glycernergic pre-motor neurons in the medullary reticular neurons, causing hyperpolarization of the motor neurons and motor paralysis during REM sleep. Disfacilitation of motor neurons as a result of a reduction of the release of serotonin and norepinephrine partially contribute to muscle atonia. While many interneuronal regions of the brain stem show increased activity in REM sleep, motor nuclei (masseter, facial, and hypoglossal nuclei) show depressed activity. This is consistent with studies that have shown glycinergic inhibition of hypoglossal neurons in REM [57], perhaps contributing to the difficulties with airway patency in this sleep stage.

Motor neuron control at the cellular level results from synaptic transmission as manifested by the presence of excitatory post-synaptic potentials (EPSPs) or inhibitory post-synaptic potentials (IPSPs). During REM sleep, motor neurons are hyperpolarized by 2–10 mv (Fig. 39.4). There is post-synaptic inhibition causing a decrease in 1a monosynaptic EPSPs resulting in motor neuron hyperpolarization. During wakefulness and NREM sleep, there are a few spontaneously occurring low amplitude IPSPs, but during REM sleep, in addition to an increase of these low amplitude ISPSs, there are additional high amplitude sleep-specific IPSPs noted. These are generated by sleep-specific inhibitory interneurons located mainly in the brain stem (immunocytochemical techniques are used to prove this observation), which send long projecting axons to the spinal and short axons to the brain stem motor neurons. Glycine, the major inhibitory neurotransmitter, is the driving force for these IPSPs. The REM-specific IPSPs are abolished after strychnine (a glycine antagonist) administration [58] by microiontophoretic application into the ventral spinal cord but not after application of bicuculline or picrotoxin (GABA antagonists), thus proving that it is glycine and not GABA which is responsible for these IPSPs. Intermittently during REM sleep there are excitatory drives causing motor neuron depolarization shifts as a result of EPSPs [28]. Glutamate is the excitatory amino acid for the REM EPSPs as these EPSPs are abolished by kynurenic acid (a glutamate antagonist) application near the ventral horn cells but not by NMDA antagonists. Muscle movements caused by these excitatory drives during REM sleep are somewhat different from the movements occurring during wakefulness. These are abrupt, jerky, and purposeless. EPSPs during REM sleep reflect increased rates of firing in the motor facilitatory pathways during REM sleep and enhanced IPSPs during REM sleep check these facilitatory discharges thus balancing the motor system during this activated state; otherwise the blind, unconscious subject will jump out of bed, as may happen in RBD (see below) [14]. Facilitatory reticulospinal fibers are responsible for transient EPSPs (phasic discharges) causing muscle twitches in REM sleep. Corticospinal or rubrospinal tracts are not responsible for these twitches because destruction of these fibers in cats [12] does not affect these twitches.

Fig. 39.4

Intracellular recording from a masseter motoneuron of a cat showing membrane potential and state changes during wakefulness, active (REM), and quiet (NREM) sleep. Note abrupt membrain hyperpolarization during active sleep accompanied by neck muscle hypotonia in the eletromyagram (EMG) and depolarization during wakefulness. In quiet sleep, membrane potential resembles that in wakefulness. Electroencephalogram (EEG); EOG, electro-otulogram: PG0, ponto-geniculo.occipital potential. (From Ref. [59])

There are several additional factors that dictate the nature of motor control during sleep.

Circadian Activity Cycles

In humans, as in most animal species, the motor system’s level of activity is dependent on an underlying circadian rhythm, even in the absence of a day–night light cycle (i.e., under constant conditions). As discussed in Chap. 6, circadian regulation of sleep, like many other important physiologic variables such as temperature, also shows circadian periodicity. While there may be a number of supplementary oscillators which control these rhythms [60–62], the suprachiasmatic nucleus (SCN) is thought to contain the most important oscillator [63] and to be the center which is responsible for the circadian variation of motor activity. The basic mechanism is a transcription–translation feedback loop [64, 65], elements of which are widely distributed in peripheral tissues [66]. Another mediator may be an expression of clock genes, such as the period gene, rPer2, which are widely distributed in different tissues and controlled by the SCN, and a large amount of SCN output is channeled through the subparaventricular zone of the hypothalamus (SPZ), which contains a specific region specialized in modulating motor activity [67, 68]. The SPZ also acts as an integrating center for various environmental influences that can impact on circadian rhythm, such as food availability, ambient temperature, and social interactions. Some of the hypothalamic regulation of motor activity is via the hypocretin system, which stimulates activity at the end of the activity period (the evening for humans) [69, 70]. In humans, activity is concentrated during the daytime hours and sleep at night; decreased movements in sleep may in fact be a marker of better quality sleep [71]. Movements in sleep are associated with autonomic surges, which may have long-term cardiovascular implications [72]. Therefore, this circadian periodicity of movement may be a protective factor. The clear separation of these periods begins to break down in normal aging and in cases of poor sleep or sleep-related disorders as well as in many movement disorders and degenerative conditions.

Development and Aging

The nature and occurrence of movements in sleep are affected by age. Movement frequency during sleep is greatest in infants, then decreases with age, even within the first few months of life [73, 74]. This is most likely due to the immaturity of regulatory mechanisms that maintain motor control in sleep. For instance, in one study, Kohyama [75] found that younger infants appear to lack the profound motor inhibition during phasic REM that is seen in older children and adults. Perhaps as a result of such immaturity, parasomnias such as somnambulism, or soliloquy are present with a greater prevalence during childhood, tending to decrease with age from early childhood on [76, 77]. Rhythmic movement disorder (RMD, see below) is also primarily a disorder of early childhood but may persist into adolescence or even adulthood in patients with significant cerebral injury or with autism [78]. With age, movements in sleep decrease; one study showed that position shifts during sleep decreased from 4.7 per hour in 8–12 year old subjects to 2.1 per hour in those 65–80 years old [79]. With aging, excessive motor activity may emerge again [80], including PLMS, RBD, and increased instability in NREM sleep manifesting as an increase in cyclic alternating pattern (CAP) [81]. In fact, these changes in CAP (see below) at various ages are increasingly being seen as a factor in age-related variation in the frequency and nature of motor dysfunction in sleep.

Drowsiness and the Sleep–Wake Transition Period

Even before sleep onset, the motor system reduces its level of activity. The sleep–wake transition period, a period of relative repose, has been called the predormitum by Critchley [82]. The subsequent transition to sleep is signaled by a variety of behavioral and EEG features [83]. The transition to sleep is a frequent inciting factor for a variety of jerks, jumps, and starts that may be a cause of concern to patients and their bedpartners. The most common of this is the “sleep start” or hypnic jerk. Another movement disorder activated in the transition to sleep is a form of propriospinal myoclonus [84]. These conditions are discussed in greater detail below. It is also during the sleep–wake transition period that the symptoms of restless legs syndrome (RLS) become prominent (see Chap. 40). RLS is relatively distinctive in that, unlike almost all other movement disorders, it is activated by rest.

Arousals, brief periods of interrupted, lighter sleep that may or may not lead to full awakening, are often associated with movements. Arousals may both follow and lead movements such as body shifts. Abnormal movements, such as parkinsonian tremor [85] may recur during arousals. Sleep-related movements, such as PLMS, may provoke frequent arousals or even awakenings and may also continue during periods of arousal from sleep.

Metabolically, physiologically and behaviorally the period just after awakening, or “postdormitum,” is distinctly different from the predormitum [86–88]. Sleep offset occurs with abrupt changes in the EEG activity, unblocking of the afferent stimuli and restoration of postural muscle tone accompanied by a reduction of cerebral blood flow with concomitant decrement of cerebral metabolism as compared with that in presleep wakefulness [89]. This is in contrast to sleep onset with gradual changes in the EEG, blockade of the afferent stimuli at the thalamic level (essentially converting an “open” brain into a “close” one) and a reduction of postural muscle tone [90]. Because of these differences between the two states certain motor or other disorders preferentially occur in either predormitum (e.g., propriospinal myoclonus at sleep onset, hypnic jerks, RMD, hypnagogic imagery, and exploding head syndrome) or postdormitum (e.g., sleep inertia, awakening epilepsy of Jang, and sleep benefit in some Parkinson’s disease patients). Sleep paralysis and hallucinations may occur in both states (hypnagogic and hypnopompic). Many of these conditions are discussed in greater detail below.

Effects of Sleep Stage on Motor Control

Changes in motor activity are dependent on the sleep state (i.e., wake, NREM sleep, and REM sleep). As discussed above, there is an orderly progression of loss of muscle tone as an individual proceeds through these stages, with muscle hypotonia being most pronounced in REM sleep, where only the diaphragm and extraocular muscles (as well as the middle ear muscles) are spared from almost complete paralysis. This is the underlying principle in measuring chin EMG during PSG. During wake, chin muscle tone is high and a tonically active chin EMG is interrupted by phasic contractions (facial expressions, tension, chewing, etc.). With relaxation and drowsiness, the level of EMG activity decreases. It further decreases as NREM sleep is achieved and deepens to SWS levels. Then, during REM sleep, EMG activity becomes minimal or even inapparent, although it may be occasionally interrupted with brief, irregular bursts of activity in phasic REM sleep, including the tongue [91]. These changes mirror, to a fair degree, the changes undergone by much of the motor system during sleep.

Any discussion on the effects of sleep stages, as determined by PSG, on motor control needs to be tempered by the fact that sleep staging in 30-second epochs, as is standard [5] (see Chap. 24) is based on rules that are arbitrary and are subject to a great degree of interscrorer variation. Physiologic processes are unlikely to adhere to these convenient timescales. For example, Mahowald and Schenck [92] reported on six patients with marked admixture of features from the different sleep–wake states (i.e., wake, NREM sleep, and REM sleep). These patients showed abnormal distribution of motor activity with relation to sleep features. Motor events, although typical of one sleep stage or state, may less commonly occur in other stages. For example, although PLMS are primarily a sleep phenomenon, these movements may also occur during arousals or periods of wakefulness after sleep onset (periodic limb movement in wake, PLMW), often as part of a periodic sequence of movements that span the sleep–wake divide [93]. PLMS occur primarily in NREM sleep but may also occur in REM sleep [94], especially in disorders of disturbed REM sleep such as narcolepsy and RBD [95, 96]. Patients with somnambulism, which typically occurs in NREM sleep, may show REM sleep motor abnormalities suggestive of RBD [97], and confusional arousals have been reported in REM sleep [98]. Even dream-enacting behavior, traditionally thought of as a REM parasomnia seen in RBD, has been reported in NREM sleep [99].

Table 39.1 summarizes the frequency of normal and abnormal motor activities that occur during the various phases of sleep and waking.

Table 39.1 Persistence of various movements in sleep

Cyclical Alternating Pattern and Movements in Sleep

The evaluation of periodic alternations in EEG activity represents an important additional means of scoring that may be more meaningful than the traditional, AASM mandated epoch-by-epoch scoring of sleep stages from a physiological perspective [5]. In NREM sleep, especially stage N2, this periodicity is common and designated the cyclical alternating pattern (CAP) [100]. First described by Terzano et al. [101], this pattern shows an alternation between bursts of both slow and fast activity (A phase) alternating with a medium frequency, lower amplitude activity (B phase). The burst-like activity (A phase) is associated with autonomic activation (Fig. 39.5). The A phases can include greater or lesser amounts faster frequencies: A1 has the least and A3 the most fast frequencies. A2 and A3 phases are often associated with arousals that can disrupt sleep [102, 103]. A number of different abnormal sleep-related movements are found to be associated with specific phases of the CAP cycle, especially the A phase. These include: PLMS [104] parasomnias or other sleep-related abnormal movements, such as bruxism [105], somnambulism [106, 107], or alternating leg movement activity during sleep (ALMA) [108], and nocturnal paroxysmal dystonia (NPD) [109]. CAP, especially phases A1 and A2, occurs more in early childhood [110], decreases during school age [111], may transiently increase during the adolescent period [112], decreases again in young adulthood, and then finally increases again in older ages [104]. CAP is a normal pattern, but deviations from normal amounts, especially excessive CAP can be abnormal [113].

