|Home | About | Journals | Submit | Contact Us | Français|
The etiology of sleep problems in PD is not well understood, as they may arise from the pathology of the disease or from other disease related-factors such as motor dysfunction, dopaminergic medication, and mood disturbances. The aim of this study was to investigate factors associated with sleep including disease-related variables such as motor symptom severity, dose of medication and mood and disease subtypes. Thirty-five non-demented patients with PD were included. Sleep was measured using 24-hour wrist actigraphy over a seven-day period, during which time participants kept a sleep diary. Subjective sleep and arousal questionnaires included the Parkinson’s Disease Sleep Scale and Epworth Sleepiness Scale. Motor symptom severity and dopaminergic medication were significantly related to measures of sleep quality. Gender differences in sleep were found, with men having worse sleep quality and more excessive daytime sleepiness than women. We also found that actigraphy may serve as a useful tool for identifying individuals with possible REM behavior disorder, a sleep disorder that has important implications in early detection of PD.
Sleep problems in Parkinson’s disease (PD) are among the most prominent non-motor symptoms of the disease, impacting the quality of life of patients and caregivers1. Patients with PD experience a variety of sleep-related problems, including REM Behavior Disorder (RBD), sleep fragmentation and reduced sleep efficiency, and excessive daytime sleepiness1.
Some aspects of sleep disruption arise from the underlying pathology of the disease2. The pathogenesis of PD, which begins in the brainstem and results in neuronal loss in the cholinergic, dopaminergic and serotonergic systems, is likely to affect the reticular formation that is implicated in sleep and wake regulation2. Sleep disturbances may also be secondary to motor discomfort3 or to non-motor features of the disease4. Dopaminergic medication may increase sleep fragmentation and reduce sleep efficiency5, or may cause daytime sleepiness3,6. Anxiety and depression, common in PD, can also cause problems with sleep onset and maintenance and daytime alertness7,8.
PD is heterogeneous in its presentation. Studies have focused on disease subtypes including PD into left (LPD) and right (RPD) side of motor symptom onset, type of initial motor symptom, and gender. Differences in cognitive performance, disease progression and non-motor features have been reported in subtypes of the disease. In regard to side of onset, LPD and RPD show distinct profiles of cognitive functioning9,10, sleep disturbances11 and visual hallucinations11,12. Findings of sleep differences in LPD and RPD, however, have been limited to use of sleep questionnaires. In regard to type of initial motor symptom, individuals who present with tremor (TD) vs. non-tremor dominant (NTD) symptoms express distinct clinical and cognitive profiles. Patients who present with primarily gait and balance disturbance tend to have more rapid progression of disease including more severe cognitive deficits and a faster onset of dementia than TD patients13. No studies have examined sleep quality with respect to TD and NTD.
Gender differences have been found in both motor and non-motor aspects of PD14. In the general population, women report more frequent insomnia and restless legs syndrome, whereas REM behavior sleep disorder, characterized by loss of normal muscle atonia and acting-out of dreams, is more common in men15. The examination of sleep quality with respect to gender has not been investigated in patients with PD.
The aim of the present study was to examine the association of sleep and varying clinical presentations and symptoms in patients with PD using an objective measure (actigraphy) and subjective assessments (sleep questionnaires). The present study focuses on the correlates of sleep disturbance in those with PD rather than the question of whether such disturbance exists, as this has already been demonstrated in numerous previous studies1.
Thirty-five patients with PD (22 men, 13 women) were recruited from the outpatient Movement Disorders Clinic at the Boston Medical Center (Table 1). The study was approved by the Boston University Institutional Review Board and all participants provided informed consent. On the modified Mini-Mental State Examination (mMMSE)16, a cut-off score of 25 was used, as this form of the MMSE is particularly sensitive to specific cognitive deficits found in PD without dementia (scores converted from the 57-point scale). Individuals with a history of substance abuse, head injury or neurologic disorders besides PD were excluded. None of the patients met criteria for Dementia with Lewy Bodies as per McKeith and colleagues17.
