An overview of the effects of levodopa and dopaminergic agonists on sleep disorders in Parkinson’s disease

Abstract

Sleep disorders are among the most common nonmotor symptoms in Parkinson’s disease and are associated with reduced cognition and health-related quality of life. Disturbed sleep can often present in the prodromal or early stages of this neurodegenerative disease, rendering it crucial to manage and treat these symptoms. Levodopa and dopaminergic agonists are frequently prescribed to treat motor symptoms in Parkinson’s disease, and there is increasing interest in how these pharmacological agents affect sleep and their effect on concomitant sleep disturbances and disorders. In this review, we discuss the role of dopamine in regulating the sleep–wake state and the impact of neurodegeneration on sleep. We provide an overview of the effects of levodopa and dopaminergic agonists on sleep architecture, insomnia, excessive daytime sleepiness, sleep-disordered breathing, rapid eye movement sleep behavior disorder, and restless legs syndrome in Parkinson’s disease. Levodopa and dopaminergic drugs may have different effects, beneficial or adverse, depending on dosing, method of administration, and differential effects on the different dopamine receptors. Future research in this area should focus on elucidating the specific mechanisms by which these drugs affect sleep in order to better understand the pathophysiology of sleep disorders in Parkinson’s disease and aid in developing suitable therapies and treatment regimens.

Citation:

Scanga A, Lafontaine A-L, Kaminska M. An overview of the effects of levodopa and dopaminergic agonists on sleep disorders in Parkinson’s disease. J Clin Sleep Med. 2023;19(6):1133–1144.

INTRODUCTION

Parkinson’s disease (PD) is the second-most-common neurodegenerative disorder¹ and is characterized by progressive neurodegeneration. PD is most commonly known for its cardinal motor features resulting from dopamine depletion. However, nonmotor symptoms, including mood disorders, autonomic dysfunction, sensory deficits, and sleep disturbances are now recognized as major contributors to disease symptomatology and continue to gain research interest. Sleep disturbances in PD can manifest as changes in normal sleep architecture, insomnia and excessive daytime sleepiness (EDS), or specific sleep disorders including sleep-disordered breathing (SDB), rapid eye movement sleep behavior disorder (RBD), and restless legs syndrome (RLS).² Sleep disturbances affect approximately 90% of patients with PD³ and have been linked to cognitive decline.⁴ Moreover, sleep problems negatively affect health-related quality of life, beginning during the early stages of PD.⁵ Thus, the management of these symptoms is relevant and necessary in order to reduce disease burden and improve patient well-being. This article reviews the effects of levodopa and dopaminergic agonists (DAs) on sleep disorders in PD and provides an overview of the impact of neurodegeneration on the development of sleep issues in PD.

REVIEW

Sleep disturbances in PD

Sleep disturbances are among the most common nonmotor symptoms.^6,7 In PD, sleep disruptions are considered multifactorial and are influenced by factors including motor and nonmotor symptoms, autonomic dysfunction, circadian dysfunction, and iatrogenic insult.⁸ Sleep disorders commonly occur in the prodromal phase of PD, before the onset of motor symptoms, and tend to progress with advancing stages of PD.² A primary cause is the progressive neurodegeneration of sleep/wake centers.⁹

The neuropathological hallmark of PD includes the degeneration of dopaminergic neurons in the substantia nigra pars compacta and dopaminergic depletion of the nigrostriatal pathways.¹⁰ Dopamine is a catecholamine involved primarily in arousal and promoting wakefulness, specifically through cortical activation and behavioral arousal.¹² The ventral tegmental area and the mesencephalic substantia nigra pars compacta are the major dopaminergic nuclei.¹³ Dopaminergic cells in the ventral tegmental area and substantia nigra pars compacta have efferent and afferent projections to and from different brain structures involved in the control of sleep and wakefulness, including the dorsal raphe nucleus, the pedunculopontine and laterodorsal tegmental nuclei, the locus coeruleus, the lateral and posterior hypothalamus, the basal forebrain, and the thalamus.¹⁴ Dopaminergic neurons do not alter their mean firing rate across sleep–wake states¹⁵ but rather are thought to modify their tonic and phasic pattern of discharge in accordance with inputs from serotonergic, cholinergic, orexinergic, glutamatergic, and noradrenergic neurons in the abovementioned brain regions.¹³

Extranigral structures including in the Braak¹⁰ and Sandyk¹¹ are important in the pathology of sleep disruptions in PD. In a postmortem study, PD patients with sleep disturbances exhibited greater pathological changes in specific regions of the brainstem and the hypothalamus than PD patients without sleep disturbances.¹⁶ Degeneration in hypothalamic regions such as the ventrolateral preoptic^17,18 and median preoptic areas¹⁹ may lead to insomnia and sleep disturbances in PD.⁸ Neurons projecting from these regions innervate and inhibit major monoamine arousal systems leading to the generation and promotion of sleep. Thus, pathological alterations in the hypothalamus can alter sleep–wake states and circadian rhythms and decrease consolidated sleep¹⁷ in PD.²⁰

