Sleep disorders in Lewy body dementia: Mechanisms, clinical relevance, and unanswered questions

In Lewy body dementia (LBD), disturbances of sleep and/or arousal including insomnia, excessive daytime sleepiness, rapid eye movement (REM) sleep behavior disorder, obstructive sleep apnea, and restless leg syndrome are common. These disorders can each exert a significant negative impact on both patient and caregiver quality of life; however, their etiology is poorly understood. Little guidance is available for assessing and managing sleep disorders in LBD, and they remain under‐diagnosed and under‐treated. This review aims to (1) describe the specific sleep disorders which occur in LBD, considering their putative or potential mechanisms; (2) describe the history and diagnostic process for these disorders in LBD; and (3) summarize current evidence for their management in LBD and consider some of the ongoing and unanswered questions in this field and future research directions.

management of sleep disorders should form an essential part of the care of LBD patients.This review aims to (1) describe the specific sleep disorders which occur in LBD, (2) describe the history and diagnostic process for these disorders in LBD, and (3) summarize current evidence for their management in LBD.It should be noted that in many cases, specific evidence for the management of sleep disorders, in DLB or PDD specifically, is lacking; therefore, much evidence is extrapolated from similar conditions, mainly PDD.As such, clinical judgement is essential when applying such evidence to LBD patients.

METHODOLOGY
This review provides a narrative overview of current evidence relevant to the etiology and clinical management of prominent sleep disorders in LBD, in particular, for the benefit of non-sleep specialists who may diagnose or treat LBD, as well as the broader LBD research community.
Literature search and synthesis have been targeted to provide a level of detail judged sufficient for this purpose.Further detail may be found in publications signposted or referenced within the text.
A "semi-systematic" search and synthesis strategy was used based on the following sources: -Scoping literature search, using PubMed, to identify publications between 1996 and November 2022 (search terms detailed in Supplementary Material, Section 2 in supporting information) on topics of physiology of sleep and arousal; sleep and arousal in LBD; sleep disorders in LBD; pathophysiology, diagnosis, and management of EDS, RBD, RLS, OSA, and insomnia.
-Key publications (reviews, clinical guidelines) and hand search of references conducted to identify relevant publications.
-Clinical and research experience of expert authors in fields of sleep medicine and LBD (authors KNA, BFB, JPT, IM).
-A systematic search strategy, using OVID Medline, to identify clinical trial evidence for treatments detailed in management sections and tabulated in Supplementary Material, Section 1 in supporting information.The search strategy is detailed in Supplementary Material, Section 2.2 in supporting information.

ANATOMY AND PHYSIOLOGY OF SLEEP-WAKE REGULATION
First, we will consider normative, physiological aspects of sleep.It should be noted that much of this evidence is derived from animal models.
Sleep and wakefulness are generated by the interactions of discrete neural circuits.Wake-promoting projections from the brainstem and basal forebrain are balanced by sleep-promoting inhibitory projections from the ventrolateral preoptic nucleus (VLPO) and parafacial zone. 12Basal forebrain projections to the cortex (GABAergic, glutamatergic, cholinergic projections) drive wakefulness. 12,13Glutamatergic projections to the cortex from the brainstem peduncu-

RESEARCH IN CONTEXT
1. Systematic Review: This narrative review aims to provide an introduction to the physiology of sleep and arousal and the pathophysiology of sleep disorders in Lewy body dementia (LBD) and offers a guide to the clinical assessment and management of sleep disorders in LBD.It has been informed by synthesis of the available literature, by a semi-systematic approach using medical literature databases (OVID Medline and PubMed), focusing on sleep, arousal, and LBD, and by expert knowledge from international experts in sleep medicine and LBD.
2. Interpretation: This review begins with an overview of sleep physiology, primarily from a functional neuroanatomical standpoint.It then proceeds to sections on specific sleep disorders occurring in LBD, exploring current understanding of the pathophysiology, diagnosis, and management of these disorders.In the first sub-sections, their epidemiology in LBD and putative or potential mechanisms are summarized.In the second sub-sections, specific guidance on the clinical assessment for sleep disorders in LBD is provided.In the third, current evidence for the management of common sleep disorders in LBD is reviewed, informed by a systematic search of treatment trials conducted for this review.Finally, we expand on and discuss promising novel therapeutic options in LBD. 3. Future Directions: Throughout this review, we highlight unanswered questions and offer suggestions for future avenues of investigations.We identify the work needed to better understand the pathophysiology and etiology of sleep disorders in LBD, to test available treatments in LBD, and to develop novel treatments for the LBD cohort.
In our conclusion, we summarize the key pathophysiological and clinical questions to be answered in future studies.
We hope that this review inspires greater use of precision sleep medicine in both clinical and research practice, such that better nights lead to better days.lopontine and parabrachial nuclei are also essential in generating wakefulness.Other brainstem, midbrain, and hypothalamic projections modulate arousal. 12,14,15These include brainstem cholinergic projections from the laterodorsal tegmental nucleus and pedunculopontine nucleus (PPN) to the thalamus; GABAergic projections from the lateral hypothalamus to the thalamus; and cortex and projections to the cortex from nuclei including noradrenergic (locus coeruleus), histaminergic (tuberomammillary nucleus), serotonergic (raphe nucleus), dopaminergic (ventral tegmental area, ventral peri-acqueductal gray nucleus), and orexinergic (perifornical lateral hypothalamus) nuclei. 12,14,15e balance between wake-and sleep-promoting circuits is conceptualized by the "flip-flop switch model" (Figure 1). 12,14,16This model F I G U R E 1 "Flip-flop switch model" of sleep/wake transition.
-Sleep and wake promoting nuclei are mutually inhibitory.
-Increasing activity on one side creates a self-reinforcing loop in which its activity increases inhibition of the opposing side, thereby decreasing inhibition of its own activity, resulting in a sharp demarcation, and a swift transition between states and avoiding prolonged periods of semi-wakefulness.
-The switch is weighted toward sleep or wakefulness by the influence of process S and process C, which alter the likelihood of sleep depending on time spent awake (sleep debt) and time of day/light levels, mediated by the effect of somnogenic molecules and suprachiasmatic nucleus and dorsomedial hypothalamic signaling on the activity of ventrolateral preoptic nucleus (sleep-promoting) and posterior lateral hypothalamus (wake-promoting) neurons.
proposes that sleep-promoting and wake-promoting projections are mutually inhibitory, creating a self-reinforcing loop and sharp transition between states.The equilibrium between sleep and wakefulness is influenced by two processes. 17Process S (homeostatic sleep drive) is an increasing drive for sleep, mediated by adenosine and other "somnogenic" molecules, accumulating during wakefulness.Adenosine accumulates in the basal forebrain (and is cleared during sleep) and increases sleep-promoting VLPO activity. 17Process C (circadian sleep drive) is a cyclical drive for sleep, synchronized to the time of day and the light-dark cycle, mediated by signaling from the suprachiasmatic nucleus (SCN). 15,17These processes promote transition from sleep to wakefulness (or vice versa), with a predictable pattern.
[20] NREM sleep is driven by GABAergic projections from the ventral periacqueductal gray (vPAG) area and glutamatergic projections from the area ventromedial to the superior cerebellar peduncle. 21The vPAG maintains NREM by inhibiting REM-generating glutamatergic neurons in the sublaterodorsal tegmental nucleus (SLD), also referred to as the subcoeruleus complex (SubC) in humans.Meanwhile GABAergic neurons in the parafacial zone inhibit the wake-promoting glutamatergic outputs from the parabrachial nucleus. 21REM sleep occurs when GABAergic projections from the extended ventrolateral pre-optic area inhibit the vPAG, releasing its inhibition of the SLD. 22The SLD also inhibits skeletal motor neurons directly, via projections to inhibitory GABAergic neurons in the spinal cord, and indirectly via GABAergic ventromedial medulla (VMM) projections, which, in turn, extend to the spinal cord.This produces motor atonia during normal REM sleep.
Meanwhile ascending projections from the VMM modulate cortical activation during REM sleep. 22This modulation is controlled by the SLD, which shapes the firing rate of the VMM across different brain states. 22ring sleep there is reduced processing of sensory information, in which information is preferentially processed when it is "salient," referred to as sensory gating. 23The thalamus is an important relay station in the brain, through which the majority of sensory information passes, and therefore, is a key structure for the gating of sensori-motor F I G U R E 2 Typical hypnograms for healthy young adult and healthy elders (>60 years).Demonstrates normal sleep cycle and sleep architecture 18 and details changes in sleep architecture observed in healthy aging 19 and in Lewy body dementia. 20EEG, electroencephalogram; NREM, non-rapid eye movement sleep; N1-N3, NREM sleep stages; REM, rapid eye movement sleep information traveling to/from the cortex in both sleep and wake. 23The thalamus coordinates the electrophysiological activity occurring during different sleep stages.It also plays a key role in the initiation and dissemination of cortical slow wave activity and the generation of sleep spindles, which are considered important for neuroplastic functions, such as memory consolidation. 24The thalamus receives inputs from cholinergic, noradrenergic, serotoninergic, and histaminergic arousal structures, along with pyramidal neurons of the frontal cortex, and GABAergic neurons of the basal forebrain facilitating the control of the thalamic coordination and gating functions, within the sleep-wake cycle. 16