Fig. 39.5

Shows a cyclic alternating pattern (CAP) as part of a CAP sequence in Stage N2 sleep. The CAP cycle (highlighted) is defined by a Phase A (classified into threee subtypes A1, A2, A3 [subtype A2 is shown here] depending on the amount of high amplitude slow waves and low amplitude fast rhythms) followed by a phase B (relatively inactive) [each phase is 2–60 s in duration]. Bipolar EEG derivation (international electrode placement). Channels 1-6 from top: FP2-F4; F4-C4; C4-P4; P4-O2; F8-T4; T4-T6. Channels 7-11 from top: FP1-F3; F3-C3; P3-O1;C4-A1. (Reproduced with permission from Ref. [114])

Classification of Motor Disorders in Sleep

Motor disorders of sleep can broadly be classified into (1) diurnal movement disorders persisting into sleep and (2) primary sleep motor disorders (exclusive to sleep). The latter category can be further subclassified into disorders of motor control as the subject lays in bed trying to get to sleep, immediately before and at sleep onset, during NREM sleep, during REM sleep, during both NREM and REM sleep, and at sleep offset. These are listed in Box 39.1, and the individual motor disorders are discussed in detail in the following sections.

Box 39.1: Disorders due to Failure of Motor Control in Sleep

I. Diurnal Movement Disorders Persisting in Sleep

• Usually persisting in sleep

  • Symptomatic palatal tremor

• Frequently persisting in sleep

  • Spinal and propriospinal Myoclonus

  • Tics in Tourette’s syndrome

  • Hemifacial spasm

  • Hyperekplexia

• Sometimes persisting in sleep

  • Tremor

  • Chorea

  • Dystonia

  • Hemiballismus

II. Disorders of Motor Control Unique to Sleep

A. Failure of motor contol while resting in bed trying to get to sleep

• Restless legs syndrome (myoclonic-dystonic movement in quiescence)

B. Failure of motor control at NREM sleep onset (including predormitum, an ill-defined stage between sleeping and waking)

• Physiological

  • Physiological body movements and postural shifts

  • Physiological hypnic Myoclonus

  • Hypnic jerks

  • Hypnagogic foot tremor

  • Alternating leg muscle activity

  • Rhythmic limb movements

• Pathological

  • Intensified hypnic jerks

  • Rhythmic movement disorder

  • Propriospinal Myoclonus at sleep onset

C. Failure of motor control during NREM sleep

• Partial Arousal Disorders

  • Confusional arousals

  • Sleepwalking

  • Sleep terrors

  • Sleep-related eating disorder (SRED)

• Others

  • Periodic Limb Movements in Sleep

D. Failure of motor control during REM sleep

• Physiological

  • Phasic muscle bursts (Myoclonus) including fragmentary hypnic myoclonus

  • Phasic tongue movements

  • Phasic rapid eye movements

  • Periorbital integrated potentials (PIPs)

  • Sleep paralysis

• Pathological

  • REM Behavior Disorder (RBD)

  • Sleep paralysis with narcolepsy

  • Familial sleep paralysis

E. Failure of Motor Control in both NREM and REM sleep

• Rhythmic movement disorder

• Catathrenia

• Excessive fragmentary Myoclonus

• Sleep bruxism

• Upper airway obstructive sleep apnea

F. failure of Motor Control during Sleep offset

• Sleep paralysis

• Hypnopompic hallucinations

• Sleep inertia (“sleep drunkenness”)

Description of Individual Motor Disorders of Sleep

A. Diurnal Motor Disorders Persisting into Sleep

As a general rule, most abnormal movements seen during the daytime show a markedly decreased frequency, amplitude, and duration in sleep and tend to be limited to light NREM sleep (stages N1 and N2) [115]. Much less commonly, they will be reactivated during REM sleep as well. Only tardive dyskinesias and primary palatal tremor may show complete cessation of movements during sleep.

The degree of persistence of various abnormal daytime movements into sleep varies greatly (Table 39.1). In one of the most informative studies on the topic, performed using EMG, accelerometry, and split screen video recording. Fish et al. [85] examined the relationship of motor activity not only to conventional sleep staging, but also to epochs with transitions (to lighter or deeper sleep stages or to wakefulness). They also monitored the 2-second periods before onset of dyskinesias in patients with Parkinson’s disease, Huntington’s disease, Tourette syndrome, and torsion dystonia (both primary generalized and secondary) and scored them for presence of arousals, REMs, sleep spindles, and slow waves. They compared these dyskinesias to normal movements both in patients and in normal subjects. Forty-one of 43 patients had characteristic movements that persisted in sleep. In every disorder, both normal movements and dyskinesias followed the same general plan: most common in wakeful epochs followed by lightening, in stage N1 sleep, REM sleep, then stage N2 sleep, with no movements in SWS. Only Tourette patients had dyskinesias during transition from wake to sleep. The 2-second period before both normal and abnormal movements showed arousals most commonly, followed by REMs, with spindles and slow waves rarely. These results support prior speculation [116] that both dyskinesias and normal movements are likely to be modulated by sleep in a similar fashion. This may be due either to the general suppression of centers for both normal and dyskinetic movements or suppression of some common descending path, such as the pyramidal tract.

It should be noted that all of the abnormal motor activities described in the experiment above, which tended to become attenuated and repressed by sleep, are thought to be generated in higher motor centers, most of them located above the brain stem. On the other hand, movement disorders associated with abnormalities of the lower motor centers, specifically the brain stem and spinal cord, have a greater tendency to persist during sleep [117]. Perhaps the best example is acquired (rather than primary) palatal myoclonus or palatal tremor (see below), which persists in sleep, although the frequency or persistence of these movements may vary with sleep stages [118]. In addition, spinal myoclonus will often persist during sleep [119]. Similar persistence may be seen in hemifacial spasm [120, 121], which is thought to involve damage either in the brain stem facial nucleus or in the peripheral nerve (cross talk due to ephaptic transmission) or both. Also, fasciculations due to damage to the lower motor neuron, whose generator lies at the spinal cord level, may persist in sleep [122].

In addition to the impact of the movements in sleep themselves, many disorders in which abnormal movements are a prominent feature impact sleep in other ways, including by causing changes in sleep architecture, mood, and level of daytime alertness, and medications treating the primary condition may equally affect the above domains, decreasing the patient’s quality of life. These aspects are discussed below.

i. Sleep-Associated Problems of the Hypokinetic Disorders

Parkinson’s Disease

Sleep impairment is a cardinal feature of Parkinson’s disease. The original quotes from James Parkinson are worthy of note [123].

But as the Malady proceeds …. (P.6)

In this stage (stooped posture with “unwillingly a running pace”… most likely stage 3), the sleep becomes much disturbed. The tremulus motion of the limbs occur during sleep and augment until they awaken the patient, and frequently with much agitation and alarm. (P.7)

…. and at the last (advanced bedridden stage), constant sleepiness, with slight delirium, and other marks of extreme exhaustion, announce the wished-for release. (P.9)

A spectrum of sleep dysfunction occurs in Parkinson’s disease (see Fig. 39.6). Many studies have confirmed the fact that sleep is a major issue for patients with Parkinson’s disease and their quality of life [124, 125]. Sleep disturbances also appear to impact cognition in Parkinson’s disease, although the mechanisms and exact relationships remain unclear [126]. Many studies have shown sleep architectural disturbances in Parkinson’s disease, including shorter total sleep time, lower sleep efficiency, and increased REM latency [127]. There have been conflicting reports on whether dopaminergic therapy improves sleep architectural changes [128, 129]. Early reports do seem to suggest that deep brain stimulation in Parkinson’s disease appears to improve NREM sleep, abnormalities of REM sleep, and improve daytime sleepiness [130]. Compared with controls, there appears to be disruption of a number of circadian processes in parkinson’s disease; including a sustained elevation of serum cortisol levels, reduced circulating melatonin levels, and altered peripheral clock gene expressions, and sleep disturbances, at least in part, may be related to this mechanism [131]; however, recently published data also suggest that dopaminergic therapy, while increasing melatonin levels (thus theoretically promoting sleep), may also show a delayed sleep onset relative to dim light melatonin onset (DLMO), suggesting that it may uncouple circadian and sleep regulation [132]. In recent years, several questionnaires have been developed and validated that can be used to assess sleep-related problems in Parkinson’s disease, and liberal use of these instruments helps detect sleep issues [133–137]. Poor sleep in patients with Parkinsons’s disease is multifactorial but can broadly be divided into being due to motor and non-motor symptoms.

Fig. 39.6

Schematic diagram showing the spectrum of sleep dysfunction in Parkinson’s Disease

• Motor Symptoms Impacting Sleep

The persistence of the parkinonian tremor into sleep, and its reoccurrence during periods of sleep–wake transition, is a major cause of sleep fragmentation. Parkinsonian tremor decreases in amplitude and duration in early NREM sleep and may lose its alternating aspects and is rarely seen in Stage N3 and often disappears in REM sleep [138]. Bradykinesia often causes trouble turning in bed, which is an additional source of discomfort and may contribute to poor sleep quality.

On the other hand, sleep can also influence these motor symptoms in a positive way; in some patients, sleep can reduce Parkinsonian disability and alleviate symptoms [139–141], perhaps due to the circadian peak of dopamine in the morning [142]. Sleep benefit may last from 30 min to three hours. This may be particularly true of patients suffering from early onset Parkinsonism such as that due to the most common recessive Parkin (PARK2) mutation. Sleep benefit is less consistent in those with the recessive Pink1 (PARK6) mutation [143, 144]. Sleep benefit in this group of early onset patients is often associated with some degree of dystonia.

A major parasomnia that causes motor dysfunction occurring in patients with Parkinson’s disease is RBD, which can disrupt sleep and be potentially injurious to the patient and the bedpartner. RBD may precede the diagnosis of Parkinson’s disease or other extrapyramidal disorders by several decades. This association may be due to the Parkinsonian degeneration affecting brain areas and systems responsible for sleep–wake regulation [145]. A recent scheme postulates that the synuclein pathology of Parkinson’s disease (Lewy Body pathology) ascends from the brain stem to the basal ganglia and finally to the cortex [146, 147], with early involvement of sleep regulatory nuclei before development of motor symptoms [148]. The combination of RBD with olfactory dysfunction may be a strong predictor for later development of Parkinson’s disease [149]. In general, RBD patients show subtle motor, cognitive, autonomic, olfactory, and visual changes that are associated with Parkinsons’s disease [150, 151], as well as brain perfusion changes determined by single photon emission computerized tomography (SPECT) imaging [152]. A recent finding is that of markedly reduced cardiac I-metaiodobenzylguanidine (MiBG) uptake, consistent with the loss of sympathetic terminals, in idiopathic RBD similar to the deficit seen in Parkinson’s disease [153]. One recent study suggests that RBD occurs primarily in Parkinson’s disease patients who have the non-tremor form [154]. The authors also suggest that RBD will not precede early onset Parkinson’s disease, an observation supported in cases of PARK6 (PINK1) early-onset familial PD [155]. RBD later in the course of Parkinson’s disease may be associated with additional complications such as hallucinations [156, 157], which may represent REM intrusions and cognitive decline [158]. A detailed analysis of movements in five patients with Parkinson’s disease and RBD found that they had many more movements in sleep than controls, but that most movements were brief and restricted in scope [159]; 3.6 % of all movements were violent while 10.5 % involved vocalizations. It has also been proposed that RBD-related movements may show “normalization” of motor control in Parkinson’s disease, with reduction in bradykinesia and vocal hypomimia [160]. Diagnosis and management of RBD is discussed in detail in Chap. 49 and 50.