While dopamine is one of many neurotransmitters playing a crucial role in the sleep–wake cycle, sleep disturbances can also result from degeneration of nondopaminergic pathways including serotonergic,²¹ noradrenergic,²² and cholinergic.²³ Additionally, EDS in PD may in part be attributed to a deficiency in hypocretin-containing neurons.²⁴ A loss of hypocretin neurons is involved in hypersomnolence and the pathogenesis of narcolepsy. In some instances, PD patients with EDS can experience sleep attacks, a narcolepsy-like symptom.²⁵ Other specific neurological regions affected in PD include the noradrenergic locus coeruleus and subcoeruleus complex,²² which are involved in sleep–wake state regulation, particularly in promoting arousal and vigilance,²⁶ with relevance to cognition.²⁷ Locus coeruleus neurons are in turn affected by sleep disturbances.^28–30 In addition, these neurons have been found to be involved in the pathogenesis of RBD. Neuroimaging has shown decreased signal intensity in the abovementioned regions in patients with PD and RBD.³¹

The brainstem is home to important components of respiratory control, including the respiratory pacemaker and central chemoreceptor.³² Thus, SDB in PD may be attributed to the degeneration of neurons in specific brainstem structures involved in the control of ventilation.³³

As per the Braak hypothesis, which suggests that neurodegeneration in PD begins in lower brainstem structures and follows an ascending pathway,¹⁵ brainstem regions could be among the first to be affected by neurodegeneration in PD,¹⁵ potentially acting as a predisposition to the development of sleep disorders. However, Braak’s hypothesis is likely not the only way by which this neurodegenerative process progresses. A postmortem study by Kalaitzakis et al found that lower brainstem structures were not affected in a subset of PD cases, indicating that different trigger sites and progression patterns of PD exist,³⁴ which could also account for differences in sleep-related manifestations.

Dopamine receptors and sleep regulation in animal models

Much of what is known about the effects of the different dopaminergic receptors in sleep comes from animal data. In rats, administration of D1 and D2 receptor agonists induces wakefulness and reduces slow-wave sleep and rapid eye movement (REM) sleep.¹⁴ However, D2 receptors exhibit biphasic effects, such that a reduction in wakefulness and increases in slow-wave sleep and REM sleep have been observed with small doses of D2 receptor agonists, and the opposite has been observed with higher doses.¹⁴ Animal data have shown that levodopa and DAs facilitate sleep at lower doses but have the opposite effect at higher doses, through the biphasic action of presynaptic D2 receptors.^35,36 Similarly, low doses of the D2 receptor agonist quinpirole decreased wakefulness and increased sleep by acting presynaptically, whereas higher doses promoted wakefulness by acting postsynaptically.³⁷ In primate models, quinpirole was shown to increase sleep latency and wakefulness after sleep onset, exhibiting a deleterious effect on sleep parameters. The DA pramipexole, which has D2 and D3 receptor affinity, also exhibited a biphasic effect in rats with increased sleep at lower doses and greater alertness at higher doses.³⁸ The D1 receptor agonist SKF38393 increased slow-wave sleep, reduced periods of wakefulness after sleep onset, albeit insignificantly, and restored levels of REM sleep to baseline values in a primate model of PD.³⁹ This effect is contrary to what was previously found in other animal models.^40–42 Thus, further experimentation is required to elucidate the effects of the different dopaminergic receptors, particularly in humans.

Pharmacology of levodopa and DAs

Levodopa is the gold standard in the treatment of PD due to its effectiveness and ability to alleviate the hallmark motor symptoms. Levodopa is decarboxylated in the striatum to form dopamine, which stimulates dopaminergic receptors. This pharmacological agent has a half-life of 50 minutes; however, this can be increased to approximately 90 minutes when levodopa is administered with a decarboxylase inhibitor such as carbidopa.⁴³ The combination of levodopa and carbidopa enables levodopa to cross the blood–brain barrier without being metabolized prematurely by the gastrointestinal system.⁴⁴ Adverse gastrointestinal effects of levodopa are less frequent with the use of carbidopa.⁴⁵ Adverse drug reactions occurring with advancing PD include motor fluctuations and symptoms recurring due to the “on–off” phenomenon. After taking levodopa, patients will be in the “on” phase when symptoms are alleviated, but due its short half-life its effects begin to wear off after a few hours and patients experience the “off” phase when symptoms can become unpredictably more severe. Common “on” complications include a delay in the onset of symptomatic relief or a less-powerful relief of symptoms than usual upon intake of medication.⁴⁶ Levodopa-induced dyskinesia is a common but physiologically complex complication in PD,⁴⁷ occurring in 3%⁴⁸ to 94% of PD patients.⁴⁹ Levodopa-induced dyskinesia occurs when dopamine concentrations are at their maximum in the brain, termed peak-dose dyskinesia, which occurs in a dose-dependent fashion^47,50 and also depends on the methods of drug delivery.⁴⁷

Long-acting levodopa (LALD) has a longer half-life than regular levodopa, maintaining a greater plasma concentration over a greater amount of time.⁵¹ LALD exhibits the same safety and tolerability as immediate-release levodopa but also has similar rates of motor complications. Its role as an alternative for immediate-release levodopa remains unclear due to its less-predictable absorption and effect.⁵²

DAs are dopaminergic drugs that can be administered either in monotherapy or with levodopa, depending on the symptomatic profile of the individual patient.⁵³ DAs act directly to continuously stimulate the central dopamine receptors⁵⁴ in situations of insufficient endogenous dopamine such as PD. DAs can be divided into 2 groups: ergoline- and nonergoline-derived agonists. Common ergoline-class drugs are bromocriptine, cabergoline, pergolide, and lisuride,⁵⁵ all now rarely prescribed in PD due to their severe adverse complications including cardiovascular,⁵⁶ retroperitoneal,⁵⁷ and pleuropulmonary⁵⁸ fibrosis.⁵⁴