Epidemiology and pathophysiology of EDS in LBD
EDS describes difficulty remaining awake and alert during the day. 3ddy et al. reported EDS in 50% of DLB and 57% of PDD patients sampled at baseline (DLB: mean 2 years after dementia diagnosis; PDD: mean 3 years after dementia diagnosis), compared to 10% in agematched controls and 18% in Alzheimer's disease (AD). 25EDS becomes more prevalent as LBD progresses, with the rate of EDS increasing to 85% in PDD and 75% in DLB, after 2 years. 25Cross-sectional studies using the Epworth Sleepiness Scale 26 and Multiple Sleep Latency Test (MSLT) 27 also identified significantly higher rates of sleepiness in DLB (81% in both studies) versus AD (39%-45%).
The pathophysiology of EDS in LBD is unclear.Degeneration of the cholinergic nucleus basalis of Meynert (nBM), in the basal forebrain, may explain cognitive fluctuations in LBD; 28 however, lesions/inhibition of equivalent neurons, in rodents, mainly affect cortical activity, increasing slow-wave electro-encephalography (EEG) rhythms, without reducing time awake. 13,29Meanwhile, GABAergic basal forebrain neurons, in rodents, do support wakefulness, 30 with activation increasing wake time and fast EEG activity, while inhibition increases sleep time by approximately 3 hours. 31Despite this, Kasanuki et al. 32 show that DLB patients with baseline EDS have greater nBM cell loss at autopsy than their non-sleepy counterparts and that sleepiness is the strongest predictor of nBM degeneration, more so than dementia severity, hallucinations, or cognitive fluctuations.Further studies are needed to explore this discrepancy.
Destruction of dopaminergic vPAG neurons, in rodents, also causes an ≈20% loss of wake time; 33 however, few studies have explored the relationship between vPAG degeneration and EDS in LBD.A post mortem study 34 found significant vPAG neuronal loss in DLB patients versus controls but, perhaps due to poor availability of retrospective EDS data, an association between vPAG degeneration and EDS symptoms was not observed.Specific inhibition of noradrenergic locus coeruleus (LC) neurons, in rodents, has little effect on general wakefulness, but appears to be critical for sustained wakefulness in response to novel environmental stimuli. 35While a relationship between LC degeneration and impaired arousal has been previously proposed, 36,37 imaging or post mortem studies have yet to test this relationship in LBD.One supplementary abstract 38 reported a significant association between LC neuronal loss and EDS in PDD patients.A cross-sectional study 36 found a positive correlation between EDS and depressive symptoms in DLB.This is relevant as LC degeneration has also been proposed as a substrate for development of depressive symptoms in LBD. 39The authors suggest that disturbance in arousal and mood symptoms may share a common etiology in noradrenergic dysfunction and could be improved in LBD, with noradrenergic pharmacotherapy. 36idence for improvement in EDS with psychostimulant treatments (Section 4.3), which enhance dopaminergic and noradrenergic function, further underwrite the relevance of these neurotransmitter systems in the etiology of EDS.
EDS and cognitive fluctuations are difficult to disentangle, 28 as both involve spontaneous alterations in cognition, attention, and arousal. 40 is unclear whether EDS and cognitive fluctuations are distinct, but associated, phenomena or if they share a pathophysiological substrate. 40Observational studies 36,41 found no associations between measures of arousal and attention in DLB, suggesting that cognitive fluctuations might be independent from fluctuations in arousal.Conversely, Matar et al. 40 argue that cognitive fluctuations and EDS arise within a common pathophysiological framework, due to degeneration in large-scale brain networks, resulting in aberrant switching between brain states, subserving sleep through to attention.Lewy body pathology, in LBD, extensively affects regions with overlapping roles in the modulation of arousal and attention (brainstem, 5 hypothalamus, 7 basal forebrain 6,42 ) and multiple studies using functional magnetic resonance imaging (fMRI) and EEG demonstrate disturbances in large-scale network regulation, related to attentional function in LBD. 4,40In keeping with the flip-flop switch model, more wide-spread degeneration would result in greater state instability and therefore, more marked abnormalities in arousal, that is, EDS.Consistent with this, Abbott et al. 43  Similar proposals of state instability have been proposed as the proximal cause of delirium but considering the role of "thalamo-cortical dysrhythmia." 44,45This describes disruption to the rhythmic oscillations of thalamo-cortical neurons 44 leading to unstable switching between brain states and so consciousness level may change from quiet vigilance to drowsiness, waking dreaming, to agitation.Emerging evidence supports this phenomenon in delirium, 44 though further confirmatory work is needed.However, as with cognitive fluctuations, it is not clear whether EDS and delirium fall along the same spectrum, or should be considered distinct entities, both pathophysiologically and phenomenologically.There is conflicting evidence for the effect of isolated thalamic lesions on arousal.Thalamic lesions, such as paramedian thalamic stroke, cause EDS, alongside apathy; 46 however, this has been attributed to the loss of ascending dopaminergic and noradrenergic inputs to the thalamus, 24 while in-depth examinations of such strokes have also shown that there is damage to regions including the paramedian midbrain and hypothalamus. 16,47rther research on the neuroanatomical basis of EDS in LBD is needed.Studies are needed to assess whether objective measures of sleepiness (e.g., MSLT) and wakefulness (e.g., Maintenance of Wakefulness Test) are associated with evidence of cognitive fluctuations.
Continuous polysomnography (PSG), during sleepiness tests and attentional tasks, could be used to assess for sleep intrusions, for example, microsleeps, which could explain attentional abnormalities observed.fMRI and EEG measures probing apparent intermediate states such as "confusional waking from sleep" or "drifting off" might be used to distinguish between sleep-wake instability and attentional deficits.The role of brain regions such as the thalamus and the cholinergic, dopaminergic, GABAergic, and noradrenergic systems require further evaluation, in the context of EDS in LBD.