There has been much research into the association between Parkinsons’s disease on the one hand, and RLS and PLMS on the other. Recent studies have suggested that the prevelance of RLS is increased in patients with Parkinson’s disease [161–163], including association in one family with a parkin mutation [164]. But this association is only poorly understood [165–167]. Some studies report that most RLS symptoms develop after onset of Parkinson’s disease and initiation of dopaminergic treatment. This suggests that RLS may be provoked by such treatment rather than the disease itself; indeed, such treatment has been shown to induce RLS through a process of “augmentation,” even in those who do not have RLS [168]. In one communication, subthalamic deep brain stimulation was reported to have induced RLS [169], but this may have been due to decreased medication doses. Similarly, some studies have found increased PLMS in patients with Parkinson’s disease [170, 171], but one small study of de novo patients found no such elevation [172]. Thus, while it remains unclear if RLS and PLMS truly occur more frequently in patients with Parkinson’s disease outside of dopaminergic treatment, when they do occur, they represent another factor that fragments sleep. Both RLS and PLMS are discussed in greater detail in Chap. 40.

• Nonmotor Symptoms Impacting Sleep

Excessive Daytime Sleepiness and Irresistable Sleep Attacks

Excessive daytime sleepiness (EDS) is a common symptoms in Parkinson’s disease and has been found to be more frequent in patients than in controls (15.5 % vs. 1 %) [173]. More advanced disease, higher frequency of cognitive decline and co-occurrence of depressive symptoms, more hallucinations,and longer time on levodopa are predictive factors [174]. This daytime somnolence can exist even without evidence for or complaints of severely disrupted sleep [175, 176]. A related condition is the occurrence of Parkinson’s disease-related “irresistable sleep attacks,” sudden episodes of sleep that appear without warning on a background of normal alertness [177]. These attacks are much rarer than pervasive daytime sleepiness that the patient is aware of [178, 179]. Genetic variants of the preprohypocretin [180] and dopamine D2-receptor genes [181] have been associated with predisposition to sleep attacks. Some patients with Parkinson’s disease have a narcoleptic phenotype, including a finding of sleep onset REM periods (SOREMPs) [182, 183]. This can be seen as 2 or more SOREMPs on multiple sleep latency test (MSLT; 4 or 5 naps scheduled during the day at 2 h intervals). Consistent with narcolepsy, hypocretin neurons are progressively lost with more severe Parkinson’s disease [184], most likely due to Lewy Body degeneration. The picture in Parkinson’s disease, however, is often more complex, while cataplexy, a key finding in narcolepsy, is generally absent [185]. Additionally, both EDS and irresistable sleep attacks can be caused by dopaminergic medication, especially at higher doses; non-ergot agonists specifically, may be more likely to cause somnolence. Treatment of daytime sleepiness may include the use of stimulants. Some recent studies have supported the use of modafinil as a relatively well tolerated stimulant [186–188], but one double-blind placebo-controlled study failed to support efficacy [189].

Sleep-Disordered Breathing

Respiratory disturbances are common in Parkinson’s disease and other related neurodegenerative disorders due to changes in upper airway function or disturbed central regulation of breathing. This may be due to Lewy Body deposition leading to cell loss. Altered upper airway function may be based on weakness of respiratory and upper airway muscles or on altered muscle tone and coordination. The prevelance of SDB in Parkinson’s disease has been demonstrated to be indepenedent of the degree of severity of motor and non-motor symptoms [190], but male gender and greater duration of illness are predictors [128]. Patients with Parkinson’s disease may have stridor or laryngeal spasm associated with off-states or dystonic episodes [191], although this is more common in multiple system atrophy (MSA, see below). Abnormal vocal cord function with regular rhythmic movements or irregular jerky movements in the glottic area may also produce changes of airflow and contribute to intermittent airway closure [192]. Similar activity persisting during sleep can lead to OSA or upper airway resistance syndrome [193]. Snoring has occurred in the majority of subjects in some series [194]. It has not been definitely established that the prevalance of respiratory dysfunction during sleep in patients with Parkinson’s disease as a whole is any higher than in healthy elderly persons [195, 196]. However, in some studies sleep-disordered breathing was more frequent and occurred in up to 50 % of patients with Parkinson’s disease [197]. Trouble turning in bed may be a contributory factor in those patients with positional OSA [198]. Patients with parkinsonism and autonomic impairment more often develop sleep apnea and related respiratory abnormalities, including central and obstructive apneas and nocturnal hypoventilation. In the presence of sleep apnea, patients with autonomic impairment are probably more likely than other patients to have nocturnal cardiac arrhythmias. An interesting recent report suggested that treatment with dopaminergic agonists predisposed patients with Parkinson’s disease to central sleep apnea [199] but more research on the subject is clearly needed.

Insomnia

Insomnia, both sleep initiation and sleep maintenance, is a common complaint among patients with Parkinson’s disease and is likely multifactorial [200]. Intrusion of motor symptoms into sleep (see above), depression and anxiety, concomitant sleep-disordered breathing and RLS, and in many cases dopaminergic treatment itself may contribute to insomnia in this population [201]. Other side-effects of medication include nightmares that may worsen underlying anxiety. Nevertheless, where motor symptoms seem to be the biggest impediment to quality sleep, appropriate therapy, including long-acting forms of dopaminergic medications to cover the night [202, 203], and, in some cases, deep brain stimulation [204, 205] may be helpful. Treatment of mood disorders with cognitive behavioral therapy and suitable antidepressants and anxiolytics is also recommended. SDB must be identified and adequately treated with continuous upper airway pressurization (CPAP) therapy.

Other Extrapyramidal Neurodegenerative Conditions

Alpha-synuceinopathies other than Parkinson’s disease (the “Parkinson’s plus” syndromes) have sleep disurbances as severe as those seen in Parkinson’s disease itself, most likely due to the same pathology of Lewy Body-related neurodegeneration. RBD occurs in most of them and is in fact more common in MSA than in Parkinson’s disease. Disturbances of sleep architecture, SDB, EDS, and insomnia are also frequent.

Progressive Supranuclear Palsy (PSP)

Patients with this condition have been reported to have severe sleep disruption with reduced total sleep, marked diminution in sleep spindles, reduced REM sleep time with abnormal REMs, disordered sleep architecture, and frequent awakenings [206–212]. RBD is frequent [213, 214], although possibly less common than with Parkinson’s disease [215]. As with Parkinson’s disease, sleep disruption increases with severity of the motor abnormalities [208–210]. The greater sleep abnormalities of PSP compared to Parkinson’s disease may be due to the greater brain stem pathology, especially that in the pedunculopontine tegmentum, a region linked to control of REM sleep; one report also found that cerebrospinal fluid (CSF) hypocretin levels were lower in patient with PSP than in those with Parkinson’s disease [216].

Multiple System Atrophy (see also Chap. 41)

Among the extrapyramidal disorders, sleep disruption appears to be worst in MSA [217–221], and these patients may be especially sensitive to the hypersomnolence induced by dopaminergic therapy [222]. Patients with severe MSA may even lack normal circadian regulation of sleep [223]. As mentioned above, RBD is very common [224], although RBD is unlikely to be seen in pure autonomic failure and provides one means of differentiating the two conditions [225]. PLMS is prevalent as well [221]. A somewhat characteristic feature of MSA is the occurrence of atrophic paralysis of the laryngeal abductor [226], or sleep-related hyperactivity of the adductors, which has been described as dystonic [227], leading to a coarse, snoring-like sound, and laryngeal stridor [228]. In fact, stridor, a potentially life-threatening condition that may cause sudden death from respiratory arrest [229, 230], has been reported as the first or even only apparent sign of MSA [231, 232]. Milder cases may be managed with CPAP [233]; more severe cases require tracheostomy. In some cases, patients with MSA may have predominantly central sleep apnea [234].

ii. Sleep-Associated Problems of the Hyperkinetic Disorders

The hyperkinetic disorders are a diverse group characterized by excessive involuntary movement, often coupled with a deficiency of voluntary movement such as bradykinesia.

Chorea

Chorea consists of movements that occur in a flowing or irregular pattern and appear to migrate from one part of the body to another. They may be increased with action and typically are seen in the face and distal limbs [6].

The best-known cause of chorea is Huntington’s disease, an autosomal dominant disease with a known mutation of the IT15 gene located on the short arm of chromosome 4. The mutation in Huntington’s disease is the expansion of a CAG repeat in the DNA that leads to increased length of a polyglutamine tract in the protein product, now called huntingtin. Currently, research is directed at finding the function of huntingtin in the normal brain and the elucidation of the toxic effect of the mutated protein. Although huntingtin is widely distributed in the brain, the pathology of Huntington’s disease is more restricted. Patients also have prominent psychological symptoms, including depression, psychosis, and behavioral disorders. Onset is typically between the ages of 25 and 50, although it may occur even in the first decade or in late adult life. Progression is slow but relentless, with eventual debility, dementia, and inanition occurring in those with onset before old age.

Sleep disturbances in Huntington’s disease have been the focus of considerable research. Recently, investigators found that sleep disturbances may be the earliest manifestation of Huntington’s disease [235]. There appears to be no correlation between CAG repeat length and sleep disturbances [236]. Investigators have shown a variable persistence of chorea during sleep, with most chorea present in awakening, and in the lighter stages of NREM sleep (stages N1 and N2), similar to other dyskinesias [85]. One study reported an increase in overall sleep movements in Huntington’s disease [237]. There has also been some research into sleep architecture in Huntington’s disease, but the results have been inconclusive. Some reported deficits include prolonged sleep latency, excessive waking, decreased SWS and REM sleep, and decreased sleep efficiency, possibly correlating to caudate atrophy [238–240]. Reports of alterations in sleep spindles in Huntington’s disease have been inconsistent [241, 242]. It has been suggested that nocturnal agitation and sleep disruption in Huntington’s disease patients is secondary to anosognostic voluntary movements on arousals, rather than to RBD [236]. A recent study of 30 patients with Huntington’s disease showed that they, compared to controls, had shorter sleep duration, reduced sleep efficiency, increased arousals and awakenings, and higher PLMS index in both NREM and REM sleep, but were not at increased risk for RBD or SDB. Greater clinical disease severity predicted decreased REM sleep percentage and greater daytime sleepiness [243]. Other studies have confirmed that unlike patients with parkinsonism, patients with Huntington’s disease have not been found to have a significant number of sleep apneas contributing to impaired sleep [244].

Sleep has not been well-studied in other conditions with predominant chorea. Broughton et al. [245] reported that four patients with Sydenham’s chorea, which follows a streptococcal infection, had reactivation of their movements during REM sleep. Neuroacanthocytosis or chorea-acanthocytosis, is an often inherited movement disorder with chorea, tics, vocalizations, and self-mutilation together with frequent seizures, associated with elevated acanthocytes (spiked red cells) in blood smears [246–248]. Silvestri et al. [249, 250] reported that in this condition, abnormal movements persisted during sleep, but with decreased amplitude, duration, and frequency. Patients frequently vocalized during REM sleep. Sleep was fragmented and of poor quality. Two siblings with neuroacanthocytosis showed EEG slowing (predominantly delta) both while awake and during REM sleep [251], indicating abnormal cerebral function. RLS has been reported to occur in this condition [252].

Dystonia

Dystonia is a condition characterized by sustained distorted or twisting postures and contorting movements, often mixed with a variety of jerk-like or oscillatory movements [253]. Dystonia can be primary or secondary and can be of variable extent, focal, segmental, or generalized, depending on the area of involvement. Dystonia includes a number of different conditions, some of which, such as early onset torsion dystonia, have a single-gene basis. The protein for early-onset torsion dystonia, torsin A, has been found to bind adenosine triphosphate, but how it causes dystonia itself remains unresolved [254–258]. Not all idiopathic dystonia patients have been shown to have a genetic mutation, however, and there are many cases of secondary dystonia that do not appear to depend on common dystonia genes [259]. One problem in evaluating sleep complaints in dystonia is that the studies so far have often examined a fairly heterogeneous collection of patients with different distributions of dystonia and different etiologies.

Although they usually subside significantly, dystonic movements may persist during sleep at a reduced frequency and amplitude. They are maximally reduced during SWS and may be partially reactivated during REM sleep episodes [260]. In the study by Fish et al. [85] of dyskinetic movements, both primary and secondary dystonic patients followed the general pattern of more frequent dyskinetic movements during wakefulness, fewer movements in stage N1 sleep, only infrequent movements in stage N2, REM, and SWS, and no movements during epochs of deepening sleep. In a study including focal and segmental dystonias, Silvestri et al. [249] found that Meige’s syndrome (oromandibular dystonia), blepharospasm, and tonic foot syndrome all showed persistent abnormal activity during sleep, with reduced amplitude, duration, and frequency of EMG bursts. The greatest suppression was in SWS and REM sleep.

