Assistant Professor of Neurology, Weill Cornell Medical College, Center for Sleep Medicine
Dr. Barone reports no financial relationships relevant to this field of study.
SYNOPSIS: Parkinson’s disease is associated with sleep disorders commonly. Disrupted sleep patterns appear to make dyskinesias worse in patients treated with levodopa therapy.
SOURCE: Amato N, Manconi M, Moller JC, et al. Levodopa-induced dyskinesia in Parkinson's disease: Sleep matters. Ann Neurol 2018; Oct 17. doi: 10.1002/ana.25360. [Epub ahead of print].
It is well known that Parkinson’s disease (PD) comprises a spectrum of motor and nonmotor symptoms that tend to change as the disease progresses. Levodopa, a medication ubiquitously used in PD patients, can have paradoxical effects on PD. Levodopa controls motor symptoms for several years, but it later induces motor fluctuation and abnormal involuntary movements, heretofore known as levodopa-induced dyskinesias (LIDs).
The synaptic homeostasis hypothesis (SHY) is a proposed mechanism to explain why the brain needs sleep for retention of memories. In wakefulness, learning requires a strengthening of certain synaptic connections, unsurprisingly increasing the need for neuronal cellular energy. In sleep, there is a restoration of cellular homeostasis through down-regulation of other synapses. Additionally, slow-wave activity (SWA) during non-rapid eye movement (NREM) sleep particularly contributes to adjustment of plasticity and cortical excitability.
In prior studies, sleep-deprived rodents that were exposed to levodopa developed earlier and more severe LID than those that were not sleep deprived. Amato et al explored the correlation, in humans, between objective sleep parameters and clinical features of different subpopulations of PD patients.
The authors recruited 36 subjects with PD and divided them into three groups: 1) de novo (DNV; n = 9): patients recently diagnosed and naïve to dopaminergic therapy other than rasagiline; 2) advanced (ADV, n = 13): patients not demonstrating LID with their habitual therapy but having the end-of-dose or wearing-off phenomenon; and 3) dyskinetic (DYS, n = 14): advanced patients experiencing motor fluctuations and showing LID.
These subjects, as well as 12 age-matched controls, underwent whole-night video polysomnography high-density EEG (vPSG-hdEEG), preceded by one week of actigraphy. Then, the SWA content of the vPSG-hdEEG was divided into 10 equal parts, noted from T1 to T10. Parts T2, T3, and T4 correlated with early sleep, and parts T7, T8, and T9 represented late sleep.
As expected, the authors found that early sleep control subjects showed a significantly greater amount of SWA compared to all PD patient groups (P < 0.01). On further analysis, SWA also occurred more often in the DNV group compared to the other patient groups (P < 0.01). SWA also was greater in the ADV group compared to the DYS group (P < 0.01); the DYS group had the lowest SWA overall. Additionally, the authors found a significant difference in SWA between early and late sleep in the control (n = 7), DNV (n = 5), and ADV groups (n = 9; P < 0.01) but not in the DYS group (n = 10). Furthermore, while the correlation between SWA and disease duration was positive in both the DNV and ADV groups, it was negative in the DYS group.
The authors believed the correlation between SWA and disease duration in both the DNV and ADV groups might reflect compensatory mechanisms within the SHY framework that could be ineffective in the DYS group. Similarly, they attempted to explain the lack of difference between the early sleep and late sleep SWA in the DYS group as a potential failure of the homeostatic mechanism proposed in the SHY.
Amato et al concluded that there is a clear association between sleep and LID. Considering the homeostatic hypothesis, and that researchers cannot determine a causative role for the lack of SWA reduction in the emergence of LID, they can suggest an association between sleep and some clinical phenotypes of PD, as well as a relationship between sleep disruption and LID.
This study was small but well done. As the authors noted, PSG studies in the PD population have shown conflicting results regarding changes in sleep efficiency, total sleep time, and sleep stages in PD patients compared to controls. What is unique about the Amato et al study is that sleep architecture was examined with respect to the disease stage and to the presence or absence of motor fluctuations and LID, allowing for a more accurate detection of differences that otherwise would not be defined clearly in a heterogeneous group. Future studies examining the link between sleep and PD should adopt a similar framework.
The link between PD and sleep disorders is becoming more prevalent in the medical literature, and an improved understanding of both facets of this equation will be of great help to researchers and clinicians. Of particular interest is the intriguing relationship between REM behavior disorder as an early marker of PD and PD-related conditions (i.e., multiple system atrophy, dementia with Lewy bodies). A better understanding of sleep, REM behavior disorder, and PD may prove vital to one day preventing or curing PD.