Introduction

Huntington’s disease (HD) is a progressive, fatal neurodegenerative disease caused by an abnormal expansion of a CAG repeat in exon 1 of the huntingtin gene (Huntington’s Disease Collaborative Research Group, 1993). The disease affects approximately 4–8 individuals per 100,000 [64] and typically presents between 35 and 45 years of age, with an average survival of 15–25 years after disease onset (Bates et al. 2002).

HD neuropathology shows a loss of up to 30% of brain weight resulting from neuronal cell death, with a direct correlation between brain atrophy and duration and severity of the disease [174]. Despite the widespread location of the mutant huntingtin, the primary atrophy of HD is located within the central nervous system. Whilst the basal ganglia have been shown to be preferentially affected with up to 60% loss of mass in the caudate nucleus, putamen, and globus pallidus [174], neurodegeneration is more widespread, probably from disease onset [140, 141], and includes a range of cortical and subcortical structures.

The classical description of HD is of involuntary choreiform movements involving the limbs and face, which reverts to bradykinesia, rigidity, postural instability, as well as axial posturing, or dystonia with disease progression [79]. HD patients also develop progressive cognitive [57, 91, 102] and psychiatric disturbances [58, 119, 120, 142, 155]. However, a number of other features have also been reported, including weight loss [6, 31, 59, 139] and changes in sleep [3, 123, 172]. Figure 1 highlights the brain regions affected in HD that are also involved in sleep.

Fig. 1
figure 1

Areas of the brain that experience pathology in HD that are also involved with sleep and circadian rhythms. Schematic diagram showing the various brain regions that experience pathology as part of the course of the disease which are also involved with the regulation of sleep and circadian rhythms

Full size image

Stages of sleep

Sleep is defined by different stages that occur in a characteristic, sequential order with specific electrophysiological patterns on electroencephalography (EEG) linked to electromyography (EMG) and electrooculography (EOG), collectively known as polysomnography (PSG) [138]. Sleep typically consists of a non-rapid eye movement phase (NREM), in which an increasing depth of sleep is seen from Stages 1–2 (‘light sleep’) to 3–4 [‘deep sleep’ or ‘slow wave sleep’ (SWS)]. The entire sequence from Stage 1–4 takes approximately 60–90 min in normal subjects, followed by rapid eye movement (REM) sleep. After approximately 10 min of REM sleep, the brain typically cycles back through the NREM sleep stages. This cyclical pattern repeats four to five times a night (see Fig. 2). Early in the night, NREM sleep is usually deeper and occupies a disproportionately large amount of time, especially in the first cycle, whilst REM sleep might be short or aborted. Later in the night, NREM sleep is lighter whereas a higher proportion of each cycle is made up of REM sleep [4], this effect is known as the polarity of REM sleep and is caused by a combination of circadian and sleep-dependant factors.

Fig. 2
figure 2

Different stages of sleep in a healthy young adult. Stages of sleep progress in a cycle from Stage 1 to REM sleep. Shaded circles represent percentage of time spent in the various stages of sleep. Adult humans spend approximately 5% in Stage 1 sleep, 50% in Stage 2 sleep, 20% in Stages 3 and 4 (SWS) and 25% in REM sleep. In contrast, infants spend about 50% of their time in REM sleep. An increase in age normally results in decreased metabolic rate and physical activity and thus a decrease in energy demands and the need to conserve energy. As a consequence, SWS and total sleep normally decreases while Stage 1 sleep increases [9, 123]. Stage 2 and REM sleep, however, remain unchanged until very old age [29, 66], although Stage 2 sleep spindles and K complexes become less well formed, less numerous, lower in frequency (spindles) and amplitude (K-complexes), and are accompanied by more awakenings during the night [123]

Full size image

Neurophysiology of sleep

Regional regulation of sleep

Traditionally, sleep has been considered to be a property of the whole organism, resulting in either an awake, drowsy or sleep state. This ‘whole-organism sleep’ was understood to be initiated and regulated centrally by interactions between specialized sleep and wake-promoting neuronal networks [78, 104, 146, 161]. More recently, a new theory has emerged suggesting that sleep is local and activity- or use-dependent, and is then consolidated by these central mechanisms. The primary units that transition between the sleeping and waking states are ‘cortical columns’, tightly connected neurons located in the cortex [82–84]. Therefore, sleep at this level of organization is auto-regulatory in that prior and ongoing activity in the network determines the probability of the network entering the sleep-like state, at least at the cortical level.