Inhibitory mechanisms are postulated to be defective in dystonia. This prompted Fish et al. [261] to study both primary and secondary dystonics to determine whether REM inhibition is intact. They found that all dystonics had normal chin EMG atonia. No patients had complex abnormal activity during REM sleep. In an attempt to analyze motor excitability, the authors successfully stimulated three normals and seven primary dystonics with a magnetic coil over the vertex to evoke a motor response in the fifth finger abductor, the abductor digiti minimi. Whereas response amplitudes were highly variable, dystonics, like controls, showed a decrease in the mean response during REM sleep relative to responses obtained before and after the sleep study in relaxed wakefulness. Latencies were prolonged on average in all groups. The findings of decreased amplitude and prolonged latency were consistent with REM motor inhibition. Occasional high-amplitude responses may have corresponded to periods of phasic excitation. These results indicate that, whatever may be the decreased inhibitory processes in dystonia, they do not involve the descending inhibitory pathways of REM sleep.

Studies of sleep in dystonia have not been systematic; studies have involved small numbers of patients on diverse medications, some of whom had prior thalamic surgery [262]. In these studies, sleep has been found to be inconsistently disrupted, with more severe fragmentation seen in more advanced cases [263]. A number of studies have reported the presence of exaggerated sleep spindles in dystonia [264, 265]. The major therapeutic effort in these patients is the attempt to reduce the dystonic movements. Successful therapy of the movements should also improve sleep.

It is not known that to what degree different forms of dystonia—early- versus late-onset, focal versus generalized—differ in their relationships to sleep, although one striking form of dystonia, variably called hereditary progressive dystonia with marked diurnal fluctuations (HPD), dopa-responsive dystonia (DRD), and the Segawa variant, often shows distinct circadian variability [266–268]. These patients typically present at a young age, often in the middle of the first decade, with postural dystonia, usually affecting one leg and sparing the trunk and neck. Thereafter, the dystonia spreads and parkinsonian signs, which are present at onset in a minority of patients, become more prominent. The condition is usually inherited in an autosomal dominant mode with a mutation in GTP cyclohydrolase I (GCHI) [269, 270]. A number of different mutations in GCHI have been described, but, less commonly, it seems that the condition can be inherited recessively with a mutation in tyrosine hydroxylase [271]. Some studies have found that even patients thought to have more typical idiopathic torsion dystonia may harbor a mutation in the GCHI gene [272]. A number of these patients may obtain significant symptomatic relief from sleep, similar to the sleep benefit seen in Parkinson’s disease, and therefore are minimally impaired early in the day, and even some dystonic patients unresponsive to l-dopa may have similar benefit from sleep. These patients do show abnormal movements in sleep. Segawa et al. [273] obtained movement counts from PSG with multiple EMG channels (8–12 surface recordings on trunk and limbs) and found that in DRD, there is a decrease in gross body movements in stage I sleep, an increase in stage II sleep, and a decrease in REM sleep. In contrast, localized twitch movements were depressed in all sleep stages, but followed the normal relative distribution between stages.

Patients with diurnal dystonia or the nocturnal sleep abnormalities of DRD are responsive to low doses of l-dopa, often as little as 50–200 mg per day with decarboxylase inhibitor [274]. Some patients can maintain a stable therapeutic effect with doses every other day. Patients with long-standing disease (24–45 years before treatment) may benefit as well as those with recent onset. DRD patients can use l-dopa without the development of the dyskinetic side effects that are so prominent in juvenile parkinsonism. A few patients may develop “wearing off” phenomena, the re-emergence of symptoms several hours after an oral dose of l-dopa. Older family members may present with a “parkinsonian picture,” but still show the same persistent, positive response to l-dopa. This finding is consistent with the idea that a single underlying disease has different manifestations that vary with age, dystonia being prominent in early and late parkinsonism [275–278].

With fluorodopa positron emission technology (PET) scanning, it has been shown in a number of families that patients with DRD have normal to modestly reduced striatal uptake of fluorodopa, including those who present with parkinsonian features later in life [276]. Because of this finding, it can be concluded that these patients have relatively intact dopamine uptake, decarboxylation, and storage systems in the striatum. The genetic abnormalities so far uncovered are involved with the dopamine synthetic system. It has also been speculated that the diurnal fluctuations that characterize DRD may be due to the circadian variation in dopamine production, with greater synthetic activity possible at night. One study found that acute dystonia secondary to neuroleptic medication also shows a circadian pattern [279], with maximal dystonia present between 12:00 noon and 11:00 PM. This could not be accounted for by sleep, fatigue, or time since the last dose of medication (in this case, injections twice daily). Some of this circadian variability may be accounted for by circadian variations in the dopamine system, which seem to show the least activity in the evening hours with maximal activity in the morning [280].

Nocturnal Paroxysmal Dystonia

Nocturnal Paroxysmal Dystonia (NPD) was first described by Lugaresi’s group as a condition which might be considered analogous to diurnal paroxysmal movement disorders [281]. Although it is now established that this and related conditions are variants of frontal lobe epilepsy, their atypical presentation often makes the diagnosis quite challenging. The characteristically short lasting attacks of NPD begin with arousal, including an abrupt autonomic activation that can include substantial tachycardia, followed by dystonic choreo-athetoid or ballismic movements and large-scale semi-purposive movements of all limbs. Vocalizations are common. The attacks are quite diverse if considered between patients but appear to be stereotyped in a single patient. Attacks are brief, last about a minute (range 15 s to 2 min for typical attacks) and may be vaguely remembered. Neither tongue biting nor urinary incontinence is common and tend to resolve without a significant period of post-ictal confusion, which is one of the reasons that it was not appreciated early on that at least the brief attacks are a form of epilepsy. In some patients, the attacks are decidedly unilateral The epileptiform nature of these attacks, together with two other conditions, paroxysmal arousals (in which patients awake abruptly from NREM sleep, perhaps with a start or cry, and have fleeting dyskinetic movements, then fall back to sleep) [282] and episodic nocturnal wanderings (attacks of sudden motor activity, including violent ambulation, loud vocalizations, and a variety of forceful gestures commonly occuring in stage 2 NREM sleep [283–285] is now clearly established [286, 287]. Therefore, the diagnosis and treatment of these disorders is discussed in detail in Chap. 44.

However, it is worth mentioning that there have been reports in the literature of attacks similar to NPD that did not respond to antiepileptic treatment. In the original description, two cases had longer duration (2–50 min) attacks, with no epileptic associations [281]. In one case, a patient afflicted with such attacks for 20 years developed Huntington’s disease. There are a number of more recently described disorders of at least uncertain etiology. Lugaresi’s et al. [288] described a periodic form of NPD which recurs every 30 s to 2 min with usually quite brief attacks (2–13 s in duration) and associated arousals which they called atypical periodic movements in sleep. While showing overlap with the short-lasting NPD, this condition was unresponsive to seizure medications, even though one patient in the original series had a vascular orbital frontal tumor on computerized tomography (CT) scanning and spikes on depth recording. Other such disorders include dystonic attacks provoked both by sleep and exercise [289], apnea-associated paroxysmal dyskinetic movements [290, 291], and post-traumatic nocturnal hemidystonia [292]. Thus, when the diagnosis is in doubt, it is prudent to order a PSG with an extended seizure montage as well as extra EMG channels to distinguish between a motor disorder and an epileptic phenomenon; in our laboratory, we employ a hybrid montage for this purpose (Table 5 from Chap. 18).

Myoclonus

The myoclonias [293] are a diverse group of conditions with abnormal movements generated at various levels of the neuraxis, from cortex (cortical reflex or epileptic myoclonus) to spinal cord (spinal or segmental myoclonus). The basic abnormal movement is a single, repeated, or periodic jerk, most typically abrupt and “lightning-like.” Most of the studies of myoclonus and sleep have focused on the persistence of myoclonic movements during sleep. Whether myoclonus persists in sleep or not appears to be related to the source of the discharge; elegant experiments have shown that myoclonus with a cortical source shows suppressed movements during sleep while (as in epilepsy) cortical discharges persisted, myoclonus of presumed subcortical origin is rapidly suppressed during sleep, and myoclonus of lower-level origins (spinal cord or secondary to peripheral damage) persists during sleep [294]. Thus, cortical and subcortical myoclonus tends to be attenuated by sleep, whereas myoclonus of peripheral origin, such as spinal myoclonus, and propriospinal myoclonus, shows persistence into sleep in varying degrees [295, 296]. Myoclonic jerks associated with startle disease also persist during sleep, although with diminished intensity [297].

Palatal Myoclonus (Palatal Tremor)

Palatal myoclonus (more accurately described as palatal tremor) is characterized by rhythmic movements of the soft palate and pharynx at a rate of 1–3 Hz. It is sometime associated with rhythmic ocular, buccal, lingual, laryngeal and diaphragmatic movements, and occasionally also movements of the upper limbs [298]. Two types have been described as follows: a primary or essential type (idiopathic) due to contraction of the tensor veli palatini muscle presenting with a clicking noise in one or both ears, and an acquired or secondary due to contraction of the levator veli palatini muscle [299]. When acquired, it is usually secondary to brain stem damage within Mollaret’s triangle (dentatorubroolivary pathways, with damage most common in the central tegmental tract which runs from the region of the red nucleus to the ipsilateral olive). While primary palatal tremor may be completely abolished by sleep, electrophysiologic studies in a small number of patients demonstrated that palatal contractions persist during sleep, albeit with shifts in amplitude and frequency or even altered rhythmicity [300–302]. The range of such cyclic motor dyskinesias may be broader than currently known: a similar tongue movement was reported to persist largely unchanged in sleep [303]. The finding of persistent rhythmicity suggests a relatively autonomous oscillator consistent with the idea that these segmental myoclonias may represent release of a primitive rhythmic center. In contrast to other forms of myoclonus, these dyskinesias appear to arise at a segmental level and to be associated with decreased motor control from higher centers. This dissociation may explain their resistance to modulation by descending inhibitory influences during sleep. The dyskinesias are not completely removed from higher motor centers or the periphery, however, because they may disappear in sleep, change with state, and be influenced by attention [304–306]. In one interesting case, palatal tremor was associated with time-locked respiration, suggesting a coupling of these two rhythms [307].

Palatal tremor is generally refractory to treatment. There are reports of occasional response to anticholinergics, botulinum toxin injections, baclofen, valproic acid, lamotrigine, tetrabenazine, and carbamazepine [308].

Tics

Tics are typically brisk, stereotyped, complex, often repetitive movements [309]. Usually, any given patient has a somewhat limited repertoire of movements that may change over a period of months to years. The prototypical tic disorder is Gilles de la Tourette’s syndrome, a condition involving multiple motor tics with vocalizations that usually begins in childhood or adolescence but may subside in later adult life [310]. Tics may be associated with a sensory penumbra and an urge to move. Tourette’s patients also have a number of commonly-associated behavioral abnormalities, especially obsessive-compulsive disorder [311, 312]. Of all tic disorders, sleep disorders have been best studied in patients with Tourette’s syndrome.

Tics in Tourette’s syndrome have been found to persist during sleep in most cases, mostly in stages I and II of NREM sleep, with fewer during SWS or REM sleep [313, 314]. Also observed is increased frequency of disorders of arousal (e.g., somnambulism and pavor nocturnus) and parasomnias in general, as well as poor quality and fragmented sleep [315, 316] in children with Torette’s syndrome. Bodily movements in general are increased in Tourette’s syndrome; Hashimoto et al. [317] found that both twitch-like and gross body movements were increased over controls during all stages of sleep, with total movements in tic patients markedly increased during REM sleep. Those authors did not attempt to analyze such movements in detail, so it is not clear what fraction of them were actual tics. In one study, patients were monitored after successful treatment of their movements with tetrabenazine, and it was found that sleep also improved [318].