Currently, the widely used nonergot-derived DAs include pramipexole, ropinirole, apomorphine, and rotigotine. These drugs bind with high affinity to the D2-like dopamine receptor family and modestly interact with the D1-like family.⁵⁴ In contrast to levodopa, DAs are appreciated for their longer half-life and can be used to reduce motor fluctuations caused by levodopa therapy.⁵⁵

Pramipexole is among the therapies with the longest half-life, 12 hours.⁵⁵ It also binds with high affinity to D3 receptors found in the limbic system⁵⁵ and is often prescribed for unipolar and bipolar depression, making it a treatment option for PD patients with mood disorders.⁵⁹ Ropinirole is pharmacologically similar in many ways to pramipexole, although it has a shorter half-life of approximately 6 hours and exhibits a preferential affinity for D2 receptors. Adverse effects of both these drugs include orthostatic hypertension, dizziness, nausea, and somnolence.⁵³

Unlike pramipexole and ropinirole, apomorphine and rotigotine are not administered orally. Apomorphine is among the shortest-acting DAs, with a half-life of 30–60 minutes, and is administered subcutaneously. It is used in cases where patients exhibit uncontrollable “off” episodes.⁵⁵ Rotigotine is primarily a D2/D3 receptor agonist administered in the form of a daily transdermal patch. It has a biphasic half-life beginning with an initial distribution of 3 hours followed by 5 to 7 hours of continuous drug availability. Rotigotine is a valuable option for patients with issues adhering to their daily drug regimen.⁵⁵ Apomorphine and rotigotine may cause reactions at the site of administration, and other adverse effects include dizziness and orthostatic hypertension.⁵³ Treatment with DAs can also lead to the development of impulse control disorders.⁶⁰

Sleep architecture and objective sleep quality

In PD, progressive alterations in architectural parameters of sleep structure have been evaluated and confirmed in numerous studies, including changes in sleep efficiency, total sleep time, sleep fragmentation, prolonged sleep latency, and a reduction in the percentage of REM sleep.^61–63 Architectural sleep changes advance progressively with disease duration,⁶¹ indicating that the neuropathological process of PD may play a pivotal role in structural sleep changes. When compared to healthy control patients, drug-naïve patients with PD present a significantly reduced sleep efficiency, a greater sleep latency, and decreased time in REM sleep, supporting the hypothesis that the neurodegenerative process of PD contributes to changes in sleep architecture.⁶⁴

Conflicting evidence exists regarding the effect of levodopa and DAs on sleep architecture. In their early study, Nausieda et al⁶⁵ concluded that sleep disruptions were associated with chronic antiparkinsonian therapy because of the increased prevalence of patient-reported sleep disturbances with longer treatment duration.⁶⁵ Furthermore, the onset of antiparkinsonian treatment in early PD patients has been linked to a statistically significant increase in awakenings and decrease in stage 1 non-REM sleep evidenced by polysomnography (PSG) recordings.⁶⁶ However, in some studies, PSG data exhibited no significant correlation with levodopa or DA dosage, suggesting that dopaminergic therapy does not result in sleep “destructuring.”^61,67 Moreover, Ferreira et al found that in drug-naïve patients started on levodopa for 2 months prior to undergoing a sleep study, sleep efficiency improved from 75.4% to 86.4% (P = .012), with improvements in wakefulness after sleep onset and sleep latency noted as well.⁶⁴ This was attributed to improved motor symptoms rather than direct changes in sleep architecture. Improvement of sleep fragmentation with dopaminergic treatment was previously shown to be related to normalization of muscle activity in sleep.⁶⁸ A study assessing the effect of LALD on microstructural sleep parameters found that this preparation of levodopa had no significant effect on objective sleep, that is, LALD did not cause alterations in sleep architecture, but it also did not reverse sleep disturbances or rescue sleep structure.⁶⁹

In a recent meta-analysis it was found that the levodopa equivalent daily dose (LEDD) contributed significantly to the heterogeneity of PSG changes in PD and that increased LEDD was associated with increased wakefulness after sleep onset and REM latency and decreased total sleep time.⁷⁰ Self-reported and objective sleep parameters assessed by SCOPA-SLEEP questionnaire and PSG, respectively, worsened with increasing dosage of dopaminergic medication taken within 4 hours of bedtime.⁷¹ However, contradictory evidence was reported when the effects of levodopa and DAs were compared separately. While levodopa was not significantly associated with altered sleep macrostructure in a cohort of 351 PD patients, DAs were associated with more awakenings (P < .001) and decreased time spent in REM sleep (P = .012).⁷² However, rotigotine resulted in improved objective quality of sleep with a 10% increase in sleep efficiency, decreased wakefulness after sleep onset and sleep latency, and increased REM sleep compared to placebo.⁷³ Although improved sleep through improved motor symptom control remains a possibility, the authors suggest a direct effect on D1 receptors,⁷³ which have been shown to be involved in sleep mechanisms in a primate PD model.³⁹

Overall, the effects of levodopa and DAs on sleep architecture appear to be variable in PD, depending on factors including PD disease duration, the dosage of medication used, and the specific pharmacodynamic profile of levodopa formulation or DA. As evidenced experimentally, the dose of levodopa and DAs is critical in both sleep promotion and sleep disruption. Thus, further experiments should focus on determining the optimal dose, formulation, and pharmacologic profile that can maintain consolidated sleep while alleviating motor symptoms.