Central regulation of sleep

Once sleep is initiated locally as a consequence of previous activity, it is then consolidated by central mechanisms [84].

Wakefulness

Centrally, several distinct neuronal populations mediate arousal and the cortical desynchrony of wakefulness via projections to the thalamus and basal forebrain [107, 146]. Circuits regulating NREM sleep include the preoptic anterior hypothalamus, which contains the ventrolateral preoptic (VLPO) area, the median preoptic area and the basal forebrain. REM sleep and the alternation between NREM and REM sleep are also under central control, with the pedunculopontine tegmental nucleus (PPT) and the laterodorsal tegmental (LDT) nuclei being especially important in this process. Multiple wake-promoting networks also exist, including the orexin system in the lateral hypothalamus. This is summarised in Fig. 3, and shows that the ascending arousal system (AAS) has two major elements to it: the ascending pathway to the thalamus which activates the thalamic relay neurons which are important for transmitting information to the cerebral cortex, and a second pathway which activates the cerebral cortex to facilitate the processing of thalamic input.

Fig. 3
figure 3

Sleep–wake control systems of the subthalamic diencephalon and associated links to input from the circadian clock and to ascending arousal systems of the brainstem. Primary interactions include disinhibition of the sleep-promoting ventrolateral pre-optic area (VLPO) neurons by adenosinergic inhibition of GABA (γ-aminobutyric acid)-containing basal forebrain neurons [163]; VLPO GABA-mediated inhibition of brainstem and diencephalic ascending arousal systems [152]; reciprocal inhibition of VLPO cells by noradrenergic and serotonergic input from ascending brainstem arousal systems [160], and by GABA-containing cells of the tuberomammillary nucleus that are co-localized with histaminergic neurons [52, 145]; wake-related orexinergic stabilization of these same ascending arousal systems [145]; and brainstem cholinergic facilitation of wake-related basal forebrain cholinergic activity through a glutamatergic intermediary. 5-HT 5-hydroxytryptamine (serotonin), ACh acetylcholine, Glu glutamate, LDT laterodorsal tegmental nucleus, NA noradrenaline, NREM non rapid eye movement, PGO Ponto-geniculo-occipital, PPT pedunculopontine tegmental nucleus, REM rapid eye movement. VLPO ventrolateral preoptic area. Diagram was modified from [115]

Full size image

Sleep

The VLPO of the anterior hypothalamus is a critical region involved in inhibiting the arousal circuits during sleep [152]. Neurons in the VLPO send outputs to all of the major cell groups in the brainstem and hypothalamus involved with arousal [152]. VLPO neurons are active during sleep, therefore specific VLPO neuronal loss produces profound insomnia and sleep fragmentation [94]. It has been shown that the sleep-promoting anterior regions directly interact with the wake-promoting posterior regions of the hypothalamus in a mutually inhibitory manner [152]; as a result, these pathways function like a classic ‘flip–flop’ switch which produces sharp transitions in state, but is relatively unstable [52, 99, 145]. The addition of the LH orexin neurons, which are located outside the flip–flop switch, stabilizes it, thereby reducing transitions during both sleep and wakefulness. Thus, narcoleptic humans who lack orexin have increased transitions between these two states [145].

Sleep regulatory substances

Many substances have been implicated in sleep regulation. Sleep regulatory substances (SRS) act on subcortical sleep-regulatory circuits [111]. It is generally agreed that nitric oxide (NO), adenosine, prostaglandin D2, interleukin-1 (IL1), tumour-necrosis factor (TNF), and growth-hormone-releasing hormone (GHRH) are all involved in the regulation of the duration and intensity of NREM sleep [111]. These substances work in the biochemical cascades that form the NREM sleep homeostat and act on the basal forebrain neurons [7], the hypothalamic preoptic neurons [111], the locus coeruleus [28] or the serotonergic neurons of the raphe to promote NREM sleep [96]. REM sleep-promoting substances include vasoactive intestinal polypeptide, while prolactin and wake-promoting substances include orexin (mentioned previously), ghrelin, adrenocorticotrophin hormone and corticotrophin-releasing hormone (Tables 1, 2).

