Most patients with PD report insufficient sleep as a result of frequent awakenings, and less commonly, early-onset insomnia and early-morning awakening [65, 70, 77, 119]. In patients with PD, poor and reduced sleep causes severe daytime fatigue and tiredness. Interestingly, as noted in primary insomnia, poor quality and reduced duration of nocturnal sleep in PD are not related to the development of EDS. [6, 103]. Surprisingly, a more robust and continuous sleep architecture has been associated with more severe sleepiness [103]. The main cause of reduced and fragmented sleep in patients with PD is the severity of parkinsonian symptoms. Although muscle tone decreases during sleep, variable degrees of rigidity can be experienced by patients with PD during the different stages of sleep. This rigidity accounts for complaints of stiffness, back pain, and leg cramps. Nocturnal akinesia is responsible, along with stiffness, for poor nocturnal mobility that manifests as difficulties turning over or getting out of bed. This situation can be extremely distressing to patients who need to go to the toilet several times at night. In patients with advanced PD, chronic treatment with bilateral subthalamic stimulation improves subjective sleep quality, something that is thought to be a consequence of increased nocturnal mobility secondary to a an improvement in rigidity and bradykinesia [5, 54]. Patients with PD who also have dementia may exhibit confusional awakenings and visual hallucinations leading to sleep fragmentation and nonrestorative sleep . The main contributors to early-onset insomnia in patients with PD are anxiety, depression, dyskinesias induced by dopaminergic drugs, and the intrinsic effect of several antiparkinsonian drugs such as selegiline and levodopa. Depression and early-morning trunk and foot dystonia cause early awakening. Circadian sleep–wake cycle disruption is another cause of disturbed nocturnal sleep in patients with PD. These individuals have an exaggerated tendency toward an advancement of phase, thereby developing an irregular sleep–wake pattern characterized by early-morning awakening and evening sleepiness. This situation is frequently associated with an advanced disease state, depression, and dementia .

Most of the studies dealing with EDS in PD have been published after 1999, following the paper by Frucht et al. [33]. It is illustrative, as an example, to review the 1998 [90] and 2001 [92] “Algorithms for the management of PD” in the journal Neurology to see how EDS was contemplated before and after the Frucht et al. [33] paper was published. In the twentieth century, EDS in PD was considered rare, and usually related to medications. This perception completely changed after the description of “sleep attacks” in PD patients treated with the dopaminergic agonists pramipexole and ropinirole.

The prevalence of persistent EDS in patients with PD ranges from 15.5 to 74 % [29, 39, 41, 47, 116, 117, 120, 126]. The main factors contributing to persistent EDS are the intrinsic pathology of PD and the sedative effects of the dopaminergic drugs. In patients with PD, the development of persistent EDS may be related to progressive cell loss in the dopaminergic and nondopaminergic brain structures, and circuits that modulate the sleep–wake mechanisms. In general, persistent EDS is associated with advanced parkinsonism and the use of dopaminomimetics. Other possible causes of persistent EDS should be considered before determining whether EDS is caused by the disease itself or by the effects of dopaminomimetics. Circadian dysrhythmia, obstructive sleep apnea, depression, dementia, and the concomitant use of other sedative drugs such as hypnotics are thought to contribute to persistent EDS .

There is a subgroup of sleepy PD patients with short mean sleep latency and the presence of REM sleep periods on the multiple sleep latency test [6, 103]. In PD, cerebrospinal fluid (CSF) hypocretin levels, however, have been reported to be normal [23] and cataplexy does not occur. In unselected PD patients, autopsies show a loss of 23–62 % of hypocretin cells, but CSF hypocretin levels are normal in PD with and without dementia [32, 122] (Table 32.1) .

Sudden onset of sleep episodes (SOS), or sleep attacks, are less common than persistent EDS in patients with PD [50]. Among patients with PD treated with dopaminergic drugs, the prevalence of SOS is estimated to range from 0 to 32 % [29, 47, 117, 120]. Episodes of SOS are considered to be the result of a dopaminergic class effect; that is, they have been shown to be associated with the use of virtually all dopaminergic drugs, occur several days or months after the introduction of the dopaminergic drug, and usually resolve or decrease after its withdrawal, reduction, or replacement. The most common variables associated with SOS episodes are therapy with dopamine agonists, duration of parkinsonism , elevated Epworth Sleepiness Scale scores, age, and sex [83, 91, 97] . In PD, modafinil [49] and sodium oxybate [94] may improve EDS.

Several studies have shown that sleep-disordered breathing is common in patients with PD, especially in those individuals complaining of sleepiness . However, it is not significantly more prevalent in patients with PD than in age-matched populations [10, 26, 27, 89, 123] (Table 32.2). In PD, the frequency of an apnea–hypopnea index greater than 5 is 27–54 %, and the frequency of greater than 30 is 4–15 %. The number of apneas does not change across consecutive nights and is not modified by the introduction of a dopaminergic agent. The number of apneas per hour in PD is not associated with age, gender, body mass index, the occurrence of RBD , oxyhemoglobin saturation at night, scales that assess sleepiness, and parkinsonism severity. Therefore, it seems that PD itself does not confer an increased risk of obstructive apneas and that the frequent presence of this condition in PD is a reflection of aging or reduction of the upper airway space. Nevertheless, patients with PD who experience EDS should undergo routine PSG to identify the potential for obstructive sleep apneas. In these cases, correct treatment of this condition with continuous positive airway pressure (CPAP) can help them greatly .