Hemifacial Spasm

Hemifacial spasm consists of intermittent contraction of one side of the face that can be repetitive and jerk-like or sustained. It is believed to arise from irritation of facial nerve or nucleus. Both central and peripheral (ephaptic transmission between adjacent nerve fibers without synapses) factors are responsible for the spasms. EMG recording shows highly synchronous discharges in upper and lower facial muscles. Montagna et al. [319] studied 16 patients, recording from upper and lower facial muscles during sleep studies. In most patients, the dyskinesias decreased during sleep, being approximately 80 % less frequent in SWS and REM sleep. One patient showed almost no change in the prevalence of spasms. Current therapy for hemifacial spasm includes medications such as carbamazepine, botulinum toxin injection into the affected muscles, or varied surgical treatments, such as vascular decompression of the facial nerve. Combined treatment with pregabalin and botulinum toxin injections has been reported [308].

Other Hyperkinetic Disorders

In hemiballismus, there are proximal flinging movements of one side of the body, which may be of a violent nature, associated with damage to the contralateral subthalamic nucleus [320]. In most cases, hemiballismus is a transient phenomenon after local injury to the subthalamus, usually ischemic, although it may be transformed into a chronic choreiform disorder. It was initially thought that the movements totally subsided in sleep. Askenasy [321], however, reported a patient whose movements persisted in sleep, and Silvestri et al. [115] found that the movements were present during stages N1 and N2, as well as during REM sleep, although diminished in intensity and frequency. Puca et al. [322] reported one case in which spindle density and amplitude were greater ipsilateral to the damaged subthalamic nucleus. There was also disrupted sleep, with prolonged latency and an absence of both SWS and REM sleep. Successful treatment with haloperidol improved the sleep and decreased the spindling.

In athetoid cerebral palsy, abnormalities of REM sleep have been noted. Hayashi et al. [323] reported on a group of severe adolescent and young adult patients. The significant motor abnormalities were associated with REM sleep: three patients had decreased numbers of REM, two had increased chin muscle tone, and seven had reduced numbers of muscular twitches. The authors suggest this may be related to brain stem pathology in these birth-injured patients. One family with five generations affected by paroxysmal dystonic choreoathetosis with dominant transmission was found to show substantial benefit from even brief periods of sleep [324]. In a Serbian family with mutations in the Myofibrillogenesis regulator 1 gene, sleep was reported to be the most effective means of terminating attacks [325].

Sleep-Associated Problems of the Ataxic Disorders

Relatively little research has been done on sleep disturbances in the ataxic disorders. Today, a large number of different genetically based variants of spinocerebellar ataxia have been described and some of these have been examined with respect to sleep. Patients with Machado–Joseph disease (Spinocerebellar atrophy type 3, SCA3) may have both RLS and RBD as common sleep-related problems. Patients with SCA2 can have reduced REM-sleep atonia. One report suggests increased PLMS and RLS in SCA6. SCA6 patients also have impaired subjective sleep quality and tend to have greater daytime sleepiness. It seems likely that the paucity of associations reported to date is more due to the lack of studies than the absence of sleep problems in these disorders [326–331].

B. Motor Disorders Exclusive to Sleep

i. Failure of motor control while resting in bed trying to get to sleep.

This category includes RLS and PLMW, which are discussed in Chap. 40.

ii. Failure of motor control immediately before and at sleep onset

Since normal individuals enter the sleep cycle through NREM sleep, these disorders can also be considered failures of motor control in NREM sleep. Many of these are of unclear clinical significance and require no treatment.

Physiological Hypnic Myoclonus

The term physiological hypnic myoclonus (PHM) was first coined by De Lisi [332] to describe brief asynchronous, asymmetric, and aperiodic muscle twitches during sleep in all body muscles of man and domestic animals resembling fasciculations seen prominently in face and distal body parts (e.g., face, lips, fingers and toes). PHM is also known as physiological fragmentary hypnic myoclonus and is seen prominently in babies and infants. Quantitative study by Dagnino et al. [333] and Montagna et al. [334] in 1988 showed the maximum occurrence of these twitches in stage N1 and REM sleep, decreasing progressively in stages N2 and N3. Presence of PHM also during relaxed wakefulness challenges the term hypnic Myoclonus [335, 336]; however, it should be noted that propriospinal myoclonus at sleep onset and intensified hypnic jerks in many patients [336] are present in relaxed wakefulness before sleep onset. The origin of PHM remains controversial. Facilitatory reticulospinal tract, pontine tegmentum, and corticospinal tract [337, 338] have all been suggested as the generator of PHM. These movements are physiologic without disrupting sleep architecture and require no treatment.

Hypnic jerks including intensified hypnic jerks

Hypnic jerks or “sleep starts” are sudden, brief contractions of the body that occur at sleep onset and are due to excitation of motor centers. They are physiological and occur in up to 70 % of the population at some point in their adult lives. They are often accompanied by a sensation of falling [339]. The movement itself is an abrupt, myoclonic flexion movement, generalized or partial, often asymmetric, which may be accompanied by a sensation or an illusion of falling. Unless very frequent (which does occur rarely) [336, 340] this is a benign movement which has little effect on sleep and carries no negative prognosis. When it occurs, it is usually a single event, which causes a brief arousal. EMG records show relatively brief EMG complexes (<250 ms in duration) that may be simultaneous or sequential in various muscles. The earliest mention of this phenomenon is credited to Mitchell [341], who described insomnia occurring as a result of hypnic jerks in 1890. Oswald [339] first described the EEG correlates of hypnic jerks. In 1965, Gastaut and Broughton [342, 343] performed the first polygraphic study of hypnic jerks. It was not until 1988 that Broughton [340] coined the term “intensified hypnic jerks” to describe the clinical phenomenon of sleep onset insomnia caused by accentuated and disruptive hypnic jerks occurring at sleep onset. More recently, Chokroverty et al. [336] performed a polysomnographic and polymyographic analysis of ten patients with intensified hypnic jerks and identified four patterns of propagation: synchronous and symmetrical patterned muscle bursts between the two sides and agonist–antagonist muscles similar to those noted in audiogenic startle reflex; reticular reflex myoclonus; dystonic myoclonus; and pyramidal myoclonus with rostrocaudal propagation of muscle bursts.

Hypnagogic Foot Tremor and Alternating Leg Muscle Activation

Hypnagogic foot tremor (HFT) (Fig. 39.7) and ALMA (Fig. 39.8) rarely come to clinical attention, being discovered as incidental findings on PSG. Both occur during lighter sleep and in transitional states into and out of sleep. ALMA has also been documented in wakefulness, all stages of NREM and also, though less frequently, in REM sleep in patients with a variety of sleep disorders [344]. Another feature of ALMA is its occurrence, in addition to the traditional tibialis anterior EMG, in gastrocnemius and sometimes in quadriceps muscles alternating between two sides. Because variant patterns are reported in each, and there is at least some plausible degree of overlap between HFT and ALMA (and also PLMS) further investigation may be required to ascertain whether they are distinct or merely variant conditions. It has also been suggested that both may be variants of RMD (see below). HFT is defined by the AASM Manual for the Scoring of Sleep and Associated Events as rhythmic contractions of foot and leg occurring during sleep onset generally bilaterally but asynchronously at a frequency of 0.5–4 Hz [5] and was first described by Broughton [340]. Wichniak et al. [345] later performed polysomnography on 375 consecutive subjects and found HFT (which they called “rhythmic feet movements while falling asleep” and described as rhythmic, oscillating movements of the whole foot or toes) in 7.5 %. Per the AASM Manual for the Scoring of Sleep and Associated Events, ALMA consists of EMG bursts that occur alternately in each leg in a rhythmic pattern of 0.5–3 Hz and was first described by Chervin’s et al. [346], who found it in just over 1 % of reviewed PSGs; most of those showing the phenomena were taking anti-depressants. The duration of an individual movement varies between 100 and 1000 ms. For both HFA and ALMA, at least 4 movements must be present in a row to make the diagnosis [5]. ALMA requires the presence of alternating activity and has been suggested to be an equivalent of a locomotor rhythm [347]. Diagnosis of either requires a PSG recording. Convincing evidence of any definite clinical consequence of these movements is yet to be presented. In one patient, pramipexole-reduced ALMA and improved sleep [347], together with a reduction of associated CAP. The clinical significance of both HFT and ALMA remains undetermined requiring no treatment.

Fig. 39.7

Hypnagogic foot termor. A 30-s epoch of relaxed wakefulness from the polysomnogram of a 69-year-old man affected by snoring and nonrestorative sleep. Note the occurrence of a series of rapid tibialis anterior activations longer than 30 s, with single burst durations of 200–300 ms. Top three channels, electroencephalography. ROC and LOC: electrooculogram channels. ECG, electrocardiogram. LtTib and RtTib, tibialis anterior electromyography channels. (Reproduced with permission from Ref. [114])

Fig. 39.8

Alternating leg muscle activation. A 10-s epoch of relaxed wakefulness transitioning to N1 sleep from the polysomnogram of a 32-year-old man with chronic insomnia and no comorbid history of restless legs syndrome, affected by snoring and nonrestorative sleep. Note the occurrence of a series of alternating tibialis anterior activations longer than 10 s, with single burst durations of 200–300 ms. Top three channels, electroencephalography. ROC and LOC: electrooculogram channels. ECG, electrocardiogram. LtTib and RtTib, tibialis anterior electromyography channels. (Reproduced with permission from Ref. [114])

Rhythmic Movement Disorder

RMD (Fig. 39.9) is characterized by repetitive, stereotyped, rhythmic movements involving large muscle groups, occurring predominantly during sleep onset or during sleep–wake transitions, at a frequency of 0.5–2 Hz. It can take many forms, including head banging (“jactatio capitis nocturna”), body rocking, body rolling, and leg rolling. Reports of events matching the current description of RMD have been abundant in the literature [348, 349] and even reported as early as in 1880 [350] although the term jactatio capitis nocturna was first used by Zapert [351]. According to the AASM Manual for the Scoring of Sleep and Associated Events, in addition to the above frequency criterion, the minimum number of individual movements to make a cluster of rhythmic movements is 4 movements and the minimum amplitude of an individual rhythmic burst must be at least 2 times the background EMG activity [5].

Fig. 39.9

Rhythmic movement disorder. Polysomnographic recording of a rhythmic movement disorder (RMD) episode. RMD typically consists of repetitive stereotyped and rhythmic motor behaviors, such as head banging, body rolling, and body rocking. These movements generally occur at sleep–wake transitions and after arousals from sleep but may also occur in wakefulness, and rarely during REM sleep. These episodes may also occur at the termination of a respiratory event in obstructive sleep apnea syndrome. (Reproduced with permission from Ref. [114])

RMD generally presents before 18 months of age and tends to occur immediately before sleep during relaxed wakefulness continuing into stage N1 and sometimes into stage N2. Rare case may show a REM predominance [352, 353]. Bouts of movements may be related to CAP [354]. While RMD is most common in pre-pubertal children, there are older children [355] and also adults [356–361] who will show persistent or emergent rhythmic movement. Most older children with persistent disorder are usually suffering from organic brain dysfunction (cerebral palsy, autism or attention deficit disorder). In developmentally normal children, however, RMD is generally benign and the child usually outgrows the movements by the second or third year of life.

Because RMD is a benign phenomenon in itself, treatment is not always necessary. However, it may cause significant injury. In addition, RMD may be secondary to frequent arousals from another condition, commonly OSA [362]. RMD may present in a rather dramatic fashion, and so may need to be distinguished from tremor or segmental myoclonias, as well as RBD or nocturnal seizures. Given this, evaluation of a patient with suspected RMD requires careful clinical history and physical examination, viewing of a video recording of the events if possible, and occasionally PSG. PSG is always recommended for most cases of RMD in patients in whom a primary sleep disorder is being considered and should be performed with an extended seizure montage if nocturnal epilepsy is suspected. In case of primary RMD, behavioral therapy and in severe cases with potential for inflicting injury clonazepam (0.5–1 mg nightly) , imipramine (10 mg at night) or melatonin [363] maybe helpful [308]. Protective measures should be used in cases with violent movements.