Subjective sleep quality and insomnia

Insomnia, the most prevalent sleep disorder among PD patients,⁷⁴ refers to difficulties in sleep initiation or sleep maintenance or early awakenings.⁷⁵ Subjective sleep quality and insomnia are related. Patients with PD experiencing sleep dysfunction and insomnia report issues primarily regarding sleep maintenance and consolidation.⁷⁶ Insomnia can be caused by other sleep disorders such as RLS, particularly for sleep-onset insomnia, and obstructive sleep apnea (OSA)⁷⁷ as well as depression,⁷⁸ pain, and nocturia. Patients experiencing insomnia often report daytime sleepiness⁷⁹ that alters their ability to perform daily tasks.

Levodopa may affect subjective sleep quality.⁶⁷ Increasing doses of levodopa administered chronically over time in PD negatively correlate with subjectively assessed sleep quality^67,80,81 and sleep maintenance.⁶⁷ The motor complications of levodopa can lead to a reduction in subjective sleep quality out of proportion to PSG-measured sleep architecture effects related, for example, to akinesia, cramps, dystonia, and pain related to “off” periods.⁶⁷ However, LALD administered at bedtime appears to improve nocturnal akinesia and increase total sleep time but has no effect on sleep fragmentation, sleep initiation, and sleep maintenance. It can also reduce rigidity and promote sleep in some patients.^52,82 Conversely, in cross-sectional analyses, Schaeffer et al⁸⁰ found that LALD may not be adequate in reducing nocturnal akinesia and impaired subjective quality of sleep and results in increased perception of nocturnal akinesia.⁸⁰ Whereas some patients may benefit from LALD to control nocturnal motor symptoms, others may not benefit sufficiently from it to experience better subjective sleep quality.

In a longitudinal study, patients with early PD who reported insomnia had a greater overall LEDD in comparison with those who did not experience sleep disturbances, indicating a potential correlation.⁷⁴ Insomnia severity assessed in PD using the Insomnia Severity Index⁸³ was not associated with LEDD.⁸⁴ Therefore, although LEDD may be correlated with the presence of insomnia it may not necessarily correlate with insomnia severity. Moreover, sleep maintenance insomnia, in a cohort of 182 drug-naïve PD patients, increased by 50% from baseline after the introduction of DAs.⁷⁵ Other investigators have established that treatment with DAs may be an independent risk factor for insomnia.⁷⁸

Three randomized controlled trials observing the effects of prolonged-release ropinirole,⁸⁵ rotigotine,⁸⁶ and pramipexole,⁸⁷ either together with levodopa or as monotherapy in early PD, reported insomnia as one of the most common adverse events experienced by participants in the groups receiving the intervention. However, a later randomized controlled trial testing rotigotine in a cohort of advanced PD patients reported no difference in reported insomnia between PD patients who received the intervention and those who did not.⁸⁸ The difference in study population, early vs late PD, might account for different effects. In the RECOVER trial, rotigotine improved nocturnal PD symptoms assessed by the revised Parkinson’s disease sleep scale.^89,90 Moreover, other randomized controlled trials using rotigotine and prolonged-release ropinirole showed an improvement in the general quality of self-reported sleep evidenced by significant improvements in Parkinson’s disease sleep scale scores.^73,91,92

Insomnia is multifactorial, and PD drug regimens may differentially influence sleep depending on the balance of beneficial vs adverse effects in a given individual. Once motor disturbances are alleviated with levodopa and/or DAs, including throughout the night, patients may be more likely to experience fewer nocturnal disturbances, reducing symptoms of insomnia and ameliorating the perceived quality of sleep. However, motor fluctuations and other adverse effects may lead to poor sleep outcomes. Side effects can also include hallucinations, which are more prevalent in PD patients taking DAs,⁹³ and may aggravate any insomnia experienced. A direct drug effect promoting alertness may also impair sleep.

In treating insomnia, nonpharmacological approaches should be favored, including optimization of sleep hygiene and cognitive behavioral therapy.⁹⁴ Pharmacological treatment options are limited and associated with potential adverse effects (Table 1). Research to further our understanding of the specific etiologies contributing to the development of poor sleep quality and insomnia in individuals with PD could aid in developing targeted and personalized therapies that will leave patients with more consolidated sleep overall.

Table 1.

Summary of the pharmacological treatment options and effects of levodopa and dopaminergic agonists on sleep disorders in Parkinson’s disease.