RBD is a parasomnia first described in five patients in 1986 by Schenck et al [106], who expanded the following year [107] the observation to ten new patients, involving predominantly older men (half of whom had a major neurologic disorder including parkinsonism) . Since that time, RBD has taken an increasingly important role in understanding sleep in PD. Before the description of RBD, episodes of nocturnal agitation were considered a variant of somnambulism, and thought to appear out of deep nonrapid eye movement (NREM) sleep and probably induced by levodopa (Scharf et al 1978) .

RBD is characterized by dream-enacting behaviors, fearful dreams, and REM sleep without atonia . RBD may be idiopathic or associated with neurodegenerative diseases [60]. Studies from three different groups have shown that patients with idiopathic RBD develop the classical motor and cognitive symptoms of PD, dementia with Lewy bodies, and MSA with time [58, 61, 101, 108, 109].

RBD occurs at least in one-third of patients with sporadic PD [34, 105] and is also frequent in patients with PD secondary to mutations in the parkin gene [71]. RBD occurs in PD patients untreated or treated with dopaminergic agents. RBD is more common in the rigid-akinetic clinical subtype of the disease than in the tremoric subtype [72]. For reasons not yet known, RBD is more common in male patients with PD than in female patients. Of note, up to 29 % of patients with PD with confirmed RBD by video-PSG are unaware of their abnormal motor and vocal sleep behaviors, and up to 13 % do not recall dreaming [57]. In these cases, history suggestive of RBD can be obtained only from the bed partner. On the other hand, hallucinations, somnambulism , confusional awakenings, and severe obstructive sleep apneas may mimic RBD symptoms in PD [52] . The sensitivity of specialized interviews for identifying RBD in patients with PD varies from 33 to 95 %. There is some evidence that RBD in PD is associated with dementia [128], age, and disease duration [115]. However, the occurrence of RBD in nondemented PD subjects with early stages of the disease is not uncommon. When comparing patients who have PD and RBD with patients who have idiopathic RBD and MSA with RBD, there are few differences in RBD-related clinical and sleep measures [57]. The pathophysiology of RBD in patients with PD is thought to result from the degeneration or dysfunction of the brainstem structures that modulate REM sleep [60]. Such areas include the medial medulla, pedunculopontine, and subcoeruleus mesopontine regions, and their anatomic connections with the substantia nigra pars reticulata, basal ganglia, hypothalamus, and limbic system. Postmortem studies examining the brains of patients with PD have shown that the degenerative process begins in the lower brainstem and advances upward through the pons before reaching the midbrain where the substantia nigra is located [17] . This might account for the observation that RBD precedes the onset of parkinsonism . However, an alternate sequence of neuropathologic events may take place to account for the more common finding of parkinsonism preceding RBD. RBD in PD, and associated with any other condition including the idiopathic form, responds to low doses of clonazepam at bedtime. Melatonin can be a therapeutic alternative in those few cases that do not respond to clonazepam or do not tolerate this drug. In refractory cases, methods of self-protection from injury during episodes of RBD may be necessary. Pramipexole does not improve symptoms of RBD in patients with PD [73], a finding suggesting that dopamine dysfunction does not play a central role in the pathogenesis of RBD.

RLS is a sensorimotor disorder characterized by an urge to move the legs. This impulse is caused by unpleasant sensations that begin during periods of inactivity at night and is relieved by movement. RLS is a condition that affects up to 5–15 % of the general population over 60 years of age and may interfere with sleep initiation and maintenance. Iron and dopaminergic dysfunction in the diencephalo-spinal system are believed to play a critical role in the pathogenesis of RLS because patients with RLS dramatically respond to dopaminergic drugs. Although the description of RLS is several centuries old, the description that levodopa improved their symptoms [3] pointed out the role of the dopaminergic system in RLS and the possible relationship between RLS and PD.

There is no evidence, however, that idiopathic RLS predisposes to develop PD [59] (Table 32.3). On the one hand, patients with idiopathic RLS do not progress to PD [132]. On the other hand, while it had been suggested that there is an association between PD and RLS, the evidence is still limited to few studies with methodological problems that have reported conflicting results [18, 44, 66, 69, 75, 76, 79, 88, 93, 98, 118, 129]. When PD patients are compared with a control group, there is no difference between the rate of RLS between the two groups [2, 19, 42]. It should be noted that RLS must be carefully distinguished by clinical history from other uncomfortable sensations commonly experienced by the PD subjects (stiffness, pain, tingling, heat, cramps) which may be related to parkinsonian features and not to RLS (rigidity, tremor, central pain, off periods, dystonia, and dyskinesias). Like in idiopathic RLS, PD patients with true RLS have low serum ferritin levels. In most of the PD cases with comorbid RLS, parkinsonism precedes the onset of RLS and severity of RLS is mild. This is probably because dopaminergic agents improve both RLS and parkinsonism. However, PD and RLS may be coexistent because they are both prevalent in the elderly. If true RLS is bothersome or prevents the PD patient from sleeping, the evening dose of dopaminergic agents can be increased or a standard formulation can be prescribed. Alternatively, other nondopaminergic drugs that are effective for RLS such as gabapentin and pregabalin can be used.