Medication information was obtained for all participants. Levodopa equivalent dosages (LED) were calculated based on previous reports with LED: (regular levodopa dose × 1) + (levodopa controlled-release dose × 0.75) + (pramipexole dose × 67.0) + (ropinirole dose × 16.67) + (rotigotine × 16.67) + (pergolide dose and cabergoline dose × 67.0) + (bromocriptine dose × 10) + ((regular levodopa dose + levodopa controlled-release dose × 0.75) × 0.25) if taking tolcapone or entacapone)18.
Motor symptom severity was quantified using the Unified Parkinson’s Disease Rating Scale (UPDRS) and Hoehn and Yahr stage. Information on the side and type of motor symptom at onset was obtained through patient report confirmed when possible by neurologist review. Mood was assessed using the Beck Depression Inventory (BDI-2) and Beck Anxiety Inventory (BAI). All patients received the Parkinson’s Disease Quality of Life questionnaire (PDQ-39)19.
To measure activity during the sleep/wake cycle, the participants wore wrist actigraphs (Actiwatch AW-64; Mini Mitter, Sunriver, OR) continuously over a 1-week period on each wrist. Data from the non-dominant wrist only was used for the current study, as per convention. The actigraphs were set to a medium sensitivity with 40 counts assessed as “wake” and an epoch length of 30 seconds. The Actiwatch data were downloaded to the Actiware sleep software version 5.3 (Mini Mitter). Results were averaged across the seven days of monitoring. The measures were sleep latency (first 10-minute period with less than two epochs of activity), sleep time (sum of time [in minutes] of epochs not exceeding the sensitivity threshold), sleep efficiency (sleep time divided by the time in bed multiplied by 100), wake after sleep onset (total time awake after the first sleep onset period) and movement and fragmentation index (number of 1-minute periods of immobility relative to the total number of immobility phases).
Sleep complaints were assessed using the Parkinson’s Disease Sleep Scale (PDSS)20, which has been used to identify disturbances such as sleep maintenance, insomnia and excessive daytime sleepiness. The scale consists of 15 common symptoms. A score of 10 indicates worse symptoms and 0 indicates no symptoms. We examined the 15 items as nine factors comprising composite scores of the individual items, as described elsewhere20. The factors were overall quality of sleep, sleep onset and maintenance, insomnia, nocturnal restlessness, nocturnal hallucinations, distressing/vivid dreams, nocturia, nocturnal motor symptoms (sensory complaints, early morning dystonia, and cramps during the night), sleep refreshment and daytime dozing. We also used the Parkinson’s Disease Sleep Questionnaire (PDSQ) from our laboratory, which captures some symptoms not assessed by the PDSS including questions assessing for RBD.
Because of the relation between sleep disorders and subjective daytime sleepiness, we administered the Epworth Sleepiness Scale (ESS)21. The score range is 0–24, with 10 or more indicating excessive daytime sleepiness. The questionnaire is commonly used in PD22,23 to assess the likelihood of dozing off in a variety of daily situations.
Patients were recruited without regard to gender or motor symptom subtype. Independent samples t-tests were used to compare demographic and clinical characteristics of the gender and motor symptom subgroups. Pearson correlations reflected associations between sleep and clinical variables such as mood and disease severity (i.e., UPDRS total, disease stage and duration, and LED). Alpha of .01 was used for all correlations and analyses by subgroups. Because there were significant correlations between all five actigraphy variables, Bonferroni adjustment was used to set alpha to .01 (.05/5). Of the nine PDSS variables, only some of them were correlated (number depending on variable), and accordingly an alpha of .01 was used here as well.
Total sleep time as measured by actigraphy (ACT) was associated with the UPDRS motor symptom total scores (r=−.43, p<.01) and LED (r=−.44, p<.009), where a higher score on the UPDRS and higher dose of dopaminergic medication corresponded to fewer hours of sleep. Duration of disease, age, education, MMSE score, BDI and BAI scores were not significantly related to ACT measures of sleep quality.