Table 1 Neural activity of neurotransmitter systems during sleep and wakefulness
Full size table
Table 2 Results from past HD-related sleep studies
Full size table

SRSs, therefore, act locally in the cortex to enhance sleep phenotypes by causing changes to the electrical and chemical outputs of a network, thereby altering their responsiveness to inputs [1, 21, 154].

Circadian rhythms

In mammals, the ‘master’ internal circadian clock is located in the suprachiasmatic nuclei (SCN) of the anterior hypothalamus and is vital for establishing the circadian rhythm of sleep–wake, Fig. 4.

Fig. 4
figure 4

The suprachiasmatic nucleus. a A schematic diagram showing a cross-section of the hypothalamus highlighting (in bold) the location of the suprachiasmatic nucleus (SCN), the body’s ‘clock’ in relation to other hypothalamic nuclei. Reproduced with permission from Sinauer Associates. b Cross-section of the human brain highlighting the location of the SCN (arrow). Reproduced with permission from J. Mai, Atlas of the Human Brain, Elsevier 2007. c Circadian rhythms of cortisol, melatonin and core body temperature

Full size image

The primary environmental synchronizer of mammalian circadian rhythms is the daily light–dark cycle. A class of intrinsically photosensitive retinal ganglion cells that express the photopigment melanopsin integrates photic information for entrainment within the retina and projects to the SCN core via the retinohypothalamic tract [11, 60]. The mechanism behind the circadian clock in mammals involves a cell autonomous transcription–translation negative-feedback loop consisting of a highly conserved set of core genes; Clock, Bmal1, period homologue 1 (Per1), Per2, Cryptochrome 1 (Cry1) and Cry2 [8, 93, 167], see Fig. 5a.

Fig. 5
figure 5

Gene expression in circadian timing. a E-boxes are regulatory DNA sequences that enhance transcription by providing a target for transcription factors. Circadian timing is sustained by three connected streams of rhythmic gene expression. The first stream involves the E-box-mediated activation of genes (including Per and Cry) by Clock/Bmal heterodimers in early circadian day. This activation is inhibited in the late circadian day by accumulation of Per/Cry complex in the nucleus, resulting in the closure of an oscillatory negative feedback loop. When Per/Cry levels decline, the circadian cycle of expression is initiated. The rate of expression is sensitive to the phosphorylation status of Per. Rev-erbα is expressed in phase with Per and Cry and acts as a negative regulator of Bmal1. Through disinhibition, Rev-erbα establishes a positive feedforward loop in part of the second stream of gene expression. This loop drives expression of Bmal1 in antiphase to the negative factors and helps to initiate the new cycle of gene expression, as well as to enhance core oscillation by segregating the intervals of peak Clock/Bmal transcription and peak Per/Cry negative feedback. Bmal brain and muscle aryl hydrocarbon receptor nuclear translocator (ARNT)-like, Cry Cryptochrome; Per Period. Diagram modified from [65]. b Schematic diagram showing SCN connectivity. The body’s biological clock has only a small number of outputs to sleep-regulatory systems. The majority of these outputs go to the area which includes the dorsal and ventral subparaventricular zone (vSPZ and dSPZ, respectively), and the hypothalamic dorsomedial nucleus (DMH). vSPZ neurons transmit information required for the organization of daily sleep–wake cycles, while the neurons of the dSPZ are vital for body temperature cycles. SPZ outputs are integrated in the DMH with other inputs. DMH neurons thus drive the circadian cycles of sleep, feeding, activity and corticosteroid secretion. Rhythms of body temperature are maintained by dSPZ projections back to the medial preoptic area (MPO), whereas the DMH is the original source of projections to the VLPO for sleep cycles, to the neurons of the paraventricular nucleus (PVH) containing corticotropin-releasing hormone (CRH) for corticosteroid cycles, and to the orexin and melanin-concentrating hormone neurons of the lateral hypothalamus (LHA) for wakefulness and feeding cycles. The connectivity of the SPZ and DMH allows circadian rhythms to adapt to environmental stimuli, such as food availability, as well emotional inputs from the limbic system, visceral sensory inputs and finally cognitive influences from the prefrontal cortex. ARC arcuate nucleus, DMH dorsomedial nucleus, LHA lateral hypothalamus, MPO medial preoptic area, PVH paraventricular nucleus, SCN suprachiasmatic nucleus, SPZ subparaventricular zone, VLPO ventrolateral preoptic nucleus, VMH ventromedial nucleus. Diagram modified from [146]