Diagnosis of EDS in LBD
The DLB international consensus criteria list "hypersomnia," that is, EDS, as a supportive clinical feature for diagnosing DLB. 2 Meanwhile, the presence of cognitive fluctuations, "pronounced variations in attention and alertness" is a core criterion for DLB diagnosis.EDS has been shown to be the most characteristic feature of fluctuating cognition, differentiating DLB from AD or normal aging. 2,48For PDD, 49 cognitive fluctuation and EDS constitute associated clinical features that allow diagnosis of probable PDD.EDS is therefore a useful clinical marker, during the assessment of suspected LBD.
EDS secondary to LBD is diagnosed clinically, based on patient and informant reports.

Key questions:
-"Is there tendency to doze or nap in the day?If so, how often, when, and how long?" -"Can the patient get through the day without sleeping?" Useful clinical questionnaires for gathering information about EDS include: -Epworth Sleepiness Scale (ESS), a retrospective subjective sleepiness measure asking about likelihood of sleep in various situations over the past 4 weeks. 50arolinska Sleepiness Scale (KSS), a retrospective subjective sleepiness measure asking about overall sleepiness in the last 5 minutes. 51S scores >10 are abnormal and are seen with LBD-related EDS, though separate causes should also be considered. 52Medication effects, nocturia, motor and neuro-behavioral symptoms and co-morbid sleep disorders, such as OSA and RBD, may all interfere with sleep and worsen EDS. 9,53EDS also poses the risk of attentional lapses during high-risk activities (e.g., driving, cooking).Therefore, safety, particularly driving safety, should be addressed where relevant. 54Use of the "Diamond Lewy diagnostic assessment toolkit for DLB" is recommended as it provides an assessment framework with questions on EDS and RBD, as well as non-sleep clinical features. 55

Management of EDS in LBD
As with insomnia, advice on good sleep hygiene is important (Section 7.3) and improving night-time sleep quality may improve EDS.
OSA is a significant cause of EDS in patients without LBD and may be a modifiable factor to improve EDS in LBD (Section 6.3).Decreasing or stopping sedating day-time medications, or those night-time sedative medications which have persisting effects into the day (e.g., clonazepam) may also be beneficial. 9,56 performed a systematic search to find trial evidence for pharmacological therapy for EDS in LBD (Table S1.1 in supporting information).Very few treatment trials were identified for EDS in LBD, though small randomized controlled trials (RCTs) and open-label trials have been conducted in PD.While clinical experience suggests that cholinesterase inhibitors (AChEi) can improve sleepiness, 9 RCTs for AChEi such as rivastigmine 57 and donepezil 58 do not report subjective or objective sleepiness measures.Psychostimulants such as modafinil, armodafinil, or methylphenidate have been trialed in PDD with positive results and may improve EDS in some LBD patients. 9,59idence for modafinil is conflicting with two RCTs in PDD indicating a significant benefit on subjective sleepiness (ESS) 60,61 and two RCTs in PDD indicating no significant benefit, in terms of subjective (ESS) and objective (MSLT) sleepiness measures. 62,63One openlabel study of armodafinil, in DLB, demonstrated improvements in subjective and objective measures of sleepiness, alongside improvement in neuropsychiatric symptoms, 64 while an open-label trial of methylphenidate, in PDD, reported improvement in ESS. 65Psychostimulants can exacerbate anxiety and psychotic symptoms, so should be used with care in LBD by specialists experienced in their use; 66 however, the identified studies did not report prominent adverse events, suggesting that these psychostimulants could be safely, and perhaps effectively, used in LBD. 59,61,62,64Large RCTs, in LBD, are strongly recommended.
Other medications, used in attention deficit hyperactivity disorder or narcolepsy (atomoxetine, sodium oxybate) have been investigated for EDS in PD.Atomoxetine (selective noradrenaline reuptake inhibitor) significantly improved ESS scores versus placebo in a RCT in PDD. 67Sodium oxybate (GABA-B receptor agonist) significantly improved subjective (ESS) and objective (MSLT) measures of sleepiness in a crossover RCT in PDD, 68 as well as ESS versus baseline in an open-label study in PDD. 69Pre-clinical work has suggested a wake-promoting effect for memantine; 70 however, it did not improve ESS, versus placebo or baseline, in two RCTs in LBD 71 and PDD. 72ough EDS is prominent in LBD, few trials report sleepiness measures as a primary outcome measure.Further trials using subjective and objective measures of sleepiness in LBD are needed to evaluate the efficacy of currently available and novel medications.

