© Springer-Verlag Wien 2015
Aleksandar Videnovic and Birgit Högl (eds.)Disorders of Sleep and Circadian Rhythms in Parkinson’s Disease10.1007/978-3-7091-1631-9_1414. Cognition and the Sleep–Wake Cycle in Parkinson’s Disease
(1)
Department of Psychology, Université du Québec à Montréal, Montreal, QC, Canada
(2)
Centre d’Études Avancées en Médecine du Sommeil, Hôpital du Sacré-Cœur de Montréal, 5400 boul. Gouin Ouest, Montreal, QC, H4J 1C5, Canada
(3)
Department of Neurology, McGill University, Montreal General Hospital, Montreal, QC, Canada
Abstract
Non-motor symptoms of Parkinson’s disease (PD) have received increasing attention in the past decade, particularly cognitive and sleep dysfunctions. Moreover, a growing body of evidence suggests an association between sleep and cognition in aging. This chapter outlines the role of sleep in the maintenance of cognition and learning and the high prevalence of cognitive impairment in PD. We then summarize the evidence for and against associations between rapid-eye-movement (REM) sleep behavior disorder (RBD), excessive daytime sleepiness (EDS), sleep-disordered breathing (SDB), insomnia, sleep quality, sleep architecture abnormalities, and cognitive impairment in PD. Three major sleep outcomes, RBD, EDS, and reduced sleep spindles, are correlated to cognitive impairment in PD, whereas insomnia, sleep quality, and SDB show modest or no correlation. Further longitudinal studies are needed to determine the role of RBD, EDS, and sleep spindle abnormalities as potential markers of cognitive decline in PD.
Keywords
Parkinson’s diseaseSleepCognitionAgingMild cognitive impairmentDementia14.1 Introduction
The non-motor features of Parkinson’s disease (PD) have been well identified by clinicians and researchers in the last decade. Two of the most devastating for patients and their caregivers are sleep disorders and cognitive impairment, which affect nearly all patients with PD over the course of the disease. Indeed, patients with PD frequently develop symptoms such as insomnia, rapid-eye-movement (REM) sleep behavior disorder (RBD), sleep-disordered breathing (SDB), excessive daytime sleepiness (EDS), mild cognitive impairment (MCI), and/or dementia. Moreover, sleep and cognition are strongly related: (1) some memory and learning process appear to be mediated by cerebral plasticity mechanisms that occur in part during the sleep–wake cycle and (2) several sleep disorders are associated with varying degrees of cognitive deficits. Both sleep and cognition present age-related changes, which are amplified in pathological aging, such as neurodegenerative diseases. Recent studies suggest that some of these PD-related sleep disorders are associated with an increased risk for cognitive impairment. This chapter reviews the evidence for an association between sleep and cognition in PD.
14.2 Cognition in Parkinson’s Disease
Cognitive impairment is a frequent non-motor symptom of PD, and can occur early in the course of the disease [1, 2]. The cognitive profile of patients with PD is heterogeneous and varies widely with the disease stage [1, 2]. The most consistently reported impaired cognitive domains in PD are attention, executive functions, episodic memory, and visuospatial abilities [1, 2]. The severity of the cognitive impairment also differs widely between patients: some patients remain cognitively intact for a long period of time, whereas others develop MCI or dementia early in the disease course.
MCI is defined as significant cognitive decline compared to age- and education-equivalent individuals, without major impact on activities of daily living. It is now a well-recognized feature of PD [1]. Cross-sectional studies using various definitions of PD-MCI have reported MCI in 19–38 % of patients with PD [1]. New criteria for PD-MCI have been proposed by a Movement Disorder Society Task Force [1]. However, the validity of these criteria remains to be determined. MCI is a risk factor for a more severe cognitive decline in PD [1, 3], which is not surprising, given that MCI is often an intermediate stage between normal aging and dementia in the general population [4].
Dementia is one of the most devastating non-motor features of PD. The point prevalence of dementia in PD has been estimated at from 8 to 48 % in cross-sectional studies [2]. This wide range is mainly due to between-study variability in methodology (i.e., population selection, dementia criteria, and population heterogeneity). A point prevalence of 30 % is suggested [2, 5]. The Movement Disorder Society Task Force has also proposed criteria for dementia in PD [2]. Whereas the cross-sectional prevalence is moderately high, prospective long-term studies have reported that most patients with PD eventually develop dementia [2, 5].