Full size image

Despite a recognized role for the SCN in governing the timing of sleep, the SCN has no monosynaptic outputs at all to the brainstem arousal sites and only minimal outputs to sleep-regulatory centers such as the LH and VLPO. The circadian regulation of sleep behavior is therefore thought to be mediated by multisynaptic projections from the SCN to sleep–wake centers of the brain [30, 146] via the ventral subparaventricular zone (SPZ) [175], followed by a secondary projection to the dorsomedial hypothalamic nucleus (DMH). The DMH sends an excitatory dense glutamatergic projection to the lateral hypothalamus and an inhibitory GABAergic projection to the VLPO [20], and is critical in the circadian regulation of sleep–wake cycles, see Fig. 5b.

The SCN clockwork also has active local “clock” systems in peripheral, non-neural tissues such as the lung, liver, pancreas, spleen, thymus and the skin [65]. These local clockworks are tuned into each other, and to solar time, by metabolic and neuroendocrine cues that depend on the SCN. Output pathways from the SCN, therefore, are not limited to the control of sleep–wake cycles and have been shown to mediate a diverse number of physiological functions both in the brain and the periphery, including the timing of hormone release, feeding behaviour and body-temperature fluctuations [2, 146].

In conclusion, even though there are central global coordinators of the sleep–wake states, such as the clock mechanisms of the SCN, the global co-ordination of NREM sleep is likely to reflect an emergent property of loosely coupled local processes. Sleep-regulatory circuits, therefore, integrate information that is related to locally induced cortical column states with information that is important for the determination of whether an animal is ultimately awake or asleep, for example the time of day, mental activity, sensory input and emotion and disease related information.

Purpose of sleep

Despite many proposed theories, the precise purpose and function of sleep remains unknown, although recently it has been suggested that sleep should be viewed as a state that increases the efficiency of behaviour by reducing energy use when activity is not beneficial. Thus sleep is a state of adaptive inactivity, rather than being considered to be a maladaptive and vulnerable state, persisting only because it contains some unknown adaptive physiological function [156].

Studies in rodents have helped to provide insight into the role of sleep; total sleep deprivation in rats produces a reliable syndrome that includes skin lesions, increased food intake, weight loss, increased energy expenditure, decreased body temperature and death [137]. Selective REM and SWS deprivation result in similar findings [87]. The effect on humans is not as striking as in other animals, however sleep deprivation or a disturbance of the sleep–wake axis has revealed a broad range of interconnected pathologies [17, 48], including, amongst other things, impaired memory and learning [15, 48], reduced mental and physical reaction times [74, 169], reduced motivation, depression [44, 48] elevated cortisol levels [92], increased susceptibility to illness [25, 149, 151] and metabolic abnormalities [148, 159]. A disturbance of sleep can therefore have a substantial impact upon an individual’s health and quality of life, which may be even more significant in disease states such as HD.

Aging and neurodegenerative disease

Predictable changes in sleep take place as part of the normal aging process in humans. Aging is usually associated with decreased metabolic rate and physical activity, and thus with a decrease in energy demands and the need to conserve energy. The resulting sleep-related changes include a decreased amount of total sleep time, sleep efficiency and percentages of SWS and REM sleep, with increases in sleep latency and Stages 1 and 2 [112]. Sleep therefore appears to become more fragmented and ‘lighter’ with age.