Epidemiology and pathophysiology of RBD in LBD
RBD is defined by REM sleep without atonia (RSWA), alongside violent and often injurious dream enactment behavior.Tonic and/or phasic muscle twitching, during REM sleep, is seen on video polysomnography (vPSG) alongside (relatively infrequent) dream enactment. 73RBD occurs in ≈1% of people ≥ 60, 74 but with far higher frequency in LBD.Matar et al. 75 identified 50.8% of LBD patients as having probable RBD at baseline, while Ferman et al. 76 found that RBD was present in 76% of autopsy-confirmed DLB patients sampled.RBD has a direct association with the neurodegenerative process occurring in α-synucleinopathies, 73 such that the majority of patients with vPSGconfirmed RBD have prodromal neurodegeneration, attributable to α-synucleinopathies. 77,78 Neurodegeneration in pontine or medullary areas is associated with loss of REM atonia. 77,79,80Dysfunction/degeneration of the pontine SubC (also known as the SLD, in rodents) has been proposed as an explanation for RBD. 77Inactivation of vGLUT2 in rodent SLD neurons induces RSWA and causes a 30% decrease in REM sleep quantity. 81other candidate is the GABAergic/glycinergic VMM, downstream of the SubC/SLD. 77VMM lesions in cats induce RSWA, 82 and genetic inactivation of VMM neurons causes RSWA, without decreasing REM sleep quantity. 83As isolated RBD (iRBD) patients do not show altered REM sleep quantity, Dauvilliers et al. 77 argue that RBD is more likely to result from specific VMM degeneration rather than SubC degeneration.
Direct evidence in humans for specific lesions occurring in RBD is lacking and does not discriminate contributions from SubC and VMM damage.Case reports of RBD, as a result of inflammatory, 84 ischemic, 85 or surgical 86 insults, implicate both pontine and medullary structures.Neuropathological evidence in iRBD is rare given its typical presentation in middle age and the high likelihood of subsequent conversion to overt α-synucleinopathy; however, one case report 80 describes an iRBD patient, diagnosed at age 52 who died at age 72, without cognitive impairment or parkinsonism, and was found to have significant Lewy body pathology in the SubC and medulla, without significant neuronal loss.Post mortem examinations of three RBD patients who developed PD/LBD, found Lewy body pathology in the SubC, gigantocellular reticular nucleus (within the VMM), and PPN, although pathology was also present in the brainstem, limbic system, and cortical areas. 87creased substantia nigra echogenicity and reduced striatal dopamine transporter density have been demonstrated in RBD and progressive changes correlate with risk of phenoconversion to overt α-synucleinopathy; 88,89 however, this likely reflects underlying progression of the neurodegenerative process, rather than directly causing RBD. 90Boucetta et al. 91 report volume loss in the pontomesencephalic tegmentum in PD with RBD versus those without; they also identified extensive volume reductions relative to PD without RBD, in the cerebellum, diencephalon, striatum, limbic system, anterior cingulate, and olfactory gray matter, but did not observe correlations between region-of-interest volumes and RBD symptom scores.
Neuromelanin-sensitive imaging studies showed reduced signal intensity in the SubC and adjacent LC in iRBD and PDD-RBD, with neuromelanin signal inversely correlating with the degree of RSWA. 92,93wever, the interpretation of these findings must be considered with caution.The SubC is small and the relatively larger, neighboring LC is a significant source of neuromelanin signal which could not be delineated from SubC signal in these studies. 94Moreover, the RSWA observed could instead be attributable to co-existing neurodegeneration in other areas (pons, medulla, and brainstem in general) 95 which are profoundly affected by Lewy body pathology.Nonetheless, these studies do implicate changes to the coeruleus/sub-coeruleus complex, in RBD etiology.In a multi-modal imaging study, Knudsen et al. 96 showed early degeneration, in iRBD, of both central (LC signal loss) and peripheral (reduced cardiac metaiodobenzylguanidine uptake) noradrenergic neurons with lesser (similar to controls) accompanying degeneration of the substantia nigra, relative to PDD.There was a non-significant but visually apparent split in the LC degeneration patterns in their PDD group, with RBD +ve PDD patients exhibiting greater LC neuromelanin loss than PDD patients without RBD.Given the proximity of the SubC to the LC, it is possible that RSWA may occur due to spill-over damage or local pathologic spread from the degenerating LC to the SubC.
Overall, it appears likely that the causative lesion in αsynucleinopathy-related RBD is localized to pontomedullary circuits, likely the SubC and VMM. 77Imaging techniques with greater spatial resolution, such as 7T structural and neuromelanin-sensitive MRI, may facilitate better characterization of the patterns of neuronal loss and dysfunction in the SubC and VMM in those with and without RBD, and provide a better understanding of RBD pathophysiology.

Diagnosis of RBD in LBD
Screening for RBD is an essential part of the LBD assessment.The DLB international consensus criteria 2 list RBD as a core clinical feature of DLB, while RSWA is an indicative biomarker, meaning that in the presence of progressive cognitive decline, probable DLB can be diagnosed based on vPSG evidence of RBD and RSWA alone.
RBD should be suspected in patients with clinical history of recurrent dream enactment confirmed with vPSG. 73Collateral history, ideally from a bed partner, is important, as patients may be unaware that dream enactment occurs. 97The first question on the Mayo Sleep Questionnaire, regarding RBD, referred to as the MSQ-1, has a sensitivity of 98% in detecting RBD (specificity 74%): 98 -"Have you ever seen the patient appear to 'act out his/her dreams' This question is incorporated into the Diamond Lewy diagnostic assessment toolkit for DLB. 55An alternative single item screening tool, the REM Sleep Behavior Disorder Single-Question Screen (RBD1Q) has been shown to have a sensitivity of 93.8% and specificity of 87.2%.
RBD1Q 99 : -"Have you ever been told, or suspected yourself, that you seem to 'act out your dreams' while asleep (for example, punching, flailing your arms in the air, making running movements, etc.)?" Other features include prominent recall of dreams and vivid dreams/nightmares, with particularly violent, aggressive, or frightening themes. 97

Management of RBD in LBD
Avoiding injury and ensuring a safe sleep environment is a key objective in managing RBD.Hard or sharp objects should be removed from the bed vicinity and bed partners should consider sleeping separately to avoid injuries from the patient.Exacerbating medications such as antidepressants (selective serotonin reuptake inhibitors, serotonin and norepinephrine reuptake inhibitors, tricyclic antidepressants) and lipophilic β-blockers (e.g., bisoprolol) should be reviewed. 59mplicating conditions (e.g., OSA) should be identified and treated.
High-quality evidence for the pharmacotherapy of RBD is limited (Table S1.2).][104][105][106][107] They also have methodological limitations, including small sample sizes and potentially unreliable outcome measures.the other hand, is well tolerated at doses between 3 and 12 mg and despite less evidence for efficacy, represents a safer option for first-line treatment. 9While published evidence is lacking, combination therapy of clonazepam and melatonin may also be effective. 109ere is some evidence in support of other medication options.
Memantine reduced RSWA in a 24-week study in 20 LBD patients versus controls. 71Two small RCTs showed that rivastigmine 110 significantly improved RBD symptoms, versus baseline and placebo, while a case report suggests that donepezil may also improve symptoms. 1113][114][115] A recent RCT, evaluating sodium oxybate in iRBD and PDD-RBD patients, 116 showed significant improvement in the frequency of RBD episodes and total movements on vPSG versus baseline; however, significant differences versus placebo were not observed.
Given the limitations of current evidence for RBD treatment in LBD, and the side effects of clonazepam therapy, there is a need for further large-scale studies of available therapeutic options and further work to identify new therapies.The development and validation of accessible, reliable, and reproducible measures of RBD/RSWA is an essential component of this, as the variability of subjective questionnaire measures and the poor feasibility of longitudinal PSG monitoring significantly impair the ability of clinical trials to obtain reliable results for RBD treatment. 102 is also important to highlight the value of iRBD as an early marker of α-synucleinopathy, which can occur two decades before conversion to parkinsonism or dementia, providing potential opportunities to study pathophysiological progression and attempt disease modification. 117Further work is needed to establish reliable methods of predicting which iRBD patients are on a neurodegenerative trajectory and their likely progression rate, to facilitate the implementation of early disease modification trials.Indeed, a recent review 117 argues that neuroprotective trials in iRBD are feasible now, highlighting, first, iRBD as a prodromal marker for α-synucleinopathies and, second, the suitability of iRBD patients as a target cohort for disease-modifying trials.Notably, large-scale natural history studies are now in progress to characterize iRBD patients longitudinally in a comprehensive manner, which should foster more optimal clinical trial design. 118,119OBSTRUCTIVE SLEEP APNEA IN LBD