Sleep Disorder	Pharmacologic Treatments	Overall Effect of Levodopa (Better or Worse)	Overall Effect of DAs (Better or Worse)
Insomnia	Eszopiclone* (off label)94,176 Melatonin* (off label)94,176	Better	Worse
Excessive daytime sleepiness	Modafinil (possibly useful)176 Caffeine (off label)176 Solriamfetol177 Sodium oxybate (off label)178 THN102 (investigational)179	Worse	Worse
Sleep-disordered breathing	No pharmacological treatment recommended	Better (under investigation)	No effect
REM sleep behavior disorder	Clonazepam176 Melatonin176	Worse	Better
Restless legs syndrome	Alpha2-delta ligands180: Gabapentin enacarbil, pregabalin, gabapentin Dopaminergic agonists180: Pramipexole, ropinirole, rotigotine Opioids180: Prolonged-release oxycodone-naloxone	Better	Better in short term but risk of augmentation (second-line treatment)†
Circadian rhythm disorders	Melatonin‡181,182 (off label) Melatonin agonists (investigational)§183 REV-ERBα agonists (investigational)¶184	Worse185,186	Worse186

*Drugs with the greatest supporting evidence for treating insomnia in Parkinson’s disease. Other pharmacological agents such as zolpidem, ramelteon, trazodone, and doxepin may be useful but there is insufficient evidence. Safety issues may also arise such as risk of falls. †Doses of dopaminergic agonists should be kept low and other treatments should be considered before additional dopaminergic agonists are prescribed. ‡Melatonin has not been directly assessed as a modulator of the circadian rhythm in Parkinson’s disease. Rather, clinical trials have observed the effects of melatonin on sleep–wake cycle alterations. §Experiments performed in vivo in animal models only and assessed the effects of a melatonin agonist on age-related changes of the circadian rhythm. ¶Experiment performed in vivo in animal models only and assessed the ability of synthetic REV-ERBα agonists to regulate circadian behavior. Das = dopaminergic agonists, REM = rapid eye movement.

EDS

EDS is classified as a state of daytime hypersomnolence where daily, unintentional sleep episodes occur inappropriately.⁹⁵ A recent meta-analysis found that based on 59 analyzed studies the estimated prevalence of EDS in PD was 35.1%, which represents a higher prevalence in PD than in the general population.⁹⁶ EDS can pose a threat to the safety of patients, particularly while they are driving.⁹⁷ The pathophysiology of EDS in PD is multifactorial, and evidence suggests that it can arise because of the nigrostriatal dopamine depletion occurring in the neurodegenerative process of PD,⁹⁸ coexisting sleep disorders, nocturnal motor symptoms,^99,100 and the use of ergot¹⁰¹ and nonergot DAs.^102,103

Routine use of DAs in PD resulted in the observation of a severe adverse effect: sleep attacks. Eight PD patients on pramipexole or ropinirole reported falling asleep while driving without feeling any drowsiness or fatigue.¹⁰⁴ The same effect of hypersomnolence while driving was reported in 3 other patients with PD on bromocriptine, lisuride, and piribedil, ergot DAs.¹⁰¹ Paus et al¹⁰⁵ surveyed 2,952 PD patients regarding the occurrence of sleep attacks or sleep episodes and found that this phenomenon occurred with all DAs available on the market at the time and with levodopa monotherapy. The latter was associated with the lowest risk of developing sleep attacks, whereas the greatest risk was associated with combination therapy of levodopa and DAs. Treatment with DAs was one of the main determining factors of sleep attacks in this cohort.¹⁰⁵ However, this finding has been challenged by other investigators arguing that neither pramipexole nor levodopa monotherapy makes patients with PD more susceptible to sleep attacks.¹⁰⁶ Amara et al longitudinally assessed EDS in PD patients over 3 years and determined that at years 2 and 3 the total LEDD was significantly higher in PD patients with EDS, setting forth the possibility of a dose-dependent association between EDS and total LEDD.¹⁰⁷ Notably, consistent evidence suggests that levodopa monotherapy confers less of a risk of somnolence in comparison with DAs alone and when compared with DA and levodopa combination therapy.^108,109 However, levodopa may exert sedative effects contributing to EDS and sleep episodes in patients with PD.¹⁰¹ Other investigators have determined that the dosage of DAs is associated with EDS but the dose of levodopa is not.¹¹⁰ Although there is conflicting evidence regarding levodopa and increased somnolence, the potential sedative effects of levodopa and the sedative effects of DAs indicate that somnolence may occur due to a drug class effect.¹¹¹ Bliwise et al¹¹² have challenged this idea, however, suggesting that daytime alertness is differentially affected by DAs and levodopa: Increasing doses of DAs resulted in reduced daytime alertness whereas higher doses of levodopa increased daytime alertness, as measured using the Maintenance of Wakefulness Test.¹¹³ The authors thus proposed a divergent, dose-dependent effect of drug class on daytime alertness.¹¹² This is different from animal data where higher doses of DA led to greater alertness, as discussed above. Differences may relate to lower dosages per kilogram used in the Bliwise study in humans compared to the animal studies, and possibly patient/PD-related factors.

EDS can also be attributed to the neurodegeneration occurring in PD. EDS is frequent in PD even before treatment initiation¹⁰³ and associated with longer PD duration.⁹⁶ Mean sleep latency from the Multiple Sleep Latency Test, an objective sleepiness measure,¹¹⁴ did not differ significantly between the rotigotine group and the placebo group in a randomized controlled trial during a 4-month period,¹¹⁵ suggesting this DA had no effect on EDS. EDS has been found to be related to Dopaminergic Nigrostriatal degeneration in PD, as assessed using dopamine transporter scanning.⁹⁸ Recently, a positron emission tomography imaging study found reductions in hypothalamic D3 receptor availability to be associated with EDS, irrespective of LEDD.¹¹⁶ This is consistent with previous studies that found daytime sleepiness was not correlated with sleep disorders or with levodopa or DAs but rather appeared to be a consequence of the pathology of PD.¹¹⁷ It has been suggested that PD patients with greater neurodegeneration, such as has been identified in ascending arousal systems, may be more susceptible to the effect of dopaminergic therapy in promoting EDS.¹¹⁶