Patients with neurodegenerative diseases such as Parkinson’s and Alzheimer’s disease often have sleep disturbances [103, 133], which creates problems not only for the patient but also for their family and carers [86, 171]. For example, REM sleep behavior disorder (RBD) is characterized by a loss of the normal muscle atonia that accompanies REM sleep, allowing the person to “act out” his or her dreams, which can often involve harmful or violent behaviour. There is a high prevalence of RBD in neurodegenerative disease ranging from ~33 to 60% in PD [22, 51], 50 to 80% in Lewy body dementia [13] and 80 to 95% in Multiple Systems Atrophy [126, 165].

Although the nature of the sleep-related impairments differs to some extent in a disease specific fashion, some common abnormalities have been observed in neurodegenerative disorders of the central nervous system. Sleep becomes progressively fragmented with disease duration and is associated with increased awakenings and overall time spent awake [103, 123]. SWS and REM sleep are often reduced and polygraphic features (spindles and K complexes) become less well defined and numerous [103, 123, 176]. This often results in poor quality sleep and excessive daytime sleepiness that can lead to, or exacerbate, cognitive impairment, mood disorders and increase the risk for accidents [62, 63, 101].

Unfortunately, PD-related treatments such as dopamine agonists, (e.g. pramipexole, and ropinirole) have all been suggested as further contributing towards the sleep problems of PD in some patients by causing ‘sleep attacks’ [49, 113, 147]. Whilst the aetiology of this is unknown, it has been suggested that dopamine agonists cause a desynchronizing effect on the EEG that is reflected in a disruption of sleep continuity [16].

Polysomnography in Huntington’s disease

Although the majority of studies investigating sleep in HD were initially carried out ~20 years ago, interest in this aspect of the disease has recently resurfaced [3, 172].

To date, sleep studies in HD patients provide evidence of a progressively worsening sleep disorder [123], which appears to be independent of CAG repeat length [3, 61]. Mild stage patients reportedly have no clinical sleep disturbance, but do have mild PSG abnormalities, with increased interspersed wakefulness and a longer time to first REM episode [3, 61]. We have also seen something similar in unpublished data from eight patients in the early stages of the disease. In addition, we found an overall loss of form and definition in the patients’ rest-activity actograms suggesting deterioration of circadian timing (Goodman AOG, unpublished data), which is consistent with previous findings [106]. However, none of the patients complained of sleep disturbance or showed excessive daytime sleepiness using standard questionnaires such as the Epworth Sleepiness Scale and the Functional Outcomes of Sleep Questionnaire, although self-reporting questionnaires designed to look at sleep in HD patients may not always provide an accurate assessment of the situation since HD patients can lack insight [68]. Nevertheless, it has been reported that up to 87.8% of patients acknowledge having sleep problems, which were rated by 61.7% as either ‘very’ or ‘moderately’ important factors contributing towards the patient’s overall problems [168].

As HD progresses, it has been reported that awake EEGs exhibit a gradual slowing and diminution of amplitude [150, 158, 177]. Patients with moderate disease also experience increased time in Stages 1 and 2 of sleep [61] and reduced SWS and REM sleep [61, 157, 177]. REM sleep percentages have been shown to decrease with disease severity [3] and, in contrast to other neurodegenerative diseases, HD patients show a higher density of sleep spindles compared to healthy control subjects [36, 157, 177]. Moderately severe HD patients experience impaired initiation of sleep with increased sleep onset latency [177], as well as impaired maintenance of sleep with increased nocturnal awakenings or arousals and a high percentage of wakefulness after sleep onset [61, 157, 177]. Thus, HD patients experience reduced sleep efficiency [19, 61, 100, 157, 177]. A well known cause of excessive daytime sleepiness is sleep disordered breathing or sleep apnoea. However to date, no significant difference has been found between HD patients and controls in terms of sleep apnoea [14]. Issues of cataplexy, hypnagogic hallucinations or sleep paralysis have not been identified as problems in HD patients [3].