Examining associations between self-report and clinical measures revealed that subjective sleep is related to mood and quality of life as well as some disease-related measures. The PDQ-39 score was significantly related to PDSS nocturnal restlessness (r=.41, p=.01), nocturnal psychosis (r=.48, p<.004), nocturnal motor symptoms (r=.58, p<.0001), sleep refreshment (r=.48, p<.004) and the PDSS total score (r=.58, p<.0001). PDSS nocturnal restlessness was also associated with the BAI total (r=.44, p<.009) and LED (r=.42, p<.01). LED was also significantly associated with the PDSS total score (r=.42, p<.01). Hoehn & Yahr stage of disease was correlated with nocturnal motor symptoms (r=.47, p<.01). The Epworth Sleepiness Scale score was correlated with BAI score (r=.499, p<.004).
ACT and self-report sleep scores appear in Table 2. There were no gender-based differences in age, disease duration, UPDRS total or Hoehn & Yahr stage. Men were taking a higher dosage or medication than women (p<.01). On the ACT measures of sleep, although not significant presumably because of high within-group variability, men had more than 10% lower sleep efficiency than women, took twice as long to fall asleep, had a total sleep time that was 70 minutes shorter, and had a sleep fragmentation index that was approximately 8 points higher.
On the PDSS, men reported significantly worse quality of sleep. On the Epworth Sleepiness Scale, significantly more men than women scored in the clinically significant range defined by the cut-off score of 10 (χ2=5.3, p<.02). No other gender group differences were seen. When these analyses were repeated controlling for LED, men still reported overall worse quality of sleep on the PDSS.
In the general population there is a higher prevalence of sleep apnea in men than women24. We therefore repeated the gender analyses excluding individuals with a diagnosis of sleep apnea (8 men and 1 woman). There were no significant ACT differences between men and women, although relative to women the men still had a 10% lower sleep efficiency, slept 50 minutes less, and their fragmentation index was 7 points higher. On the self-report measures, a greater proportion of men still reported significant daytime sleepiness (χ=7.5, p<.006), and there was a trend for men to still report worse quality of sleep than women (p<.05),
Comparisons by side and type of symptom at onset appear in Table 3. There were 19 RPD and 15 LPD participants (one patient reported bilateral onset of motor symptoms) with no significant differences on any clinical or demographic characteristics, or on any objective or subjective sleep measures or in regard to daytime sleepiness. Three RPD and three LPD patients were on sleep medications. When these patients were removed from the analyses, there was a trend for LPD patients to report a higher score on the PDSS sleep refreshment subscale (p<.05), indicating feeling less rested in the morning.
Data from the same group of 35 PD patients was subject to a separate set of analyses using initial symptom for categorization (16 TD, 19 NTD). They were matched on age, UPDRS total, disease duration, stage of disease, medication dose and BDI score. There were no group differences on any of the sleep measures. Two NTD and four TD patients were on sleep medication. When the data from these patients were removed, there were still no significant group differences between these patients. Because NTD patients had significantly more years of education than TD patients, analyses were repeated using education as a covariate, and showed a trend for NTD patients to report more difficulty with sleep onset and maintenance on the PDSS (p<.05).
Comparisons of PD patients with and without RBD, and patients who were and were not taking dopamine agonists appear in Table 4. Nine PD patients (7 men, 2 women) reported acting out their dreams at night, indicating possible RBD. There were no significant clinical or demographic differences between these patients and the rest of the group. These patients had a higher ACT sleep fragmentation index (p<.002). Those with possible RBD also reported more nocturnal psychosis on the PDSS (p<.0001), and had a higher Epworth score indicating more daytime sleepiness (p<.01) than the others.
As dopaminergic medication has been reported to increase the severity of sleep problems, we compared patients who were on dopamine agonists to those who were not. Eight patients were not taking any dopamine agonist medication. These individuals were significantly older than the rest. They had significantly longer ACT total sleep time (p<.01). There were no other ACT or self-report sleep differences between these groups.