Epidemiology and pathophysiology of OSA in LBD
OSA is characterized by snoring, with recurrent episodes of partial or complete collapse of the upper airway during sleep, associated with cortical arousal, and recurrent oxygen desaturation. 120OSA can cause EDS and impair sleep quality. 120Untreated OSA can also cause cognitive impairment and worsen mood and quality of life (QOL) in individuals without dementia. 120Therefore, while EDS, cognitive impairment, and depression may be directly attributable to LBD, OSA is an important differential diagnosis and additional, modifiable factor when managing these symptoms in LBD.
OSA is present in > 30% of people > 65, 121 increasing in prevalence among frail older adults. 122The prevalence of OSA in PDD and DLB appears similar to the background rate, with evidence of sleep-disordered breathing in 20% to 30% of PDD patients 123 and 18% to 35% of DLB patients. 27,124Studies in PDD demonstrate mild sleep apnea overall, 125,126 but with substantial inter-subject variability, with some demonstrating moderate to D and DLB, based on apnea-hypopnea index (AHI), found mild OSA (AHI 5-15) in 19.2% of PDD patients versus 8.6% of DLB patients, moderate OSA (AHI 16-30) in 19.2% of PDD patients versus 17.4% of DLB patients, and severe OSA (AHI >30) in 7.6% of PDD patients versus 8.6% of DLB patients. 124neral risk factors for OSA include age, high body mass index (BMI), and male sex; 129 however, in LBD, additional factors may increase risk of OSA, including reduced tone in upper airway muscles, 130 irregular respiratory flow oscillations, 131 use of sedating medications, 132 and dysfunction of the autonomic regulatory mechanisms for respiration. 133Subclinical respiratory dysfunction has been identified in DLB.Mizukami et al. report decreased ventilatory response to hypercapnia in DLB, versus AD and controls. 133Similarly, Hibi et al. report abnormal breathing rhythms in DLB, during quiet wakefulness. 131A neuropathological study 134 found that neuronal loss in the pre-Bötzinger complex, important for respiratory rhythmogenesis, and the medullary raphe nuclei, important for respiratory chemosensitivity, 134 were more severe in DLB than controls, but less severe than in multiple system atrophy.These findings support the argument that neurodegenerative respiratory disturbance is a feature of DLB; however, large-scale studies are needed to measure respiratory abnormalities during sleep and wakefulness in LBD.

Diagnosis of OSA in LBD
While OSA is not a core or supportive feature for LBD diagnosis, it is clinically important itself, and is a key confounder in the diagnosis and treatment of RBD, 135 EDS, 52,135 RLS, 52 and insomnia. 52For example, Iranzo and Santamaría 135  The presence of the following symptoms may be indicative of OSA and are used to support the diagnosis: -Fragmented, unrefreshing sleep.
-Waking up gasping or choking or with sore dry throat.
-Witnessed (typically loud) snoring and/or breathing interruptions during sleep, noted by an informant (ideally bed partner).
The STOP-Bang questionnaire is a widely used screening tool with eight questions capturing information on snoring, EDS, apneic periods, hypertension, BMI, neck circumference, age, and sex. 136A score of >3 predicts > 50% chance of subsequent detection of OSA on a respiratory sleep study. 136The diagnosis can be made based on overnight, respiratory sleep-study evidence of 15 or more "predominantly obstructive respiratory events" per hour without the presence of clinical symptoms, or based on evidence of five or more "predominantly obstructive respiratory events" per hour, in the context of indicative symptoms noted above. 100

Management of OSA in LBD
OSA management is undertaken by specialist sleep services.Therefore, early identification and referral are essential. 9Weight loss, if BMI is raised, and reducing smoking and alcohol use are important supplementary measures. 120Medications such as opioids, clonazepam, or antipsychotics may need to be rationalized. 120ere are no trials evaluating treatments for OSA in LBD, and few in PDD 137,138 (Table S1.3).Nonetheless, CPAP is the first-line therapy for OSA, including in LBD. 120Treatment of OSA with CPAP has been shown to improve cognitive performance, EDS, mood, and QOL 139,140 and the benefits of treatment may be shared by caregivers, as well as patients, as CPAP has been shown to improve sleep quality in bed partners of OSA patients. 140,141A systematic review 142 supports the benefits of CPAP on cognition, with 9 out of 11 studies reporting protective effects of CPAP on cognition, reducing the incidence and progression of cognitive impairment in mild cognitive impairment (MCI)/AD patients.
CPAP is tolerated in ≈50% of cognitively impaired older adults. 143,144Issues applying and adjusting the mask due to cognitive impairment and poor dexterity; nasal dryness; and interruptions in using the device, due to nocturnal toilet trips, are frequently reported issues. 145Patient and caregiver education, frequent followup, and heated humidification improve compliance; 145 however, providing effective therapy remains a challenge for those who cannot tolerate regular CPAP.Mandibular advancement devices (MAD) can be effective for mild to moderate OSA, with some evidence of benefit in PDD, 138 but MAD require good dentition and may be less appropriate in older patients with LBD. 120 Given the cognitive and neuropsychiatric benefits of OSA treatment, there is need for additional treatment options in LBD.

Epidemiology and pathophysiology of insomnia in LBD
Insomnia disorder describes difficulties initiating sleep or maintaining sleep with subsequent daytime impact. 146Insomnia symptoms increase with age, with prevalence rates approaching 50% in people > 65. 147 In LBD, insomnia prevalence (23%-43%) 148 is similar to the agecomparable background rate.Sleep architecture changes with age with increased awakenings and decreased slow-wave sleep (Figure 2). 149fragmented night may also be due to LBD-related factors including medication effects and symptoms such as nocturnal hypokinesia, dystonia, pain, mood changes, and nocturia 150,151 or may be secondary to other sleep disorders, for example, RLS. 52 However, these factors do not fully account for abnormalities in sleep architecture observed in LBD.PSG assessments in DLB patients have shown that patients have poorer sleep efficiency compared to healthy age-matched controls and show evidence of sleep fragmentation, not attributable to age, or respiratory or motor issues, suggesting that there is also a primary neurodegenerative basis for insomnia in LBD. 20,124,152PO dysfunction has been proposed as a neuro-physiological substrate of primary insomnia. 153Animals receiving lesions to the VLPO exhibit reduced sleep time as well as sleep instability. 154Lewy body pathology occurs in the hypothalamus in PDD and DLB, 7 while abnormal thermoregulation in PDD has been attributed to neurodegeneration in the median preoptic area, 155 close to the VLPO and also involved in sleep-promoting signaling. 156Fragmented sleep in AD and healthy older adults is associated with loss of VLPO neurons. 157Neuropathological studies focusing on the VLPO in LBD have not been performed, so it remains unclear whether VLPO degeneration occurs in LBD and, in particular, in patients with prominent insomnia symptoms.
Abnormal circadian regulation of sleep (process C) is also proposed as a mechanism for primary insomnia. 146Animal models support the concept that circadian rhythm may be affected in αsynucleinopathies. 158Transgenic mice, overexpressing α-synuclein (Th1-αSyn mice), exhibit increasingly fragmented circadian behavioral patterns over time, alongside decreased excitability of the SCN. 158Post mortem evidence, in advanced PDD, 159 observed Lewy body pathology in the SCN of 9/13 participants and in the pineal gland in two patients.
Evidence for circadian dysfunction, in humans, has mainly been published for PDD rather than DLB.Alterations in circadian patterns have been identified by multiple measures in PDD.Nocturnal hypertension occurs commonly in PDD and patients show an inverted circadian pattern with higher nocturnal versus daytime blood pressure. 160PDD patients exhibit a loss of normal variation in levels of plasma melatonin and it has been proposed that degeneration of dopamine-containing cells in the retina of PDD patients may negatively affect the alignment of dark/light cycles mediated by the SCN. 161Core body temperature variation has also been shown to be abnormal in PDD and these abnormalities are correlated with more significant sleep issues, including prolonged sleep latency, decreased REM sleep, and greater severity of RBD symptoms. 162Though few studies have evaluated the presence of circadian abnormalities in DLB, Raupach et al. 163 found a significant decrease in amplitude of core body temperature in isolated RBD, PDD with concurrent RBD, and DLB versus controls, while PDD patients without RBD did not differ from controls.By contrast, Harper et al. found that post mortem-confirmed DLB patients showed greater disturbances of circadian locomotor activity than AD patients, while variations in core body temperature in DLB did not differ significantly from healthy controls. 164Firm evidence for circadian dysrhythmia as a prominent feature of DLB is lacking, therefore, and the degree to which circadian dysfunction contributes to sleep disturbance, in both DLB and PDD, is unclear.Further work is needed to characterize the patterns of physiological and hormonal changes occurring over 24 hours in LBD patients to understand whether these represent a primary circadian dysfunction or a consequence of dysfunction in other neuronal systems.