Brain-derived neurotrophic factor (BDNF), a neurotrophin that increases the resistance of neurons to metabolic, excitotoxic and oxidative insults [18, 33], has also been associated with slow-wave activity, with reports showing that a decrease in BDNF leads to a decrease in slow-wave activity [40]. As decreased production of brain-derived neurotrophic factor has been identified in HD patients [46, 179], it is therefore possible that this decrease contributes towards the decreases in SWS observed in HD patients [157, 177].

Sleep related movement disorders in Huntington’s disease

A range of different parasomnias have been described, of which the most common is Restless Legs Syndrome (RLS). Although a genetic association between RLS and HD has been suggested [37], RLS has not been reported to be significantly increased in the HD population.

Periodic limb movement (PLM) disorder (PLMD) is characterized by periodic episodes of repetitive and highly stereotyped limb movements that occur during sleep. Studies have found that HD patients have more frequent periodic leg movements compared to controls [3, 61], although there is debate as to how accurately the defined PLM criteria can distinguish between true PLMD in patients compared to those who simply experience chorea as part of their movement disorder.

Parasomnias in Huntington’s disease

Rapid eye movement sleep behaviour disorder (RBD) is a parasomnia characterised by vivid, often frightening dreams associated with simple or complex motor behaviour during REM sleep. Patients appear to act out their dreams, in which the exhibited behaviours mirror the content of the dreams and have been heavily linked to α-synucleinopathies. The risk for example of developing neurodegenerative disease in idiopathic REM sleep behaviour disorder is substantial (12-year risk of 52.4%), with the majority of patients developing Parkinson’s disease or Lewy body dementia [131]. The most recent polysomnography paper to publish on RBD in HD reported a 12% prevalence (three out of 25 HD patients) [3].

Possible causes of sleep disturbance in HD

Sleep and circadian systems are distinct, although they interact in a variety of different ways; not all sleep disorders result from a problem of circadian rhythmicity, but all circadian disorders produce sleep disturbances. Do patients with HD therefore experience a ‘pure’ sleep disorder, a circadian disturbance, or a combination of both? This is an important question to address, since a sleep disorder without an associated circadian abnormality can be treated as a ‘pure’ sleep disorder.

Transgenic mouse models of HD have shown that improving circadian and sleep patterns using Alprazolam can actually improve cognitive function and survival in R6/2 mice [116, 117]. Another recent study has shown that arrhythmic hamsters fail to perform in a hippocampal-dependant learning task and that learning can be restored using a GABA antagonist [143]. Therefore, if in patients there is circadian involvement, the impact on the disease could potentially be more widespread, since by treating the sleep and circadian disorder (with, for example, melatonin, chronotherapy or light therapy), one could potentially affect other aspects of the disease, including the cognitive deficits that are a core element of HD and which are a major problem for many patients and their families.

Neural pathway disruption

As a consequence of the widespread neurodegeneration in HD, areas directly or indirectly involved in sleep and circadian rhythms are likely to be affected, see Fig. 1. This includes the striatum as, although it is not considered to be a key part of the sleep–wake process (see above), studies involving striatal lesions have nevertheless suggested a possible involvement with sleep [23, 100, 173]. This effect may be mediated through the outflow of the basal ganglia to critical thalamic nuclei and the PPT [56, 77].

Significant volume reductions in the brainstem, a key structure involved in the regulation of sleep, have been reported in HD patients [140], with some evidence of such reductions worsening with disease severity and increasing UHDRS motor score [72]. This brainstem atrophy in HD has been described to precede even caudate atrophy [98] and thus, like PD, it may be that disturbances of sleep could precede the motor aspects of HD and could be a core feature of this disorder.

The hypothalamus, particularly the lateral hypothalamus, is crucial for the regulation of sleep and metabolism [164]. Neurons containing orexin found in the lateral hypothalamus [27, 124] have been reported to be abnormal in HD. Significant atrophy and loss of orexin neurons in the lateral hypothalamus of HD patients’ brains have been identified [5, 81, 122]; a subsequent neuroimaging study identified hypothalamic atrophy in patients, even in the early stages of the disease [76]. However cerebrospinal fluid samples have not always found evidence for a loss of this peptide [53].