Propriospinal Myoclonus at Sleep Onset

Propriospinal myoclonus is a form of spinal myoclonus in which the excitatory impulses are believed to travel through relatively slow-conducting intersegmental propriospinal pathways [364–366]. However, propriospinal myoclonus at sleep onset was described fairly recently by Montagna et al. [367, 368] who performed polygraphic studies that showed that the myoclonic activity began in spinally innervated muscles, propagating at low speed to rostral and caudal muscular segments, and hypothesized that a spinal generator may be facilitated by changes in supraspinal control related to vigilance levels. They identified it as a potential cause of severe anxiety and insomnia. The myoclonic movements typically involve the trunk with possible extension into the limbs. In a recently described form of this myoclonus, the myoclonic jerks are only evident during relaxation or recumbency [369], especially when the patient is drowsy. Unlike PLMS, the movements are relatively easily abolished by even light sleep. They may, however, produce a substantial difficulty with sleep induction and can therefore be a cause of significant insomnia. Cases have been described that are associated with RLS [370] and an important consideration in the differential diagnosis of propriospinal myoclonus at sleep onset is the myoclonic form of PLMW seen while sitting or lying in patients with RLS [371]. One case of propriospinal myoclonus that occurred during sleep was reported after a thoracic spine fracture that progressed to “myoclonic status” and respiratory failure [372]. The treatment of this condition is challenging and some cases respond to clonazepam, zonisamide and other antiepileptic drugs used in the classic propriospinal myoclonus [308].

iii. Failure of Motor Control During NREM Sleep

This includes PLMS (discussed in Chap. 40) and the disorders of partial arousal, such as sleep terrors, confusional arousals and sleepwalking/Parasomnias (discussed in Chap. 50).

iv. Failure of Motor Control During REM Sleep

This includes RBD (discussed in Chaps. 49 and 50). The scoring criteria for REM without atonia, an essential neurophysiological component in the diagnosis of RBD, is provided in Table 39.3.

v. Failure of Motor Control in both NREM and REM Sleep

Benign Sleep Myoclonus of Infancy

This is a transient, sometimes familial condition that begins soon after birth and resolves within months [373–377]. It may, however, persist for up to a year after birth, hence the recent change in nomenclature from “benign neonatal sleep myoclonus” to “benign sleep myoclonus of infancy.” The myoclonic jerks are brief, asynchronous, and repetitive, involving primarily the distal limbs, especially the arms, but also the trunk; the jerks are often generalized. The jerks occur during all stages of sleep, with most occurring in NREM sleep, and typically do not arouse or wake the infant [378]; waking the child will cause them to cease promptly. The movements do not occur continually in sleep and, when not present in sleep, they may be precipitated by rocking the infant or by gentle restraint during sleep [379]. The exact pathophysiology is unknown, but these movements most likely represent an exaggeration of the normally greater sleep-related movements in infants [380]. Although completely benign, self-limiting and with no long-term sequelae, the clinician is often called upon to reassure frantic parents that their baby is not having seizures, which it may superficially resemble [381–383]. When in doubt, EEG or PSG may help alleviate some of the concern.

Sleep Bruxism

While bruxism or teeth grinding can occur during the day, nocturnal bruxism (Fig. 39.10) is to be clearly differentiated from daytime bruxism. Nocturnal, or sleep bruxism, when frequent and intense enough, can interrupt sleep and cause significant dental wear [384]. It is associated with arousals and autonomic activation during sleep [385, 386]. SPECT studies show an asymmetry in D2 dopamine receptor binding in bruxism patients at the level of the basal ganglia compared to controls suggesting that dopaminergic cell dysfunction may play a role in the pathogenesis of bruxism [387]. Bruxism tends to decrease with age, although bruxers may also have increased movements during sleep in general, and may be more common in the supine position [388]. Bruxism may also be a sign of recurrent OSA-related arousals, and thus any patient with bruxism should be screened for possible sleep-disordered breathing.

Fig. 39.10

Sleep bruxism. A 20-year-old woman referred for nocturnal awakeings with vocalization and sleepwalking. Polysomnogram shows arousal from slow-wave sleep with subsequent rhythmic masticatory muscle activation and teeth grinding (note electromyographic [EMG] artifacts on electroencephalogram [EEG] channels) typical of sleep bruxism. Top 15 channels, EEG. Mylo EMG, mylohyoid EMG activity. R-EOG + L-EOG, right and left electrooculogram; ECG, electrocardiogram; THO-ABD, thoracoabdominal respiration. (Reproduced with permission from Ref. [114])

Bruxism may need to be distinguished from other dyskinetic movements which involve the jaws, including oromandibular dystonia and idiopathic myoclonus in the oromandibular region during sleep. Idiopathic myoclonus in the oromandibular region (e.g., faciomandibular myoclonus) during sleep is an apparently isolated, non-epileptic condition that occurs predominantly in stages 1 and 2 NREM sleep [389–391]. It consists of isolated or short runs of shock-like jaw movements with brief EMG bursts.

Sleep-related bruxism has been described in every stage of sleep. While highest level of activity occurs during stage N3 and wakefulness, no difference has been described with regard to percentages of the sleep stages [392]. A close association between sleep bruxism and REM sleep has also been described [393, 394]. In 2008, Manconi et al. [395] published an interesting case report of a patient with sleep bruxism and catathrenia (see below) occurring in a synchronized fashion. They hypothesized about the presence of a common trigger mechanism for both phenomena.

According to guidelines put forth by the new AASM Manual for the Scoring of Sleep and Associated Events [5], bruxism can be identified by either brief (phasic) EMG elevations of 0.25–2 s and sustained (tonic) EMG elevations of >2 s. These EMG elevations must be at least twice the amplitude of the background EMG. Phasic bruxism events must occur in a sequence of 3 or more and this sequence can be said to comprise a bruxism episode. At least 3 s of stable EMG must be present before a new episode of bruxism can be scored. Bruxism can be reliably scored by audio in combination with polysomnography by a minimum of 2 audible tooth grinding episodes/night of polysomnography in the absence of epilepsy. In addition to chin EMG, additional masseter electrodes may be placed at the discretion of the investigator or clinician for optimal detection of bruxism. An alternate term for the phasic type of bruxism is Rhythmic Masticatory Muscle Activity.

The treatment for bruxism has yet to be standardized; various modalities have been employed, including dopaminergic agents [396], anticonvulsants [397], or with botulinum toxin injections [398]; dental devices may also help [399, 400] but some studies suggest caution in their use [401].

Catathrenia (see also Chap. 41)

Catathrenia, or nocturnal groaning, is a relatively newly described entity characterized by loud expiratory vocalization, whose exact pitch and timber may vary from individual to individual but is fairly stereotyped in a given patient (see Fig. 41.12). Not strictly a disorder of motor control, it may rather represent a disorder of breathing in sleep and is classifies as such according to the ICSD-3, although this is disputed by some [402]. While far more frequent in REM sleep, it may also occur in NREM sleep and alternates with normal breathing. It was actually first described by Pevernagie et al. [403] but was first named by Vetrugno et al. [404]. The same group subsequently reported in 2007 [405] that the groaning was accompanied by disproportionately prolonged expiration causing reduced tidal volume and bradypnea without oxygen desaturation, and that patients experienced no additional symptoms after a mean follow up of 4.9 years. They speculated that catathrenia was due to persistence of a vestigial type of breathing pattern. In 2011, Ott et al. [406] performed laryngoscopy under deep sedation in a patient with catathrenia and found that while the glottis was open at inspiration, there was subtotal closure of the glottis at expiration, resulting in the characteristic groaning. The following year, Koo et al. [407] performed acoustic analysis of catathrenia and found that it had morphologic regularity, with two types of sound pitches (either a monotonous sinusoidal pattern or a sawtooth-shaped signal with higher fundamental frequency), as opposed to snoring which was distinct from catarthrenia and had an irregular signal. Several authors have reported the efficacy of CPAP in treating this benign but socially awkward condition [408–410]. The anatomical factors that predispose to catathrenia, namely broad upper airway, yet protrusive upper incisors and flat mandibular angles, have recently been described [411].

Excessive Fragmentary Myoclonus

Excessive Fragmentary Myoclonus (EFM) is essentially a variant PSG finding (Fig. 39.11), often found incidentally, that has yet to be demonstrated to be of clinical relevance. This condition may be another in which inadequate inhibitory drive fails to block descending activation from higher centers or it may represent a condition of excessive activation of higher centers during sleep. A neurophysiologic analysis by Vetrugno et al. [412] failed to disclose any cortical prepotential on EEG–EMG backaveraging suggesting a subcortical origin. The condition has been found in degenerative developmental disease (Niemann-Pick) [413] and as a consequence of brain stem lesions [414] but usually occurs in isolation [415]. Generally, EFM is not accompanied by gross visible movements; if movements are present at all they are small movements involving the corner of the mouth or small movements of the fingers or toes. In most cases no movement across a joint space occurs and the movements may resemble fasciculations, mere dimplings seen over the muscle associated with very brief EMG potentials (<50 ms). According to the AASM Manual for the Scoring of Sleep and Associated Events [5], EFM is present when at least 20 min of NREM sleep is recorded on PSG with the characteristic EMG pattern present (bursts typically <150 ms and of variable ampltude) and at least 5 EMG potentials per minute. Although classically described in the lighter stages of NREM sleep, they may occur in REM sleep, where the pattern resembles the normal phasic twitches seen in REM sleep, except they are more evenly spread throughout an individual epoch and not clustered as are phasic REM twitches. Some, [416] but not all [417] reports suggest that they are least common in slow-wave sleep. Given the lack of known clinical consequences, treatment is not required.

Fig. 39.11

Excessive fragmentary myoclonus. Poysomnogram epoch from the study of a 71-year-old man who presented a 2-year-history of continuous twitch-like movements of the arms and legs throughout the night, which did not wake him from sleep. Note the brief, asynchronous, asymmetric potentials in the limb electromyography (EMG) channels (RtTib and LtTib). Visually, they presented as brief twitch-like movements not causing movement of major joints. Excessive fragmenrtary myoclonus is considered a benign phenomenon with no clinical consequence. Top three channels, EEG. ROC-A1, LOC-A2, electrooculography channels. CHIN, chin EMG. ECG, electrocardiography. (Reproduced with permission from Ref. [114])

vi. Failure of motor control at sleep offset

Sleep paralysis

Transitions out of sleep may also be associated with sleep paralysis, a condition in which an individual is paralyzed while awakening from sleep. Weir Mitchell [341] is given credit for an early description of the condition in 1876 and he termed it “night palsy”. Adie [418], in the 1920s observed occurrence of sleep paralysis in narcolepsy patients and Wilson in 1928 [419] introduced the actual term. There are earlier descriptions in the Chinese, Indian, Persian and Greek cultures and mythologies, as well as in famous novels such as Herman Melville’s Moby-Dick (1851).

During episodes of sleep paralysis, breathing and eye movements are usually preserved. This condition is thought to represent a variety of REM sleep tonic motor inhibition [420]; recordings of the state can show REMs together with an electrophysiological pattern consistent with REM sleep [421]. Sleep paralysis is generally associated with arousal from a REM period (hypnopompic) or, less commonly progress into REM sleep from wake (hypnagogic) [422]. The latter would be very unusual in the normal course of events and more likely to occur with narcolepsy, though it may occur in many non-narcoleptic individuals, sometimes with a familial pattern. There are three forms of sleep paralysis, isolated or recurrent isolated sleep paralysis (physiological occurring mostly in adults up to 30–50 % of the population), familial sleep paralysis and sleep paralysis as part of narcolepsy. Several studies suggest that, at least in some populations, sleep paralysis may be quite common [423–425]. When it does occur in normal individuals, it is generally infrequent, but may cause significant anxiety, especially the first time that it occurs. A similar condition, nocturnal alternating hemiplegia of childhood, involves paralysis limited to one side while awakening from sleep [426]. This may be a variant of hemiplegic migraine, a complicated headache disorder with paralysis due to suppressed activity in certain brain regions.