There is considerable variation between the time of dementia onset and the time of PD onset [2]. Some demographic and clinical risk factors for dementia in PD have been identified, namely older age, depression, more severe parkinsonism with rigidity, akinetic-rigid subtype, freezing, postural instability, and gait disturbance [2, 5]. In addition, hallucinations, MCI, and apathy increase the risk for dementia in PD, mainly because they are early symptoms of cognitive impairment [1, 2, 5]. Various candidate biomarkers for cognitive decline in PD are currently being investigated, including waking EEG and structural and functional neuroimaging abnormalities, biomarkers in cerebrospinal fluid, and genetic polymorphisms [5]. This chapter focuses on sleep abnormalities as potential risk factors for and markers of dementia in PD.
14.3 Sleep and Cognition in Aging
Sleep is a complex physiological state characterized by coordinated cyclic changes in the activity of diverse neuronal systems. Thus, major changes in neuronal molecular, electrophysiological, and neurochemical activity occur throughout the sleep–wake cycle [6, 7]. Although the specific roles of these physiological events have not been determined, growing evidence suggests that sleep has a significant impact on cognition, particularly memory consolidation and cerebral plasticity [6–8].
In normal aging, age-related changes in sleep are common, and are characterized by increased wake after sleep onset (WASO), reduced duration of slow-wave sleep (SWS), and increased time spent in stage 1 sleep [9, 10]. Consequently, there are more awakenings and sleep quality is poor. Although stage 2 sleep remains relatively unchanged, the EEG features of stage 2 sleep are less pronounced, with reduced frequency of sleep spindles and lower amplitude of K complexes. The REM sleep percentage remains constant, with only a small decrease in older age. Circadian phase advances, characterized by a shift in the sleep–wake cycle toward morningness, have also been reported. Recent studies suggest that age-related changes in sleep parameters are associated with age-related cognitive decline [11, 12].
In addition to sleep-related architecture changes, aging is associated with increased prevalence of sleep disorders such as SDB, EDS, insomnia, and RBD. These disorders can also impact cognition. In older adults with SDB, impaired vigilance, attention, executive functions, and learning have been reported [13, 14]. In a prospective study in older women, SDB was identified as a risk factor for developing MCI or dementia [15]. However, the mechanism of this effect is unclear (i.e., hypoxia, disrupted sleep, comorbidities such as cerebrovascular disease). Idiopathic RBD is also associated with cognitive impairment, characterized by reduced performance in attention, executive functions, learning, and visuospatial tasks and increased risk for MCI. This is almost certainly due to the fact that RBD is a well-recognized risk factor for dementia with Lewy bodies and PD [16, 17]. Moreover, alterations in the sleep–wake cycle (insomnia, EDS, WASO, and poor sleep quality and duration) may be associated with altered cognitive performance and cognitive decline in the elderly population, although this relationship remains controversial [18–21].
All of these sleep abnormalities are associated with PD, which is characterized by widespread degeneration of the neurons of the reticular activating system and the diverse brain stem structures involved in sleep regulation [9, 10]. This suggests that in pathological aging such as PD: (1) sleep alterations may impact the cerebral plasticity-related mechanisms underlying certain cognitive functions; and/or (2) sleep alterations may indicate more widespread and severe neurodegeneration, particularly in the brain stem, the thalamus and the cortex (thalamo-cortical loop), which indirectly impacts neurobiological systems related to cognitive functions.
14.4 Sleep and Cognition in Parkinson’s Disease
As reported in other chapters of this book, sleep complaints and sleep disorders are major non-motor symptoms in PD [9, 10]. The most frequently reported sleep disorders in PD are RBD, EDS, SDB, and insomnia. What are the relationships between these disorders and cognition in PD?
14.4.1 Rapid-Eye-Movement Sleep Behavior Disorder
In studies using polysomnography (PSG) for diagnosis, RBD affects approximately 30–45 % of patients with PD [22, 23]. RBD has been related to disturbances of the brain stem neural networks underlying REM sleep muscle atonia and motor control [24]. These brain stem regions are known to be disrupted in PD [24]. Given that RBD affects only a subgroup of patients with PD, those with concomitant RBD could have a more severe form of PD characterized by more diffuse neurodegeneration, leading to functional impairments such as cognitive deficits.