Diagnosis of insomnia in LBD
The diagnosis of insomnia is primarily clinical, with diagnosis possible with careful history alone, based on ICSD-3 criteria 100 for chronic insomnia disorder.
The key features, relevant to LBD, are as follows: (1) Disturbance of sleep onset or sleep maintenance: ○ Difficulty initiating or maintaining sleep?
○ Waking earlier than desired/early morning waking?
(2) Occurring at least three times a week for at least 3 months.
(3) Causing distress or further impairment in social, occupational, or other areas of functioning: ○ Irritability or low mood.
Poor sleep environment and sleep hygiene; medical co-morbidities, age-related or otherwise; concomitant medications; and night-time LBD-related symptoms may significantly contribute to disturbance of sleep and should be considered. 52,109Patient and collateral histories should cover other medical conditions and medications, including: -Nocturia.
-Use of alcohol, caffeine, nicotine, and illicit drugs throughout the day.
When daytime sleepiness is reported, the ESS is useful to quantify the sleepiness. 50Typically, in insomnia, the ESS score is low (2 or less), while scores > 10 indicate EDS. 52While EDS is very common in LBD, the co-occurrence of apparent night-time sleep disturbance and EDS should increase suspicion of confounding sleep disorders.If there are features suspicious of OSA or RLS or RBD, referral for specialist sleep assessment should be considered.
In many cases, the most helpful question is to ask the patient and their caregiver to take you through a typical 24 hours, to gain an overall picture of their sleep-wake cycle.Alongside history, a sleep diary is recommended, as it provides longitudinal information on sleep timing and duration. 52A 1-week sleep diary should document: time to bed, lightsoff and lights-on times, and the total time spent asleep during the day and night.Records showing frequent day-time naps and altered sleepwake cycles can help to differentiate insomnia from EDS or circadian rhythm disturbance alone.Due to cognitive impairment, caregivers will likely be required to assist in completing the diary in LBD, potentially limiting its feasibility/utility.Actigraphy (the measurement of limb movement over time, using wearable sensors) is emerging as an additional method for obtaining similar information. 165The availability and use of actigraphy in clinical practice has yet to become widely established; however, the AASM have offered "conditional" support for its use in the assessment of insomnia and circadian rhythm disturbance in adults, noting that actigraphy may be especially useful in those who are unable to reliably complete sleep logs, such as individuals with cognitive impairment. 165

Management of insomnia in LBD
Good sleep hygiene is essential for managing insomnia in LBD. 9 Measures include daily natural light, a fixed morning rising time, avoiding stimulants (caffeine after lunch), avoiding excess alcohol intake, establishing a conducive sleep environment (comfortable bedding; appropriate temperature; a quiet, dark room), restricting daytime naps, and taking regular exercise. 9,109Exercise, in particular, may be beneficial with an RCT in PDD 109 showing significant improvements in sleep efficiency and wake after sleep-onset (WASO), versus sleep hygiene advice alone (Table S1.4).Co-morbidities, concomitant medications, and night-time LBD-related symptoms may contribute to sleep disturbance (e.g., nocturia, RLS, pain, nocturnal hypokinesia, neuropsychiatric symptoms) and should be treated. 9,109The Diamond Lewy management toolkit provides recommendations on the management of urinary, motor, and neuropsychiatric symptoms in LBD. 66In cognitively intact older adults, cognitive behavioral therapy for insomnia (CBTi) is the first-line treatment for insomnia; 52 however, evidence for the efficacy or feasibility of CBTi in MCI-LB or LBD is lacking. 166A systematic review 166 found evidence that adapted CBT paradigms can improve anxiety, depression, and QOL in MCI/dementia but there was insufficient data on the use and efficacy of CBTi specifically.
In some cases, pharmacological options may be required to treat insomnia in LBD; however, few trials have been conducted in LBD (Table S1.4).Melatonin is first-line pharmacotherapy for insomnia in people > 55, according to the British Association for Psychopharmacology. 167Evidence in LBD is lacking but RCT evidence in PDD indicate the benefit of melatonin on subjective measures of sleep quality, [168][169][170] though one RCT, 170 using PSG, did not observe significant changes in PSG measures with melatonin versus placebo.
A meta-analysis, 171 pooling data from nine RCTs in neurodegenerative diseases, showed that melatonin improves subjective sleep quality.
The Diamond Lewy consensus group also recommends melatonin as a first-line treatment for insomnia given its benign side effect profile. 9 Non-benzodiazepine Z-drugs, such as zopiclone or eszopiclone, do not have specific evidence in LBD but expert opinion also supports their use for short-term management of insomnia. 9No evidence was identified for zopiclone in PDD; however, an RCT assessing eszopiclone in PDD was identified, which observed significant improvements in the number of awakenings and subjective sleep quality versus placebo, though significant change in total sleep time (vs.placebo) was not observed. 172Given their negative effects on cognition, EDS, and risk of falls and fractures, z-drugs should be used with care. 9o open-label, single arm studies of galantamine in PDD and DLB provide rare direct evidence in LBD, reporting significant improvements in total Pittsburgh Sleep Quality Index and/or Neuropsychiatric Inventory sleep sub-scores. 173,174The benefit of galantamine, as well as other AChEis, warrants further investigation in LBD.An update on the Movement Disorder Society evidence-based recommendations for treatments of non-motor symptoms in PDD highlights the likely efficacy of rotigotine for insomnia in PDD. 59Our search identified two RCTs evaluating the effect of rotigotine on sleep in PDD, with evidence of significant improvement in sleep efficiency, WASO, sleep latency, 175 and subjective measures of sleep quality 175,176 (Parkinson's Disease Sleep Scale-2); however, the potential to exacerbate EDS may limit the utility of dopaminergic treatments, such as rotigotine, for insomnia in LBD.Rotigotine decreases sleep latency 175 and a meta-analysis by Sun et al. reports a statistically significant risk of daytime sleepiness with rotigotine. 177Further investigation in LBD is needed to better understand the risks and benefits of its use.