The anterior, ventral region of the hypothalamus contains the SCN [125]. Abnormalities within it have been described in neurodegenerative diseases [65] with a loss of circadian synchrony in both the central nervous system and peripheral tissues and cells [125]. In HD, the reported sleep disturbance may reflect a more fundamental problem in circadian rhythms and their central and peripheral regulation; indirect evidence seems to support this. The primary symptom of a circadian rhythm disorder is the inability to sleep during the desired sleep time. HD patients reportedly lie sleepless at night with insomnia and have a tendency towards increased daytime somnolence and sleep episodes [172] with naps at unpredictable hours [3, 61]. Second, studies using actigraphy have found that HD patients make significantly more movements and have increased activity during sleep compared with controls [71]. This apparent reversal of the day–night pattern of sleep seems to support the suggestion of a disturbance in circadian rhythm [61]. Third, a disturbed circadian rhythmicity of hormones, such as prolactin, has been identified [45]. Finally, findings of increased REM latency in HD [3] are consistent with a phase-delayed circadian rhythm, further suggesting a problem with an internal misalignment of the body clock and sleep–wake cycle. Although a thorough and comprehensive study investigating circadian disturbance in HD patients has not been undertaken to date, this is not the case experimentally, where the R6/2 transgenic mouse model of the disease has been systematically investigated [39, 106]. So for example, Morton et al. reported that R6/2 transgenic mice had disturbed night-day activity that worsened with disease progression. These abnormalities were also accompanied by a marked disruption of expression of the circadian clock genes mPer2 and mBmal1 in the SCN, as well as in the motor cortex and striatum.

Other causes of disordered sleep in HD

Depression has been associated with difficulty falling asleep, frequent nocturnal awakenings, early morning awakening, decreased total sleep and non restorative sleep. In HD, affective symptoms such as depression and anxiety are common even before disease onset [24, 34]. Furthermore, commonly prescribed medications for HD patients, such as antidepressants, neuroleptics, dopamine antagonists and tetrabenazine, can impact on sleep and wakefulness by causing, for example, insomnia and increased drowsiness.

Silvestri et al. [157] reported that existing chorea decreases and may even disappear during sleep. Other studies report that both normal semi-purposeful sleep movements and dyskinesias are markedly suppressed during the deeper phases of sleep, yet reappear after EEG evidence of arousal, following a shift to a lighter sleep stage [19, 47]. Therefore, although chorea during sleep arousals is likely to affect sleep [97] it has, however, been discounted as the principal cause of the sleep disturbance in HD [47].

The presence of dystonia and age-related problems such as dementia, nocturia, body pain and other physical or mental health conditions [123], in addition to environmental factors, all may contribute to sleep abnormalities. For example, as the disease progresses, patients may choose to retire from their daytime jobs, resulting in an unstructured lifestyle, which may result in irregular sleep–wake behaviour and circadian rhythms (disturbed entrainment).

The consequences of disturbed sleep in HD

Considerable circumstantial evidence exists to suggest a link between the HD-related symptoms and sleep deprivation. HD has a characteristic cluster of symptoms that may include several of the following: a loss of motor control [79], changes in mood [75], memory [70] and cognitive impairment [102]. Sleep deprivation has also been implicated in many of the same repertoire of symptoms and signs, including profound effects on cognitive and motor function [35, 48, 90] and memory. For example, SWS is involved in hippocampus-dependent declarative memory consolidation [121, 127]. Evidence suggests that a loss of circadian timing reduces the ability of the hippocampus to encode learned information [143]. This may be relevant in HD, since impaired declarative memory has been found in HD patients, including presymptomatic gene carriers [54]. It is possible, therefore, that the memory deficits found in HD are related in part to a reduction in SWS and/or circadian disturbances. Perturbations in clock genes, both in humans and other animals, have also been associated with psychiatric conditions [89], addiction [95], and metabolic syndromes [150], all of which occur frequently in HD.