The current study examined multiple clinical and demographic variables that are known to impact objective and subjective sleep. We found that self-reported quality of life and mood are associated with subjective but not objective measures of sleep quality. Our findings also confirm previous reports that the severity of disease and the dose of dopaminergic medication are associated with sleep problems in PD as UPDRS total and LED were significantly related to ACT total sleep time. One possible explanation for a relation between sleep and motor function is that actigraphy may be unable to distinguish between sleep and motor problems in PD. Arguing against this notion is the observation that only total sleep time, and not sleep efficiency or fragmentation, was significantly related to these variables. This suggests that actigraphy is in fact sensitive to sleep symptoms apart from general disease severity. The association between the disease-related variables and total sleep time may be a marker of worsening disease25, which is related to an increase in severity of motor symptoms, a higher dose of dopaminergic medication, and worse sleep problems.
We did not find a relation of mood to ACT measures of sleep quality. Sleep problems experienced by individuals with anxiety and depression as a primary diagnosis (without PD) may be more related to difficulties with sleep initiation and total sleep time than with sleep maintenance26. This is the converse of what is experienced by patients with PD, for whom difficulties with sleep initiation are less frequent than sleep fragmentation27. Anxiety was associated with self-reported nocturnal restlessness, which may reflect the self-perception of bad sleep in individuals with high levels of anxiety and other mood problems28. Anxiety was also significantly related to increased daytime sleepiness, a finding we have previously reported that has also been confirmed by others7,8.
Men with PD reported worse quality of sleep and more excessive daytime sleepiness than women. The dose of dopaminergic medication did not account for the differences in sleep quality between men and women, as after controlling for LED, men still reported worse overall quality of sleep and had more daytime sleepiness. We also examined whether there was an impact of sleep apnea on the gender differences in sleep and daytime somnolence, and found that taking out those with a diagnosis of sleep apnea eliminated the gender difference in self-reported quality of sleep. By contrast, excluding those with sleep apnea actually strengthened the finding of significant gender differences in excessive daytime sleepiness. In light of this information, it is important to further investigate gender differences in sleep of patients with PD as these may be relevant to differential disease management for men and women.
We had previously found significant differences by side of motor symptom onset on several subjective sleep measures including daytime sleepiness and nocturnal hallucinations11. These findings were not replicated in the current study. Several differences between the two studies may be relevant. First, the previous study included a men-only sample from a Veterans Affairs population, in comparison to a community sample with both men and women included. Second, men with RPD in that sample were significantly more depressed than those with LPD; depression was related to the sleep factors, and controlling for depression strengthened the findings. In the present study, RPD and LPD participants were not different on any mood measure. Excluding patients who were on medications did not reveal any significant LPD/RPD differences in sleep quality. The incongruity between the current findings and those of the previous study merits further investigation of the interaction between gender and side of onset as well as the impact of mood on sleep in patients with PD.
No significant differences in quality of sleep were found between TD and NTD patients. Despite the different progression and profile of disease in TD and NTD patients, it may be that both of these subgroups experience significant sleep problems. Neuropathologically, it has been found that individuals with NTD PD have a more widespread distribution of Lewy bodies in their neocortex31, which explains increased incidence of cognitive impairment and dementia but not necessarily having a greater impact on areas that would affect sleep quality. Alternatively, our results do show evidence of trend-level differences in self-reported sleep quality between TD and NTD patients after controlling for education. To clarify this and further our understanding of meaningful subtypes of PD a more thorough study of the quality and type of sleep problems in motor symptom subtypes is needed.
RBD is one of the more commonly studied sleep disorders in PD, reflecting its potential as a harbinger of the disease32. RBD is characterized by vigorous motor activity and vivid dreams that appear to be acted out33. The gold standard for diagnosis of RBD is patient/caregiver report of dream enactment and a sleep study confirming loss of muscle atonia. We found that patients who reported acting out their dreams at night had a significantly worse ACT sleep fragmentation index, indicating that this ACT variable may be a sensitive marker of the presence of RBD in conjunction with patient report of dream enactment. This was recently corroborated in another report that also found actigraphy to be sensitive to distinguishing between those with and without RBD34. This is an important finding as many patients may present with RBD but are not diagnosed due to the high cost and burden of a full-scale overnight sleep study. Actigraphy can serve as a convenient and cost-effective diagnostic and research tool for RBD in patients with PD and in those with idiopathic RBD that may be a precursor to PD.