From 2006 to 2009, some studies reported decreased cognitive performance in patients with PD with RBD. Sinforiani et al. showed altered executive functions performance in patients with PD with clinical RBD compared to patients with PD without RBD [25]. Compared to normative values, patients with PD without RBD did not present any cognitive deficits. A 2-year follow-up of these cohorts confirmed executive dysfunction in patients with PD with RBD, particularly in older individuals at risk for more rapid progression of motor symptoms and hallucinations [26]. Moreover, patients with PD without RBD remained cognitively intact. In 2007, our group performed complete neuropsychological assessments to compare two groups of patients with PD, with and without RBD, to a group of healthy subjects [27]. We found decreased performance on cognitive tests measuring attention, executive functions, verbal learning and memory, and visuospatial abilities in patients with PD with RBD. patients with PD without RBD showed equivalent performance to healthy subjects on all cognitive measures. These results were confirmed in a larger sample drawn from the same cohort in 2009 [28], showing additionally that the presence of MCI in PD is strongly related to RBD. Thus, MCI was present in 73 % of patients with PD with RBD compared to 11 % of patients with PD without RBD and 8 % of controls [28]. Some subsequent studies have confirmed the association between RBD and cognitive dysfunction in PD [29–32] whereas others have not [33–38]. Two recent studies assessed cognition in treatment-naïve newly diagnosed patients with PD [39, 40]. The first reported equivalent cognitive performance and MCI frequency in patients with PD with and without RBD [39]. The second showed no difference in cognitive profile between patients with PD with and without REM sleep behavior events [40]. These two studies suggest that PD duration and the use of dopaminergic medication may modulate cognitive impairment in patients with PD with RBD. Indeed, higher age and more advanced duration of PD are two well recognized risk factors for cognitive decline in PD [41]. Moreover, although still unclear, dopaminergic medication seems to modulate cognition in patients with PD but the positive or negative impact varies according to the type of treatment and concentration used [5]. Note that most previous studies on the impact of RBD in PD have been cross-sectional with certain methodological limitations. Many used screening cognitive tests only, which have poor sensitivity to detect cognitive impairment in PD [31, 34, 36, 38]. Moreover, RBD diagnostic criteria varied, with some studies using clinical criteria without PSG confirmation or nonstandard PSG criteria, which may affect the accuracy of RBD classification, falsely reducing differences between groups [25, 26, 30–32, 34–38, 40]. Others have not included a healthy control group, which limits the interpretation of the results [25, 26, 30–39]. Furthermore, studies of the highest quality, which used PSG criteria for RBD and standard cognitive batteries, were performed on relatively small samples of patients with PD, reducing the statistical power [27–29, 33, 39]. Hence, further longitudinal studies in larger samples are needed, including PSG to confirm RBD, a healthy control group, and complete neuropsychological assessments, in order to better understand the association between RBD and cognition in PD.
The association between RBD and the development of dementia in PD has also been investigated. Marion et al. reported higher prevalence of clinical RBD in PD with dementia and faster cognitive decline in PD with RBD [42]. In a prospective study over a mean 4-year follow-up, we found that 48 % of patients with PD with RBD at baseline developed dementia, whereas no patients with PD without RBD converted to dementia [43]. This finding was confirmed by Nomura et al., who observed faster occurrence of dementia in PD with RBD compared to PD with normal REM sleep features [44]. Taken together, these studies suggest that RBD in patients with PD may be associated with more rapid cognitive decline, and that RBD could be a clinical risk factor for dementia in PD.
Other studies have found functional and structural impairments specific to RBD in PD, which can explain to some degree the higher risk of cognitive deficits in patients with PD with RBD. A distinct clinical profile, often associated with the presence of cognitive impairment in PD, has been identified in patients with PD with RBD, with autonomic dysfunction, higher incidence of visual hallucinations, more freezing and falls, a non-tremor dominant subtype, and symmetric disease [45–50]. Other studies have demonstrated abnormalities in quantitative waking EEG and event-related potentials, mainly in posterior cortical areas, in patients with PD with RBD [51, 52]. Another study using positron emission tomography ([11C]methylpiperidyl propionate acetylcholinesterase) reported relative neocortical, limbic cortical, and thalamic cholinergic denervation in patients with PD with clinical RBD [53]. A further study using diffusion tensor imaging and voxel-based morphometry (VBM) identified subtle (in uncorrected analyses) reductions in cortical gray matter volume (parietal and temporal lobes) and widespread white matter microstructural abnormalities in patients with PD with clinical RBD [54]. Taken together, these studies provide functional and anatomical support, which could explain the higher prevalence of cognitive impairment in PD associated with RBD.