Physiological sleep paralysis is generally brief, lasting for seconds to a few minutes, but sometimes may last longer, particularly recurrent isolated sleep paralysis. On occasions the episodes are accompanied by hypnagogic or hypnopompic hallucinations. The episodes may be triggered by sleep deprivation, stress, physical exertion or supine position. Isolated or recurrent sleep paralysis does not require any specific treatment other than reassurance, life-style changes, regularizing sleep–wake schedule but in severe cases causing anxiety and panic short-term treatment with selective serotonin reuptake inhibitors (SSRIs) or tricylic antidepressants may be beneficial.

Sleep Inertia

This is most likely not a disorder of motor control in sleep in the strictest sense, but is discussed here for conveneience. Sleep inertia, also known as sleep drunkenness is a transient physiologic state of hypovigilance, confusion, impaired cognitive and behavioral performance, and grogginess that immediately follows awakening from sleep [427]. Put simply, the subject is physiologically awake (body awake) but cognitively asleep (brain asleep). EEG of sleep inertia is characterized by a generalized decrease of high frequency beta-1 and beta-2 EEG power but an increase of delta power in the posterior scalp region concomitant with decreased frontal delta power [428]. This state can last from minutes up to four hours, most commonly about five minutes and rarely may exceed 30 min. Prior sleep deprivation, awakening from SWS and short naps may aggravate sleep inertia. It is also more intense when awakening from near the trough rather than the peak of the circadian core body temperature rhythm. Sleep disorders, particularly idiopathic hypersomnia, as well as narcolepsy-cataplexy syndrome and obstructive sleep apnea syndrome may be associated with prolonged sleep inertia. Bedrich Roth and collaborators were probably the first to describe idiopathic hypersomnia with sleep drunkenness in the 1950s [429]. One suggestion for the pathogenesis of sleep inertia is build-up of adenosine and this state can be reversed by caffeine acting through adenosine A2a receptors, thus explaining the well-known reinvigorating effects of an early morning cup of coffee.

Scoring Criteria for Sleep-Related Movements

These are summarized in Tables 39.2 and 39.3.

Table 39.2 American Academy of Sleep Medicine (AASM) polysomnography scoring criteria for sleep-related movements [5]

Table 39.3 Scoring criteria for REM without atonia

Methods for Studying Sleep-Related Movements

As with any branch of clinical medicine, there is no substitute to a well-taken history and thoroughly-conducted physical examination. Not only is this good clinical practice that ensures that the appropriate test is ordered, but in many cases where the underlying disorder is clearly benign, and reassurance and observation is desirable (such as hypnic jerks), may obviate the need for testing at all.

A. Accelerometry Based Testing

i. Actigraphy

Actigraphy is a validated, relatively cost-effective and convenient alternative to expensive, cumbersome in-laboratory procedures in the assessment of sleep–wake cycles and movements in sleep, in both clinical practice and research. Depending on the equipment and technique used, recordings can be made for many days or even months. In assessing sleep disorders, this extended recording can allow for the capture of rare events, overcoming the problem of variability which can limit the accuracy of more abbreviated studies, and repeated measurement of sleep in different conditions (evaluation of sleep patterns, disease progression or remission, therapeutic responses). In addition to the cost consideration, the small size, light weight and ease of use of most of these devices allows for its application in multiple settings; the activity monitors can be taken out of the laboratory, self-applied, and even transmitted by mail. They may be particularly useful in uncooperative patient groups with degenerative disease who would not tolerate a laboratory sleep study.

The limitations on activity monitoring result from the relatively non-specific results and the limited information monitored. All movement, even transmitted movement, is recorded. There is generally no information about cerebral state (EEG), eye movements (too small to be reflected in a limb monitor), or breathing. Therefore, they do not provide much useful information about physiological state and crucial information about exact sleep stages.

Typically, activity monitoring devices use accelerometry to quantify movement. Several small self-contained devices currently available on the market provide a direct assessment of the amount of activity or body movement at the point of the body where they are attached. These are all derived from the work of Colburn and Smith who produced the first of these meters and documented the methods for others to use [431]. Virtually all of these use a piezoelectric sensor (usually a ceramic bender unit). The ceramic bender generates its own electric current that is directly proportional to the amount of acceleration. The activity devices usually include a volatile memory chip and a small computer or micro-controller chip. They are programmed to determine the amount of activity in a unit time and record that amount at a determined storage rate. The activity accepted by these devices is usually filtered so that they cover the dominant frequency ranges for human movement of about 0.5 to 10–15 Hz [432]. Later the data are downloaded to a computer, typically a desktop or laptop PC, usually through a special interface device. Various manipulations can then be performed on the downloaded data for further quantification or illustration. The activity data are maintained with a time-date code so that the activity can be analyzed by the time of each day recorded. The self-contained units are battery powered; current models provide batteries capable of actively recording for from 14 days to 4 years. Although shorter battery life does limit the maximum duration of the recording, battery life may increase as this technology develops. Another limitation on the duration of monitoring is the amount of computer memory available to retain the stored values. Currently, the memory size available for these monitors is 4 MB–1 GB. Duration of monitoring is inversely proportional to the rate at which values are stored. For low storage frequencies (e.g., once every minute), these capacities translate into a total monitoring period of 3–720 days. However, at high storage rates useful for examining individual movements (e.g., 10 per second), total monitoring would only be from about 7 min up to 29 h. These devices all use internal circuitry to sample the output voltage at a certain frequency (sample frequency or rate). The amount of activity can be determined by checking the number of times the voltage reaches or exceeds a minimum criteria (threshold crossing) or by some integration or summation of the total voltage from the individual samples. Integration provides the more sensitive approach, especially for examining individual movements as opposed to total activity. After a certain number of samples, the result either in total threshold crosses or integrated voltage is stored. The storage frequency or rate limits the time resolution of this technique. For assessing total activity occurring in spans of a few seconds to minutes, the digital sampling can be at relatively low rates (e.g., 4–8 Hz) and still provide an adequate measurement. But for higher storage frequencies designed to examine individual movements, a sampling frequency of 10–40 Hz is probably necessary. Storage rates of 10 Hz or more would be ideal although slower movements can be analysed with storage rates perhaps as low as 1 Hz.

In sleep–wake detection for the evaluation of various circadian rhythm disorders and in insomnia, the patient wears the device on the non-diominant wrist, and simultaneously keeps a sleep log for comparison. There have been several validation studies published, in both children and adults as well as in special populations, under a variety of conditions, evaluating a large number of wrist actigraphs from various manufacturers. Most devices show good sensitivity and specificity (of the order of 86–96 %), but low specificity (30–40 %); detection of wake is usually unsatisfactory [433–437]. These caveats should be borne in mind when interpreting data from actigraphy.

For movement disorders, the activity monitor is placed at the site of the abnormal movement. In general, the goal of such recording is to count and quantify such movements, not merely to indicate when movement occurs. Various earlier studies showed that abnormal movements associated with hyperkinetic disorders could be quantified using actigraphy [438, 439], if appropriate filtering was used to select for frequencies associated with the movements. Early studies attempted such a quantitation in PLMS. The total movement activity during sleep for patients was determined from activity monitors worn on the ankle of the affected leg, but the correlations between overall activity and the number of PLM were not high (values of about 0.6) [440]. More recently, sophisticated systems for counting movements have been developed and validated [441–443], making these systems useful for therapeutic monitoring or assistance in diagnosis of PLMD and RLS [444–446]. A much better correlation between total activity and specific abnormal movements may be obtained with a finer-grain analyses [447]. Recognizing the distinctive profile of individual movements requires matching the descriptive powers of an EMG record. To detect the onset and end of a specific movement requires sensitivity to higher-frequency components of the movement, necessitating sampling rates in the range of 10–40 Hz. Moreover, there are major data-storage problems for this condition. PLM are, by definition, greater than 0.5 s in duration (see Chap. 40). Activity measurements to detect PLM should have storage frequencies of at least 4 Hz and preferably 8–10 Hz to enhance measurement accuracy. A fine-grain analysis with 40 Hz sampling and storage at 10 Hz available from one of these monitors (PAM-RL, Respironics, Pittsburg, PA) provides a description closely matching the EMG recordings for these movements (Fig. 39.12). The recording at 10 Hz can then be saved for up to 7 days depending on the memory size in the units. In the more advanced activity meters such as the PAM-RL detections are based on sampling at 40 Hz with data stored for the activity summed over 4 samples (10 Hz data storage). The descriptive information about the movement along with total activity per 0.1 s permits a review of machine scoring to determine if criteria are met for periodic movements of sleep. The data provides an excellent agreement with the nocturnal PSG for number of leg movements (Fig. 39.13) with a correlation of 0.997 and an average error for rates per hour of less than 1.0. when done in the laboratory setting with calibrated meters [448]. The monitors when used off the shelf in a standard clinical setting also have very good agreement with results from the PSG and are considered validated for this use [443]. The PAM-RL has an advantage for home recordings since it records separately the PLM rates when the legs are stretched out from when they are upright (subject sitting or standing). Thus they give the PLM rates for the sleep position (although not sleep, per se) (Fig. 39.14).

Fig. 39.12

Example of a high-precision activity monitor worn on the ankle to detect leg movements

Fig. 39.13

Example of the real-time output of a high-precision activity monitor worn on the ankle (bottom line) compared to anterior tibilais EMG activity (top line). The middle line shows the real-time automatic detection of a significant leg movement made by the activity meter. The decision rules for the real-time leg movement detector create a 7-s delay in the detection

Fig. 39.14

Example of a computer display of stored leg activity and leg position data from a high-precision activity monitor worn on the ankle. The lighter areas on the graph (white background) indicate the leg is in a mostly horizontal position

The use of the new ambulatory monitors that provide this fine-grain analyses of movements might be further extended to assess other movement disorders in sleep, such as RBD or rhythmic movement disorder. But even such a development would fail to provide relevant information about the patient’s sleep–wake state. This can be approached by adding illumination or position information. To detect body position, a system has been developed [449] which requires wearing small monitors on the trunk and also on the leg just above the knee. Each monitor records position in three-dimensional space for each epoch (30 s to 1 min), and the combination of the two provides a description of the overall body position as standing, sitting, reclining, supine, prone, or lying on the right or left side. These monitors, when compared to direct observation of a subject’s body position, show an excellent overall agreement (contingency coefficients C = 0.85–0.91, maximum value of C for these data = 0.913). Activity data collected at the same time as position data permits differentiating abnormal movements that occur while the patient is lying down from those while standing or sitting. It also permits the detection of events during the sleep time when the patient sits or stands up, such as occurs for sleepwalking.

ii. Consumer-oriented sleep technology

In recent years, there has been an explosion of inexpensive, consumer-oriented, readily available technology that is meant to monitor, among other parameters, sleep quality and duration. Entries into this category include standalone wearable devices (e.g., FitBit, Jawbone) as well as smartphone-based software programs (“apps”) [450, 451]. While they all essentially use accelerometry-based techniques, as does actigraphy, to score sleep or wake (and in some cases distinguish between “light” and “deep” sleep) based on body movement, in most cases, the exact technology is proprietary, which limits detailed evaluation. “No-contact” bedside devices that detect sleep through radiowaves have also recently become available [452]. While such consumer-oriented sleep technology is very popular, mainly due to its convenience and easy accessibility, there is very little data validating it against established means of evaluating sleep [453]. Therefore, the sleep community remains uncertain as to how to approach this technology and how to interpret the information obtained from such devices and apps that patients bring with them to their clinic evaluations [454]. Recently published studies do suggest that this technology, while showing variable correlation with PSG-based scoring, has a sensitivity and specificity for sleep–wake detection that may be comparable to actigraphy [452, 455,456,457,458]. However the data are preliminary as of now. Thus, although the use of consumer-oriented sleep technology is likely to increase in the coming years, until larger studies that evaluate this technology are available, its exact role in clinical sleep medicine and research, if any, remains unclear.