Insomnia and EDS are both associated with DAs.⁷⁷ DAs, as well as levodopa, may promote insomnia in several ways. They increase the risk of hallucination, which can cause sleep disturbances.^118,119 Motor fluctuation such as “off” periods related to medication can also lead to poor sleep. Moreover, as discussed above, DAs have a biphasic effect on D2 receptors such that higher doses may lead to insomnia whereas lower doses could exert a soporific effect.^36,37,78 Both EDS and insomnia in PD can be caused by disturbed sleep from motor or other symptoms, including other sleep disorders, and neurodegeneration. Specific alterations in sleep/wake circuitry may result in different clinical manifestations. For example, hypothalamic D3 receptors appear to be involved in EDS but not insomnia.¹¹⁶ Consequently, the manifestations of insomnia and EDS in a given patient might result from the interaction of the specific neurodegeneration pattern and medication type, dose, and method of delivery.

SDB

SDB in PD can present in different forms, the most prevalent form being OSA, which affects between 20% and 60% of all PD patients.^120–122 It is suggested that OSA in PD can develop and worsen due to factors such as upper airway motor instability and impaired ventilatory control.³² Whether OSA is more prevalent in PD than in the general population remains a topic of debate. Nevertheless, OSA results in fragmented sleep and intermittent hypoxemia, which can exacerbate symptoms of poor sleep, EDS, and cognitive dysfunction, greatly impairing quality of life.³² In our group’s work and that of others, greater cognitive dysfunction was found in PD patients with OSA and greater cognitive impairment was positively associated with increasing OSA severity.^123,124 Furthermore, treatment of OSA in PD using continuous positive airway pressure was found to improve nonmotor symptoms, reduce global cognitive dysfunction¹²⁵ but not specific cognitive domains,¹²⁴ and reduce daytime somnolence.¹²²

The effect of levodopa and DAs on SDB in PD has not been well-described. Vincken et al reported a positive effect of levodopa on upper airway obstruction in 1 PD patient. Spirometry was performed on this patient before intake of levodopa, immediately after, and approximately 1 hour later. The flow rate increased substantially after levodopa treatment, in parallel with parkinsonian symptoms and dyspnea improvement. The authors thus showed that upper airway muscles are involved in disorders affecting motor activity, such as PD.¹²⁶ In a cohort of 21 PD patients, 5 were found to have evidence of upper airway obstruction after levodopa withdrawal. Reintroduction of levodopa resulted in improvement in upper airway obstruction.¹²⁷ However, airway obstruction may not ameliorate following levodopa treatment and may be due to comorbid chronic obstructive pulmonary disease.¹²⁸ Additionally, it has been postulated that the DA apomorphine can improve dysfunction of the musculature in the airway, potentially reversing obstruction that could result in OSA.^127,129 Interestingly, irregular and tachypneic breathing has been reported as a symptom of levodopa-induced dyskinesia in 2 PD patients. The authors postulated that, in some patients, respiratory dysfunction induced by levodopa may be a result of neurodegeneration of dopaminergic neurons in the respiratory control centers. Dopamine is involved in both the peripheral and central circuitry of respiration. Exogenous dopamine may result in denervation hypersensitivity of peripheral chemoreceptor dopaminergic neurons, causing levodopa-induced respiratory dyskinesia.¹³⁰ Moreover, the degeneration of dopaminergic neurons in the brainstem that control breathing and compromised dopamine receptor function may also contribute to the development of levodopa-induced respiratory dysfunction.¹³¹

Higher LEDD is associated with less-severe OSA.¹³² A retrospective study conducted by Valko et al¹³³ compared patients with PD with or without SDB. Patients in the group with PD and SDB were divided further into cohorts of central and obstructive SDB predominance. Here, AHI was calculated using overnight PSG, where central apnea was characterized by an absence of respiratory effort after the event and obstructive apnea was characterized by the presence of respiratory effort after the event. Patients with central SDB predominance (n = 7) had a greater overall LEDD than patients with obstructive SDB predominance (n = 50). Moreover, 100% of patients exhibiting central SDB were on dopaminergic treatment, compared to only 56% with obstructive SDB (P = .03). The authors suggested that central SDB predominance may be more likely to occur in patients on both levodopa and DAs; however, further studies with a larger sample size are required to confirm this finding. Additionally, Valko et al speculated that PD patients with central SDB may be predisposed to sleep attacks, considering EDS was highly prevalent in the central SDB group. Interestingly, in patients with PD and SDB treated with DAs, the authors observed a significant decrease in the AHI during REM sleep that occurred exclusively in this group. The authors attributed this lower REM SDB severity to the close interplay between the neurodegenerative processes affecting REM-sleep-related structures in the brainstem and the effect of DAs.¹³³ While important observations were made in this study, the exact effect of DAs and levodopa on SDB in PD remains ambiguous and warrants further investigation, considering the impact of SDB on cognition, among other clinical variables.