Metabolic consequences

Epidemiological and clinical data indicate that voluntary sleep curtailment [118, 166] and disorders that impair sleep architecture, such as narcolepsy [88] and sleep apnoea [134], are associated with an increased incidence of disrupted metabolic regulation. In rats, chronic total sleep deprivation leads to pronounced changes in energy regulation, glucose metabolism [148], plasma ghrelin and leptin concentrations [12, 109], increased food intake, progressive weight loss and hyperthermia [10, 38]. Indeed, HD patients experience hyperphagia, weight loss [42, 43, 80, 105, 132, 144, 162], increased 24 h energy expenditure [50, 55, 132] and glucose metabolism [41, 128] and altered leptin and ghrelin concentrations [130, 132] which could be related to changes in sleep. Furthermore, recent data have emerged to suggest that circadian processes are also critically involved in energy homeostasis [170].

HD patients with longer CAG repeats have been found to have a lower body weight, which the authors attribute to be the result of a hypermetabolic state [6]. A disturbance in sleep in this already vulnerable group is therefore likely to exacerbate any existing difficulty in maintaining weight.

The significance of disturbed sleep in HD

The question that therefore remains is which symptoms exist first in HD? Do sleep disturbances in HD not only exacerbate other pre-existing symptoms, but actually cause others? These are both relevant and important questions, since in an already vulnerable population, a disturbance of sleep would increase vulnerability. It is highly likely that in HD many of the symptoms experienced by patients are adversely affected by disturbances of sleep, especially since these problems can be found even in the early stages of the disease. The connection between abnormalities in sleep and circadian rhythms in relation to other signs and symptoms of the disease may be critically important. A decline in quality of life can directly result from disturbed circadian-regulated sleep, and changes such as this can have a profound impact on both patient and carer [136]. Defining the abnormalities of sleep that exist in HD and understanding their influence on other disease related features is important for enabling clinicians to initiate appropriate investigations and to instigate treatments that could dramatically improve quality of life in both patients and their families/carers; this may even have an effect on the natural history of the condition. Indeed, in this respect it has been shown that modification of sleep patterns using Alprazolam can actually improve cognitive function and survival in the R6/2 transgenic mouse model of HD [116, 117]. However, Alprazolam has been shown to cause circadian shifts in rodents but not in humans and, in fact, reduces restorative SWS and REM sleep [67], the key aspects of sleep that are already jeopardised in HD. Nevertheless, this study does highlight the point that modifying/correcting sleep abnormalities may improve a whole range of other abnormalities in HD, including cognition.

One study in PD involving 391 patients found that the clinical factors that showed the highest predictive value for worsening health-related quality of life were non-motor symptoms, such as sleep disorders [135]. Defining the extent and nature of sleep-related abnormalities in HD more thoroughly is not only critical in improving the quality of life in both patients and carers, but also in revealing additional clinically relevant information about the disease itself. If sleep abnormalities were predictive of HD, as they may be for PD [131], it may be possible to intervene to treat the sleep problem and consequentially reduce symptom severity. This may involve not just drug therapies, as in transgenic mouse models of HD; environmental enrichment has been shown to slow disease progression [69]. It is possible that lifestyle modifications such as a program of regular exercise may be a useful therapy in the treatment of patients with sleep disorders, since it has been shown that there is an association of regular exercise or physical activity with a lower prevalence of symptoms of disturbed sleep [153]. Indeed, a number of studies in healthy subjects have documented that both acute and long term exercise increase slow-wave sleep and total sleep time [85, 110, 178].

Conclusion

In conclusion, sleep sustains cognitive and physical performance, health, well-being and productivity and even mild sleep deprivation degrades performance over a few days. Widespread pathology is observed in HD, including regions that are involved in both sleep regulation and circadian rhythms. Evidence suggests that HD patients may experience disturbance in both of these processes, which in turn may have effects on a wide range of other essential functions, such as metabolism. Abnormalities in sleep and circadian rhythms also have negative impacts upon cognitive and psychiatric function as well as memory, all of which are problems that lie at the heart of HD. Treating these disturbances as part of the general care of the patient has the potential to dramatically improve quality of life, as well as potentially reducing the severity of other symptoms and could even affect disease progression.