14.4.2 Sleep-Disordered Breathing and Excessive Daytime Sleepiness
Unlike the relatively clear relationship with RBD, the role of SDB in cognition in PD is much less clear. SDB is commonly present in PD, with prevalence varying from 22 to 66 % depending on the apnea-hypopnea index (AHI) cut-off used [55]. However, studies with control subjects have not found increased SDB prevalence in PD compared to healthy age-matched individuals [55]. Despite the high incidence of SDB in PD and its deleterious effects on health reported in the general population, the impact of SDB on motor and non-motor symptoms in PD appears to be minimal. Only a few studies have investigated the impact of SDB on cognition in PD. Cochen de Cock et al. found no significant effect of sleep apnea on nocturia, sleepiness, depression, cardiovascular events, or cognitive impairment in PD [56]. However, they used the standardized mini-mental state examination to measure cognitive functioning, a screening test that is insensitive to cognitive impairment in PD [57]. Our group recently studied 92 patients with PD who underwent PSG, an extensive neurological exam, including several non-motor measures, and a complete neuropsychological assessment [58]. The prevalence of SDB in our cohort, depending on the AHI cut-off, was 11 % (with AHI >15), 21 % (AHI >10), and 33 % (AHI >5). We found no significant differences in motor and non-motor symptoms between apneic and non-apneic patients with PD, regardless of the AHI cut-off. Of note, the two PD groups did not differ on any cognitive measures (attention, executive functions, learning and memory, or visuospatial abilities) or in the proportion of patients with MCI. These results confirm the absence of a strong relationship between SDB and most major outcomes in PD, including cognitive decline.
EDS is another frequent non-motor symptom in PD. Although EDS has been related to cognitive dysfunction in PD [59, 60], several factors may confound this relationship, and other studies have not confirmed it [61–63]. In fact, only a few studies have formally examined the relationship between EDS and cognition in PD. Gjerstad et al. reported that patients with PD with EDS more often had dementia at baseline and showed faster progression of cognitive impairment and disability after a 4-year follow-up [64]. In a subsequent 4-year follow-up on their cohort, the association between cognitive impairment and EDS was confirmed by univariate analysis, but not by multivariate analysis [65]. In another study, Compta et al. reported no significant difference on the Epworth sleepiness scale (ESS) scores between patients with PD with and without dementia [66]. However, the EDS prevalence (ESS > 10) was higher in patients with PD with dementia. Goldman et al. compared EDS symptoms between cognitively intact patients with PD, patients with PD with MCI, and patients with PD with dementia [37]. They found higher ESS scores and higher EDS prevalence in patients with PD with dementia compared to the two other groups. ESS scores correlated with several cognitive measures. However, patients with PD with and without MCI were equivalent on the ESS. In non-demented patients with PD, a significant correlation between EDS and slowed processing speed has been reported, with no differences in other cognitive measures [30]. We recently compared daytime sleepiness symptom severity in 16 patients with PD with MCI, 20 cognitively intact patients with PD, and 36 healthy subjects [67]. All groups were equivalent on sociodemographic variables, and the two PD groups did not differ on clinical signs of PD, including dopaminergic medication dosage. We found no difference on the ESS scores between patients with PD with MCI (mean = 9.60) and patients with PD with normal cognition (mean = 9.55). However, the two PD groups scored higher on the ESS than healthy subjects (mean = 6.42). Of note, EDS in PD may be heterogeneous: in early stages, medication side effects may play a prominent role, whereas EDS becomes less dependent on medication doses as the disease progresses (i.e., it is mostly a primary disease manifestation). Using structural neuroimaging and VBM, Kato et al. found that patients with PD with EDS had marked gray matter atrophy in several brain regions compared to patients with PD without EDS and healthy subjects, whereas patients with PD without EDS showed no gray matter atrophy compared to controls [68]. Taken together, these results suggest that in PD: (1) EDS is a common feature regardless of the presence of cognitive impairment; (2) the severity of EDS increases with cognitive decline and the symptoms are more manifest in patients with dementia; and (3) EDS is associated with more severe neurodegeneration. This suggests that EDS may be a risk factor for dementia in PD, although this hypothesis needs further validation.
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