Epidemiology and pathophysiology of RLS in LBD
RLS has a well-defined clinical phenotype, with diagnosis based on ICSD-3 criteria requiring five essential features: (1) the urge to move the legs (not necessarily accompanied by dysesthesia); (2) occurring primarily with rest/inactivity; (3) occurring primarily in the evening or night; (4) partially or totally relieved by movement, as long as the movement continues; and (5) associated with distress, sleep disturbance, or functional impairment. 100Abnormal brain iron metabolism, altered opioid physiology, and dopaminergic dysfunction are all impli-cated in the pathophysiology of RLS (without LBD). 178RLS is nine times more prevalent in people with iron-deficiency anemia than the general population. 179S responds well to levodopa leading to hypotheses that RLS was the result of brain dopamine deficiency and could be associated with α-synucleinopathy-related dopaminergic dysfunction; however, several lines of evidence suggest that the underlying dopaminergic abnormalities in LBD and RLS are distinct.178 First, there is little evidence for an epidemiological association between RLS and PDD or LBD.The prevalence of RLS increases with age, with rates of 10% between 30 and 79-year-olds and 19% for those > 79. 180 The prevalence rate of RLS in PDD has been reported as 11% 181 and there is insufficient data on the incidence of RLS in DLB.Methodological issues include confounding from both symptom mimics in PDD (unpleasant sensory symptoms and akathisia), and the treatment effect of dopaminergic medications on RLS, which potentially depresses prevalence estimates in PDD.179,181 Second, neuropathological studies in idiopathic RLS are opposite to typical findings in PDD: in the substantia nigra, neuronal loss is absent, Lewy bodies are absent, and iron staining is decreased.182,183 Functional neuroimaging studies in RLS show normal presynaptic nigrostriatal dopaminergic function, while this is impaired in early PDD without RLS.184 Unlike in PDD, positron emission tomography and single-photon emission computed tomography studies indicate that dopamine levels are increased and D2 receptor density is decreased in RLS, while cerebrospinal fluid metabolite studies have demonstrated increased dopamine turnover is occurring suggesting a hyper-rather than hypo-dopaminergic state occurring in RLS.178 This raises the question of why levodopa or dopamine agonists would be beneficial in a hyperdopaminergic state.Evidence suggests that brain dopamine signaling follows a circadian pattern 185 with levels decreasing in the evening.Therefore, post-synaptic adaptation to excess dopaminergic signaling, during the day, results in a relative dopaminergic signaling deficit at night, despite overall increased dopamine levels.Dopaminergic supplementation corrects the relative deficit; however, increased dopaminergic stimulation can lead to increased downregulation of D2 receptors, worsening the underlying neurotransmitter-receptor discrepancy and increasing the dose of dopaminergic medication needed for symptom relief.This phenomenon, known as augmentation, can limit usefulness of dopaminergic agents. 178Therefore, while RLS and LBD share common treatment options, current evidence does not support an association between the conditions.

Diagnosis of RLS in LBD
RLS may be diagnosed clinically, based on the ICSD-3 criteria detailed in section 8.1.
History taking should focus on identifying these features: (1) Urge to move the legs (with or without dysesthesia): ○ A restless feeling in the legs or an urge to move?
○ Daily or more infrequently?
-Dysesthesia accompanying this urge is variable and poorly characterized.
(2) Occurring primarily with rest/inactivity and (3) occurring primarily in the evening or night: ○ Does it occur when relaxing, particularly in the evening, or when trying to sleep, at night? (4) Partially or totally relieved by movement, as long as the movement continues: ○ What does the patient do when experiencing the feeling?
○ Is it improved by movement?
○ Does it come back when movement stops?
(5) Associated with distress, sleep disturbance, or functional impairment: ○ Is there associated distress or night-time agitation?
○ Is there evidence of broken sleep or EDS?
Patients with cognitive impairment may have difficulties recalling or describing these features.The bed partner may provide a description of night-time periodic limb movements of toe, foot, and ankle, every few seconds. 52,109dical and drug history is important.Medications which can worsen RLS include antidepressants such as tricyclics and selective serotonin-reuptake inhibitors, antihistamines, antipsychotics, lithium, and metoclopramide.Quantifying nicotine, alcohol, and caffeine use may also be useful, as limiting these can improve symptoms.RLS can occur due to iron deficiency.Measurement of serum ferritin should be used to screen for iron deficiency, if RLS is identified. 109

Management of RLS in LBD
The intermittency and functional consequences of RLS symptoms, including distress, should be considered when deciding upon treatment.With intermittent symptoms (less than daily), lifestyle modification should be considered first-line. 186This includes decreasing alcohol, nicotine, and caffeine consumption; taking light exercise, stretching, and warm baths. 109When iron deficiency is present (ferritin level < 75 ng/mL), replacement is recommended as those with low iron may have a limited response to other RLS treatment. 109A combination of oral iron replacement and vitamin C is effective and well tolerated, but improvement may take at least a month. 109In some, iron infusions can be used to normalize ferritin levels and provide relief from RLS symptoms. 109ose with distressing and persistent symptoms should be offered pharmacological therapy. 186Though no studies have evaluated treatment for RLS in LBD (Table S1.5), dopaminergic medications (e.g., rotigotine, pramipexole) and α2-delta calcium channel binding ligands (gabapentin, pregabalin) are effective in treating idiopathic RLS. 9 LBD patients may be receiving dopaminergic therapy, which can be adjusted to address RLS.Long-acting oral and transdermal preparations may be particularly useful. 9LBD patients can also be sensitive to the side effects of dopaminergic medication, however; therefore, optimizing levodopa monotherapy to address RLS may be helpful.While augmentation appears to occur less frequently in RLS with PDD, 145 increasing doses of dopaminergic therapy may still be required to manage RLS, due to augmentation.EDS and psychotic symptoms, which can be exacerbated by dopaminergic therapy, may limit options for dose escalation.Gabapentin and pregabalin, at night, can help avoid the need for dopaminergic treatments; however, these medications often have a sedative effect and detrimental effects on cognition and should be avoided where possible, in the day. 9

NOVEL AND EMERGING THERAPEUTICS FOR SLEEP DISORDERS IN LBD
Novel pharmacotherapies targeting sleep-or wake-promoting neuropeptide receptors have been developed and released in the last 15 years, but availability is geographically limited.Melatonin receptor agonists (e.g., ramelteon) and dual orexin receptor antagonists (e.g., lemborexant, daridorexant) are now available, in the United States and Japan, as treatments for insomnia.Lemborexant 187 and daridorexant 188 have RCT evidence of benefit in older adults and daridorexant is currently under review by the National Institute for Health and Care Excellence for use in the UK.Unfortunately, two RCTs of ramelteon for iRBD were terminated due to poor subject recruitment. 189Specific evidence for the tolerability or efficacy of such medications in LBD has not yet been published. 189Orexin receptor agonists (e.g., danavorexton) are in development, with phase 1 trial evidence showing significant improvements in objective and subjective sleepiness with danavorexton, in idiopathic hypersomnia. 190Similar agents may offer an alternative to psychostimulants for treating EDS in LBD in the future.
Bright light therapy (BLT, a lightbox that exposes the patient to bright light in the day) may be a non-pharmacotherapeutic option.BLT significantly reduced night-time awakening and improved sleep quality in dementia, in a recent meta-analysis of 18 RCTs. 191Some included studies also reported improvements in EDS with BLT. 192The optimal timing, duration, and intensity of BLT has not been established.
Accordingly, treatment regimes in this meta-analysis varied significantly, potentially compromising the comparability of findings.There have not been studies specifically testing BLT in LBD and more work is needed in this regard.