Recently, our group found that patients with PD taking LALD before bedtime had a lower incidence of OSA events with a lower AHI and fewer respiratory arousal events compared to patients not taking LALD. These results suggest that LALD might be a potential treatment for OSA in some PD patients. Moreover, DAs were not significant predictors of OSA in this cohort and adjusting statistical models for LEDD did not alter results, suggesting that DAs do not affect OSA in PD.¹³⁴ A randomized controlled trial assessing the effect of LALD on OSA in PD is ongoing.

RBD

RBD is characterized by the loss of muscle atonia during REM sleep accompanied by enactments of dreams with or without vocalizations,⁹⁵ which can be bothersome for patients and family members. The incidence of RBD increases throughout disease progression,¹³⁵ and it is estimated that up to 50% of PD patients will eventually be diagnosed with RBD.^136,137 RBD may be a prodromal feature of PD. The loss of dopaminergic midbrain neurons is associated with idiopathic RBD. Consistent with the hypothesis that RBD precedes disorders involving nigrostriatal dopamine depletion such as PD,¹³⁸ up to 82% of patients diagnosed with idiopathic RBD develop a neurodegenerative syndrome including parkinsonism after 12 years, on average, of follow-up.¹³⁹

Patients with PD frequently experience psychosis, most commonly in the form of visual hallucinations¹⁴⁰—a phenomenon that has been investigated with RBD in PD.^141,142 RBD in PD is significantly related to the development of hallucinations. RBD and visual hallucinations likely have overlapping neuropathological mechanisms, including cholinergic dysfunction and degeneration.¹⁴³ The latency with which psychosis in PD develops depends on differing clinical correlates. In early-onset psychosis, RBD was more frequently present than in patients with late-onset psychosis. Moreover, the development of hallucinations was found to be correlated with the dosage of DA therapy.¹⁴¹ Later onset of psychosis in particular correlated significantly with LEDD. Generally, as disease duration increases and PD progresses, there will be an increase in LEDD. Therefore, this correlation may be influenced more so by disease duration rather than by the actual drug effect.¹⁴⁴

LEDD was found to be among the statistically significant clinical correlates of RBD when comparing patients with PD who either did or did not have RBD.¹⁴⁵ Longer disease duration was also a significant clinical correlate, further contributing to the likelihood that the association of LEDD and RBD may be confounded by disease duration. A recent preliminary study¹⁴⁶ has confirmed the finding that LEDD is associated with RBD symptoms. Meloni et al sought to evaluate how dopamine replacement therapy affects the presence of symptomatic RBD in a cohort of 250 patients with PD. Intriguingly, levodopa therapy was directly and positively associated with RBD severity, as measured with the RBD Screening Questionnaire, after adjusting for age and PD severity.¹⁴⁶ Unlike levodopa, DA therapy was not associated with RBD Screening Questionnaire scores.¹⁴⁶

The potential prodromal nature of RBD in PD further points toward an association with dopaminergic loss. Patients with idiopathic RBD appear to experience progressive nigrostriatal dopaminergic dysfunction.¹⁴⁷ Hence, DAs have been evaluated in the treatment of RBD in patients with and without PD. In 10 patients with RBD only, pramipexole was highly efficacious, with 89% of these patients reporting that RBD symptoms were moderately reduced or completely resolved.¹⁴⁸ These results were in line with other findings indicating that clinical manifestations of RBD decreased in both frequency and intensity,¹⁴⁹ as well as the severity of RBD symptoms,¹⁵⁰ after treatment with pramipexole. However, Kumru et al assessed the effect of pramipexole in a cohort of PD patients undergoing PSG and found that while parkinsonism improved in all patients the frequency and severity of RBD symptoms remained unchanged.¹⁵¹

Rotigotine may be beneficial for RBD symptoms in PD, as reported in an exploratory, open-label study assessing RBD-related symptoms in PD subjectively and objectively using the RBD Questionnaire – Hong Kong¹⁵² and PSG, respectively.¹⁵³ The preliminary findings of this study present an improvement in the severity of RBD symptoms with rotigotine.

The presented studies on DAs in RBD are observational, lack a control group, and involve small sample sizes. Furthermore, studies evaluating idiopathic RBD^148–150 differ from the study of Kumru et al,¹⁵¹ in which secondary RBD (RBD present with another neurological disorder) was assessed. Idiopathic RBD and secondary RBD may present distinct stages of this disorder, potentially changing response to treatment.¹⁵⁴

The potential improvement of RBD symptoms in PD with the administration of rotigotine suggests that a potential relationship exists between dopamine and the pathogenesis of RBD.¹⁵³ Likewise, Meloni et al speculate that the lack of association between selective D2/D3 receptor agonists and RBD scores implicates the involvement of D1 receptor activation in the underlying pathophysiology of RBD in PD.¹⁴⁶ Indeed, REM sleep is modulated by D1 receptor activation,⁴⁰ a hallmark of rotigotine’s mechanism of action, as confirmed by in vivo modeling.¹⁵⁵ Randomized controlled trials evaluating the effects of rotigotine in PD patients with RBD should be employed to confirm these observations,¹⁵³ and further experimentation is required to clarify these neurological mechanisms.