Dopaminergic medication has been reported to be one of the key causes of impaired sleep in individuals with PD. Although the literature reveals some dissent on this topic, our findings show that dopaminergic medication plays at least some role in the sleep problems of PD. LED was significantly related to ACT and self-reported measures of sleep. There were also significant sleep quality differences between those who were taking dopamine agonists and those who were not, again confirming that dopaminergic medication may be a factor in impaired sleep in treated PD. It is still not clear that dopaminergic medication can contribute to the severity of the sleep problems of PD alone. There is sufficient evidence in the literature that the pathology of PD itself affects brainstem and midbrain areas that are important for control of sleep and wake mechanisms2. In the current study, the dose of the dopaminergic medication was related to ACT total sleep time and self-reported measures of sleep quality but not to ACT sleep fragmentation and reduced sleep efficiency, which are the most frequent nighttime complaints in PD3.
A limitation of this study was that we used actigraphy rather than polysomnography (PSG), which is the gold standard for studying sleep. Acigraphy cannot differentiate between nocturnal activity due to wakefulness and nocturnal activity due to other common sleep disturbances of PD such as REM behavior disorder, periodic limb movements, or sleep apnea. All of these nocturnal sleep problems disrupt sleep and therefore decrease sleep quality, which is the main area of interest of our study. In order to evaluate the presence of the common sleep disturbances in PD, we used a self-report measure in addition to actigraphy. Unlike PSG, actigraphy is relatively inexpensive and is not as burdensome to the patient. The ecological validity of actigraphy is enhanced in comparison to PSG because it permits multiple-night recording in the patient’s home environment. Reliability and validity studies in healthy adults have demonstrated that actigraphy is highly correlated with polysomnography for differentiating sleep from wake states35,36 and it has been frequently used to assess sleep and wake behavior in healthy individuals and those with insomnia37. Although ACT is yet to be validated with PSG in patients with PD, it has been shown to be correlated with self-report measures of sleep38 and has been used to study efficacy of medication39 and transcranial magnetic stimulation40 related to sleep in PD.
A second limitation of the current study is the small sample sizes for our subgroup analyses. With larger samples it is possible that more subgroup differences would emerge. We did, however, find significant differences with moderate effect sizes for the subgroup analyses looking at gender and dopamine agonists and possible REM behavior disorder, indicating that at least for those variables the sample sizes were sufficient to detect meaningful subgroup differences. Though significant differences were found with small subgroup sizes, these results should be confirmed by future studies with larger samples.
In summary, our findings underscore that sleep problems in PD are multifactorial in nature. There are disease-related and individual factors (i.e., gender) that contribute to the degree, severity and type of sleep problems. Motor symptom severity and medication are among the primary correlates of sleep disruption, but additional factors differentiate sleep in PD including subtypes of the disease, with NTD patients reporting worse sleep quality, and gender, with men having worse sleep quality and more excessive daytime sleepiness. We also found that actigraphy may serve as a useful tool for identifying individuals with possible RBD, a sleep disorder that may presage PD. Understanding the cause and nature of sleep disturbances in PD and identifying them early in the disease process may lead to the development of new therapies for symptom management and enhancement of quality of life.
We would like to thank Patrick McNamara for his generous support over the course of the project and Sandra Neargarder her helpful input with data analysis. Marie Saint-Hilaire, MD, FRCPC, Cathi Thomas, RN, MS, and Denyse Turpin, RN, MPH of the Department of Neurology, Boston Medical Center provided valuable support of our participant recruitment efforts. Chase Ansok, Erica Appleman, Sophie Blease, Eric Griffith, Patricia Johnson, Barrett Phillips, and Jessica Saurman assisted with testing participants and scoring data. We acknowledge with gratitude the efforts of these colleagues and especially the efforts of all of the individuals who participated in this study.
This study was supported by a Ruth L. Kirschstein National Research Service Award (1F31NS061555-01) and a Clara Mayo Foundation fellowship from the Department of Psychology, Boston University, to K.S., and by NINDS grant R01 NS050446-01A2 to A.C-G.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.