Neuro-modulation presents another alternative to
pharmacotherapy. 193There is an emerging evidence base for the safety and efficacy of hypoglossal nerve stimulation (HGNS) in moderate to severe OSA. 194While stimulator insertion is invasive, it is less invasive than other surgical options for OSA and so could be more suitable for LBD patients, when unable to tolerate CPAP or oral appliances.However, any invasive intervention must be carefully weighed in LBD, and specific trials in patients with dementia will be needed to facilitate a balanced judgement on the risks and benefits of HGNS in LBD and other dementias.Transcranial magnetic stimulation (TMS) may have potential for the treatment of multiple symptom domains in LBD, including sleep disorders. 195,196Studies have reported beneficial effects for TMS on RLS 197 and EDS (in a recent small RCT in PDD), 198 while a study 199 showed that TMS can also increase respiratory airflow in OSA, by stimulating submental muscles without inducing wakefulness, suggesting that TMS could be a viable non-invasive strategy for treating OSA in the future.Vagal nerve stimulation (VNS) has been shown to improve EDS in treatment-resistant epilepsy. 200,201MSLT and ESS scores were not correlated with decreased seizure frequency, suggesting a separate mechanism outside of seizure reduction. 201n-invasive transcutaneous cervical VNS also significantly improved ESS scores in Sjögren's syndrome. 202Degeneration of the vagus nerve is known to occur in early in LBD, but its role in sleep and arousal abnormalities in LBD is notably under-researched. 203Similarly, the neurophysiological effects and potential treatment mechanisms for VNS are poorly understood.VNS may increase brain noradrenergic and cholinergic function, 195 lending face validity to the hypothesis that VNS could offer viable treatments for EDS in LBD, but more work is needed to understand its mechanisms and demonstrate its utility in LBD.
Invasive brain stimulation may facilitate the modulation of sleep and arousal in LBD.Deep brain stimulation (DBS) 193 is increasingly used clinically in PDD, when drug therapy for motor symptoms has failed 204 and effects on sleep/arousal have been reported as secondary outcome measures in several DBS studies in both LBD and PDD.DBS targeting the nBM bilaterally has been trialed in six PDD patients, 205 observing a non-significant improvement in caregiver-rated EDS between active stimulation and sham conditions, while overall Neuropsychiatric Inventory score improved significantly.A further study of nBM-DBS in five DLB patients did not report EDS measures, but observed a non-significant reduction in a measure of cognitive fluctations. 206bthalamic nucleus stimulation in PDD improves PSG measures, increasing total sleep time, decreasing WASO, and increasing periods of continuous sleep. 207It also significantly improves EDS (compared to baseline) in PD. 208 Meanwhile, the effect of DBS on RBD and RLS is unclear with conflicting observations of improvement and worsening of both disorders with subthalamic nucleus and PPN stimulation.○ α-synuclein seeding assays.
○ structural and/or functional neuroimaging markers.
In BOX 1, we summarize key questions, highlighted in this review, with potential methods to address them in future studies.We suggest a precision sleep medicine approach for our LBD patients, with assessments at multiple levels, including further epidemiological assessments; behavioral, molecular, electrophysiological, radiological, and neuropsychological phenotyping; as well as the evaluation of novel diagnostic and therapeutic strategies.While many of these can be both while sleeping?(punched or flailed arms in the air, shouted or screamed)?"• If yes, (A) Duration in months or years?(B) Has the patient ever been injured from these behaviors (bruises, cuts, broken bones)?(C) Has a bed partner been injured from these behaviors (bruises, blows, pulled hair)?(D) Has the patient told you about dreams of being chased, attacked, or that involve defending himself/herself?(E) If the patient woke up and told you about a dream, did the details of the dream match the movements made while sleeping?
While the sensitivity and specificity of the direct questions used in the MSQ-1 and RBD1Q is high, other sleep disorders (OSA, narcolepsy, sleep terrors) may confound the diagnosis.Therefore, referral to a sleep specialist is important, for exclusion of mimics and confirmation of the diagnosis using vPSG.Diagnosis of RBD based on the International Classification of Sleep Disorders Third Edition (ICSD-3) criteria 100 requires evidence of RSWA on vPSG.RSWA is quantified by formal PSG criteria such as the SINBAR (Sleep Innsbruck Barcelona) or AASM (American Academy of Sleep Medicine) criteria (for further detail see review by Cesari et al.).101RSWA may also occur in individuals without overt dream enactment behavior, so called "isolated RSWA."A proportion of these individuals may subsequently develop RBD, indicating the existence of a prodromal RBD state, which requires further study.101

10 CONCLUSIONBox 1 :○○○○
Disorders of sleep and arousal are common in LBD.They are disabling, with considerable caregiver burden, but often have identifiable causes and many have effective therapies.Therefore, an understanding of sleep-wake circuits and a targeted sleep assessment is an essential part of the LBD clinic.There are also significant gaps in LBD-specific therapies for sleep/arousal problems, and remaining questions regarding LBD-related pathophysiology in each disorder.Unanswered?? questions remain regarding the neuro-physiological/anatomical substrates for insomnia, EDS, and RBD, while the role of neurodegeneration in the occurrence of OSA and RLS requires further study.Work is needed to validate currently used therapies in the LBD population and discover novel therapeutic options for improving sleep and arousal in LBD.Key questions highlighted in this review which may be answerable in future studies -Testing the inter-relationship between deficits in arousal and attention in LBD: Examine degree to which objective measures of sleepiness (e.g., MSLT, Maintenance of Wakefulness Test) correlate with report-and task-based evidence of attentional deficits.Test whether EEG monitoring can detect evidence of sleep intrusions (i.e., microsleeps) during apparent attentional lapses.-Identifying functional neuroanatomical substrates for EDS: Neuropathological and neuroimaging studies examining association of dysfunction and degeneration of cholinergic (basal forebrain [BF]), noradrenergic (LC), GABAergic (BF), and dopaminergic (vPAG) nuclei with EDS.-Establishing an accurate prevalence estimate for RLS in PDD and DLB: ○ Well-controlled, long-term, prospective observational studies.-Characterizing changes in circadian physiological patterns in LBD: Longitudinal monitoring using actigraphy, temperature sensors, heart rate, and blood pressure monitoring as well as serial saliva sampling, sweat, and (pinprick) blood analysis over 24 hours.-Characterizing the structural, neurochemical, and functional networks underpinning iRBD and RBD in LBD: ○ Longitudinal structural and functional imaging studies and quantitative assessment of RBD symptomatology.○ Neuropathological studies to better characterize patterns of degeneration in structures implicated in RBD pathophysiology including the SLD and VMM.-Identifying reliable markers of α-synucleinopathic diseases and symptom trajectory in iRBD to facilitate early treatment trials.Potential examples include: 209 10808Gilat et al.108collated results from all published clinical observations of clonazepam or melatonin treatment. For conazepam, results from 1 RCT, 3 observational studies, 16 retrospective medical histories, and 31 case reports showed "clear" or "partial" improvements in RBD symptoms in 82.2% of patients (843/1026 patients).For melatonin, results from 3 RCTs,