RLS

RLS is a sleep-related movement disorder characterized by waking dysesthesia and an urge to move the legs that occurs at rest and primarily in the evening and at night, resulting in delayed sleep onset. These symptoms can be relieved by movement.⁹⁵ They may be accompanied by periodic limb movements during sleep. RLS is commonly diagnosed in PD patients, with its prevalence ranging from 0–52%.¹⁵⁶ RLS in PD is typically associated with an older age of onset, a lower rate of family history, and a lower periodic limb movement index on PSG.¹⁵⁷

It is postulated that RLS is an early clinical feature of PD; however, recent evidence suggests that symptoms of RLS do not predate a new diagnosis of parkinsonism.¹⁵⁸ Moreover, the pathogenesis of RLS in PD likely differs from that of idiopathic RLS.¹⁵⁹ Dopamine dysregulation, not dopamine deficiency, may be the predominant mechanism underlying idiopathic RLS,¹⁶⁰ differing from the hypodopaminergic state of PD. The lack of association between RLS and periodic limb movements in PD also suggests different pathophysiology in PD.¹⁶¹

Treatment of RLS should initially focus on correction of exacerbating factors, including iron deficiency.¹⁶² Pramipexole,^163,164 ropinirole,¹⁶⁵ and rotigotine¹⁶⁶ are all approved for the treatment of RLS¹⁶⁷ and have shown considerable efficacy and safety.¹⁶⁸ However, long-term use of these drugs often leads to augmentation: the apparent worsening of RLS symptoms related to RLS treatment symptoms.¹⁶⁹ Additionally, the risk of impulse control disorders with DAs also remains an issue in RLS treatment. Augmentation also often occurs with levodopa treatment.¹⁶⁷ Thus, recent guidelines indicate that alpha₂-delta ligands should be the first-line agents used to treat chronic persistent RLS and DAs should only be prescribed if alpha₂-delta ligands are contraindicated.¹⁷⁰

Maestri et al have reported that extended-release formulations of DAs may resolve the issue of augmentation due to their longer half-life.¹⁷¹ There appears to be an inverse relationship between drug half-life and the development of augmentation. In a cohort of 24 patients with RLS, augmentation was resolved after switching from immediate-release DA treatment to extended-release pramipexole.¹⁷¹ However, further investigation is required to determine if the risk of augmentation is reduced using extended-release DAs, including in PD.

The pharmacological treatment of RLS in PD aligns with that of idiopathic RLS. While RLS in PD can be treated by optimizing and adjusting dopaminergic drug regimens to alleviate symptoms, the same adverse effects exist as in idiopathic RLS.¹⁷² Small doses of pramipexole and ropinirole have shown great efficacy and tolerability. In the double-blind, placebo-controlled RECOVER trial, rotigotine showed a significant overall improvement of sleep complaints in PD (assessed by the revised Parkinson’s disease sleep scale), including on specific questions regarding RLS symptoms such as the urge to move arms and legs and restlessness felt in the arms and legs.⁹⁰ In patients already on dopaminergic therapy, it has been proposed that the schedule of levodopa intake can be altered to optimize its effectiveness in alleviating RLS, with administration of an evening dose.¹⁷² In older PD patients with greater cognitive dysfunction, DAs can lead to confusion and hallucinations and, in rarer cases, impulse control issues.¹⁷² Levodopa could then be considered over DAs. Nonetheless, to reduce the risk of augmentation and avoid increasing what may be an already high LEDD for some PD patients, it is recommended that other treatment options be thoroughly discussed and considered, namely alpha₂-delta ligands.¹⁷³ Studies have shown that subthalamic nucleus deep brain stimulation can also reduce RLS symptoms in PD,^174,175 although this requires further investigation.

CONCLUSIONS

Sleep disturbances and disorders are prevalent nonmotor symptoms warranting safe and efficacious treatments to prevent cognitive decline and improve quality of life. The development of sleep issues in PD is likely multifactorial and influenced by the complex interplay of neurodegeneration, disease progression, and iatrogenic insult. Levodopa and DAs continue to be effective in the treatment of motor symptoms, and preliminary studies suggest they may be beneficial for certain sleep disorders such as OSA and RBD. However, these drugs can pose a risk for developing sleep disturbances including daytime somnolence, a common adverse effect of DAs. Moreover, insomnia may also be an adverse effect associated with DA use. Insomnia can result from a multitude of adverse events in PD such as levodopa-induced dyskinesia and nocturnal motor complications and thus may be likely to improve when motor symptoms are alleviated by treatment. Overall, larger randomized controlled trials and longitudinal studies are necessary to evaluate the effects of levodopa and DAs on sleep and sleep issues in PD. Additional experimental evidence is required to elucidate the specific mechanisms of sleep disorders in PD, and, consequently, develop a greater understanding of how levodopa and dopaminergic drug therapy may influence these neuropathological mechanisms.

ABBREVIATIONS

DA,

dopaminergic agonist

EDS,

excessive daytime sleepiness

LALD,

long-acting levodopa

LEDD,

levodopa equivalent daily dose

OSA,

obstructive sleep apnea

PD,

Parkinson’s disease

PSG,

polysomnography

RBD,

rapid eye movement sleep behavior disorder

REM,

rapid eye movement

RLS,

restless legs syndrome

SDB,

sleep-disordered breathing

DISCLOSURE STATEMENT

All authors have seen and approved the manuscript. Work for this study was performed at the Respiratory Epidemiology and Clinical Research Unit, Research Institute of the McGill University Health Centre, Montréal, Québec, Canada. A.-L.L. reports advisory board honoraria from Abbvie, Sunovion, Paladin Labs, and Merz and speaker honoraria from Sunovion and Paladin Labs. The other authors report no conflicts of interest.