B. Neurophysiological Studies

Polysomnography

Video PSG is the gold standard in diagnosing a variety of sleep disorders, specifically sleep-disordered breathing, abnormal movements in sleep, and nocturnal seizures. PSG techniques are discussed in detail in Chap. 17, and the scoring of sleep stages and respiratory events in sleep is discussed in Chaps. 24 and 25 respectively. Many abnormal movements in sleep as well as parasomnias are induced by arousals, which can be due to factors not obviously associated with the movements themselves, most commonly OSA. Treating OSA will decrease arousals, in turn decreasing the abnormal movements and parasomnias. Thus, even patients with what appear to be typical abnormal movements in sleep (HFT, ALMA, RMD, etc.) should be screened for possible OSA and undergo a PSG if they present with risk factors for sleep-disordered breathing.

When dealing with patients with abnormal movements in sleep, certain modifications to the PSG montage are helpful (see Chap. 18). While the standard PSG with a single EMG lead for the legs does provides a certain degree of information about certain motor disturbances, specifically PLMS, they may not be helpful when dealing with abnormal movements involving the upper extremities, the cranially innervated muscles, or even more proximal or more distal lower extremity muscles. For this reason, a multiple muscle montage that includes extra EMG channels recording from additional cranially innervated muscles (such as the sternocleidomastoideus, masseter and other muscles), upper limb (e.g., biceps, triceps, extensor digitorum communis, flexor digitorum subliminis), and lower limb muscles (e.g., quadriceps, hamstrings, gastrocnemius, and extensor digitorum brevis) and axial muscles (e.g., cervical, thoracic and lumbar paraspinals, rectus abdominis, intercostal muscles) (see Table 18.4) is recommended in those patients who have more complex movements by history. Similarly, an extended seizure montage with extra EEG channels (see Table 18.2), or a hybrid montage using select additional EEG and EMG channels (see Table 18.5) may be of benefit in patients in whom abnormal movements in sleep are suspected to be secondary to nocturnal seizures. Technician observations are invaluable in the documentation and description of events, and where the question is one of RBD, in eliciting dream recall. The sleep specialist can ask for no stronger ally than a vigilant technician who is aware of the clinical question being asked and is able to focus the camera on the area and movement of interest when abnormal movements occur.

Where patients complain of daytime sleepiness or abnormal movements during daytime naps, multiple sleep latency testing (MSLT), with multiple muscle montage if indicated, may be considered (see Chap. 22).

While ambulatory sleep studies (or home sleep tests [HST]) are gaining increasing acceptability and use in the evaluation of OSA, they suffer from significant limitations in the evaluation of abnormal movements in sleep, including a limited number or most often no EEG or EMG channels, no corresponding video recording, and no observer to document unusual behavior. Therefore, at the present time, HST is not recommended in the evaluation of patients with abnormal movements in sleep, such patients need to be evaluated by in-laboratory PSG.

Motor Evoked Responses

To more directly examine the impact of sleep on the motor system itself, motor evoked potentials (MEPs) can be studied. In one study of MEPs evoked by stimulating the motor cortex with a strong magnetic stimulus during sleep, it was noted that the MEPs decreased during NREM sleep [459]. Results during REM sleep have shown a much greater degree of variability in amplitude of evoked responses. Hess et al. [459] found that responses were of normal or increased amplitude, suggesting enhanced cortical excitability during REM sleep. In contrast, Fish et al. [261] found that average amplitude was decreased in 3 normal subjects with prolonged latencies in REM sleep compared to wakefulness despite variability of response amplitudes indicating maintenance of motor inhibition during REM sleep. In a group of narcoleptic patients, stimulation during cataplexy resulted in apparently normal MEPs [460]. While these results remain to be harmonized, the variability is consistent with the fluctuating balance between inhibitory and excitatory processes in REM. A finding of decreased mean amplitude, however, is more consistent with the general inhibitory balance of REM sleep in normals. When sleep apneas are superimposed on sleep, MEP amplitude may decrease further [461]. The use of MEPs in the evaluation of abnormal movements in sleep is mainly in the realm of research at this point and not of much value in everyday clinical practice.

Other methods used in Special Circumstances

In selected cases, a number of specialized techniques to evaluate for abnormal movements occurring in sleep can be performed. These include EEG–EMG studies with back averaging, reciprocal inhibition, long loop reflex (the “C” reflex), startle reflex, and somatosensory evoked potentials (SEP, e.g., giant SEP in cortical myoclonus). A detailed description of these techniques is beyond the scope of this chapter, but the reader is referred to other sources for further information [462].

C. Neuroimaging Studies

The development of new imaging techniques that permit assessment of activity in the waking brain provides an additional method of studying regional contributions to state-dependent motor activity. Studies of cerebral blood flow and metabolism have largely paralleled those of cellular activity. Techniques, including functional magnetic responance imaging (fMRI), SPECT, and PET scans including ligand studies, MR spectroscopy, functional or resting connectivity, diffusion tensor imaging (DTI) and tractography (for white matter imaging), voxel-based morphometry (for gray matter imaging), and transcranial sonography have all been employed in research and are discussed in greater detail in Chap. 21.

Blood flow and metabolism may be greater during REM sleep than in waking but are widely depressed during NREM sleep, especially SWS [463, 464]. Examining differential regional activities in relation to sleep states or features can provide insights into sleep mechanisms. In one study, Hofle et al. [465] correlated activity in different brain regions with power in different EEG frequency domains (e.g., delta, here 1.5–4.0 Hz). The greatest decrement associated with increased delta power (characteristic of SWS) was in the thalamus, consistent with the depressed thalamic activity of sleep. The presence of sleep spindles, most common in stage N2 sleep, is associated with activation of the thalamus, paralimbic areas, and the superior temporal gyrus [466]. Slow (11–13 Hz) and fast (13–15 Hz) spindles show this common activation, plus distinctive activations of the superior frontal gyrus (slow spindles) compared to sensorimotor cortical areas (fast spindles), medial frontal areas, and hippocampus. During REM sleep, in contrast to NREM sleep, there is activation of the brain stem core and thalamus as well as limbic areas of the brain and primary and secondary sensory areas [467–469], including visual cortices [470]. Hong et al. [471] examined the association between REMs and blood flow and found associations both with the midline attentional system active in REM sleep and areas involved in generating waking saccadic eye movements and subserving visual attention. Higher cortical areas, including prefrontal cortex and multimodal sensory and associative cortex, remain suppressed during all sleep stages [472]. REM sleep can be divided into those baseline periods without REMs and the periods with REMs, during which sensory receptivity is decreased [473]. During actual rapid eye movements, fMRI studies have shown additional activation in posterior thalamus and occipital visual cortex [474] or within a thalamocortical network including limbic and parahippocampal areas. Additional studies have shown that the basal ganglia are suppressed in SWS, but very strongly activated in REM sleep [463]. The significance of these basal ganglia changes for the motor system and for movement disorders in sleep remains unclear but is of great potential interest. These results in imaging studies suggest an evolving of sleep states in terms of the involved brain structures, including the motor system.

Imaging studies can also begin to assess potential deficits due to altered sleep conditions, such as sleep deprivation [475], which can cause depression of frontal lobe activity that is only partially restored after a compensatory sleep.

Principles of Treatment of Sleep Disorders Related to Abnormal Movements

The first and foremost step is to determine if sleep dysfunction is related to these abnormal movements at night, or if these are due to an associated common primary sleep disorder (e.g., OSA which has a prevalence in the population of about 14 % in men and 6 % in women between 30 and 70 years old; or a persistent insomnia disorder which may affect the quality of life in about 10 % of the population), or a comorbid psychiatric illness (e.g., anxiety, depression).

The basic treatment can be divided into two categories:

• 1. Treatment of the abnormal movements at night possibly responsible for sleep dysfunction using standard treatment for these movements. If these movement disorders are causing sleep dysfunction, optimal treatment of these involuntary movements should improve patient’s sleep dysfunction. Pharmacologic treatment of these jerks and shakes has been addressed briefly in the text (see above). Most of these abnormal movements do not require a specific treatment as these are mostly benign and will disappear in time but sometimes may persist into adulthood. However, treatment may be required if the movements are violent, injurious, or potentially violent. Pharmacotherapy usually includes benzodiazepines, selective serotonin reuptake inhibitors (SSRIs), or other antidepressants or anxiolytics for short-term relief. Non-pharmacological treatment may consist of psychotherapy, reassurance, and education of the patient and the family, progressive relaxation, and good sleep hygiene (sleep health) practice (e.g., regularize sleep–wake schedule, avoid alcohol, coffee consumption, and smoking near bed-time.). Also, attention should be paid to environmental safety measures, such as removing harmful sharp objects from bedroom, placing a mattress or other soft surface next to the bed, locking doors, and windows to prevent the patient from injuring himself/herself.

• 2. If sleep dysfunction is due to a primary sleep disorder, this should be treated using the standard method (e.g., CPAP for OSA and cognitive behavioral therapy (CBT) for insomnia with or without hypnotics used intermittently). Any comorbid psychiatric illness should be treated using generally accepted measures (e.g., SSRIs or other antidepressants and anxiolytics for depression and anxiety). Patients should always follow some common sense sleep hygiene measures. In a subset of patients with sleep onset or maintenance insomnia and circadian disruption associated with abnormal movements, especially when caused by neurodegenerative disease (e.g., Alzheimer’s disease and Parkinson’s disease) appropriately timed bright light therapy has been found to be useful. A word of caution is in order. Pharmacotherapy should be initiated at low doses, particularly in the elderly and those with neurodegenerative diseases, to minimize side effects. Finally, patients should be advised to avoid prolonged use of sedative-hypnotics and reduce or eliminate medications that may contribute to sleep dysfunction or OSA.

Box 39.2 lists these general principles of treatment.

Box 39.2: Principles of Treatment of Sleep Dysfunction Related to Abnormal Nocturnal Movements

• First determine if sleep dysfunction is related to abnormal nocturnal movements or a primary sleep disorder or a comorbid psychiatric-illness

• Treat primary movement disorder if it is causing sleep dysfunction

• Treat associated primary sleep disorder

• Treat comorbid psychiatric illness

• Initiate good sleep hygiene measures including regular sleep–wake dysfunction

• For pharmacotherapy try a non-benzodiazepine receptor agonist or a melatonin receptor agonist for short-term hypnotic use

• Start with a small dose and gradually increase the dose to minimize side effects in the elderly or those with neurodegerative disease

• Reduce or eliminate medications that may contribute to sleep disturbance or sleep apnea

• Attend to environmental safety precautions to avoid injury to patients

• Use appropriately timed bright light exposure in a subset of patients with insomnia and circadian rhythm disruption.

Summary and Conclusion

This chapter discussed an important aspect of human motor control (and dyscontrol) that has been largely neglected for a long time because it sits in the borderland of two important disciplines in medicine—those specializing in movement disorders and those specializing in sleep medicine. An understanding of motor control mechanisms is important for both fields. A breakdown in the delicate balance of motor control due to an affection of the afferent, central, or efferent structures can cause a dysfunction of voluntary movements or appearance of abnormal movements causing both positive and negative symptoms. Sleep modulates motor phenomena with progressive decline of motor activity due to increasing dominance of central inhibitory drive; concomitantly, the excitatory mechanism breaks through the inhibitory phase causing the appearance of motor events (some are physiological, but others are clearly pathological) Movement disorder specialists deal with diurnal involuntary movements, whereas sleep specialists encounter abnormal motor activities during sleep that may disturb sleep and result in impaired daytime functioning. The question often arises as to whether these are diurnal movements persisting during sleep, or abnormal movements triggered by sleep or intruding into sleep. This dilemma is highlighted by the fact that there are considerable similarities and overlaps between nocturnal and diurnal movements, and sleep may be disturbed by both diurnal and nocturnal motor events. There is a growing realization that both diurnal and nocturnal motor events may result from a common neurobiological alteration in the molecular mechanisms of motor control and sleep wakefulness. In this chapter, we briefly outlined motor control of human voluntary movements in wakefulness and sleep as well as suggested some pathophysiological mechanisms for abnormal jerks and shakes at night. We also provided a brief description of these conditions based on a method of classification for easy comprehension. Finally, we summarized principles of treatment of sleep dysfunction associated with these abnormal nocturnal motor events.