Delayed emergence of a parkinsonian disorder in 38% of 29 older men initially diagnosed with idiopathic rapid eye movement sleep behavior disorder

REM sleep behavior disorder (RBD), first described as a distinct syndrome less than 10 years ago, is characterized by the release of vigorous and injurious behavior in REM sleep that usually represents attempted enactment of vivid, action-filled, and violent dreams. ^[1,2] Polysomnographic monitoring has identified REM sleep as the state responsible for RBD. ^[1] The diagnosis of RBD requires polysomnographic monitoring, with the hallmark EMG findings being the loss of generalized muscle atonia of REM sleep or prominent phasic muscle twitching in REM sleep, or both.

RBD usually affects older men; nearly 60% of RBD patients have symptomatic RBD, with the onset being linked with the emergence of narcolepsy, neurodegenerative, cerebrovascular, or other neurologic disorders, and other conditions. ^[1-5] The pons and medulla in the brainstem are usually spared in RBD, as determined by extensive clinical examinations and neurophysiologic/neuroimaging techniques, ^[3,5,6] although an animal model of RBD is produced by pontine lesions. ^[7,8]

We report a novel finding in RBD: the delayed and selective emergence of a parkinsonian disorder (presumably Parkinson's disease [PD]) in 38% of older male patients initially diagnosed with idiopathic RBD.

Methods.

In a series of 96 RBD patients (mean age, 52 yr; range, 9-81 yr), ^[5] 33 were men more than equals who had idiopathic RBD at the time of initial diagnosis; 88% (29/33) of these men were available for follow-up. The remaining 12% (4/33) of patients were not available for follow-up, and their data are excluded. Two of the 29 patients in this report eventually developed Alzheimer's dementia or cerebrovascular disease 7 and 5 years after RBD was diagnosed, and their data are not included.

All patients completed a comprehensive sleep-wake and medical questionnaire. Patients had been married for a mean 33.3 plus minus 10.1 (SD) years, and 25 wives participated in the clinical interviews, along with psychiatric evaluations, ^[3,9] that were conducted by one author (C.H.S.). Two other authors conducted the neurologic evaluations (S.R.B. examined 10 of 11 patients with eventual PD). Beginning with our center's first report on RBD, ^[1] in which two of five patients had neurodegenerative disorders linked with the onset of RBD, all patients diagnosed with RBD at our center have been carefully questioned and examined for PD and other neurodegenerative disorders, both during the initial evaluations and during periodic follow-up examinations. There was no identified referral bias of RBD patients, as our sleep center is a multidisciplinary, regional referral center.

Standard methods were used in recording and scoring the hospital-based polysomnographic studies, ^[10] with allowance for scoring REM sleep despite intermittent or persistent loss of REM sleep atonia, including electro-oculography, EMG (submental, bilateral anterior tibialis, and extensor digitorum muscles and additional muscles in most cases, as shown in the Figure 1, A and B), EEG with eight-channel scalp EEG montage and paper speeds of 15 and 30 mm/sec, ECG, and nasal-oral airflow monitoring using thermistors and more extensive respiratory monitoring whenever indicated (i.e., respiratory effort by inductance plethysmography and hemoglobin oxygen saturation by oximetry). There was continuous audiovisual recording of each polysomnographic study. Psychotropic medication was not taken for 1 month or longer before a polysomnographic study. A multiple sleep latency test (MSLT) ^[11] was completed in patients suspected of having daytime somnolence or who complained of daytime fatigue. The MSLT is a structured daytime polysomnographic study in which patients can nap for 20 minutes every 2 hours. Polysomnographic studies and MSLTs were interpreted by experienced clinical polysomnographers (M.W.M., S.R.B.).

• Download figure
• Open in new tab
• Download powerpoint

Figure 1. Abruptly shifting patterns of abnormal phasic and tonic motor activity during REM sleep in a 74-year-old man with REM sleep behavior disorder who subsequently developed Parkinson's disease. (A) There is constant augmentation of background muscle tone (7-submental EMG) without any of the customary atonia of REM sleep; there is also phasic motor activation with REMs (1-2-electro-oculogram), twitching of flexor (Flex.) and extensor (Ext.) upper extremity muscles ^[8-11], and minimal muscle twitching in the lower extremities (12-15). The muscle twitching (8-11) suddenly ceases in tandem with REM burst cessation (1-2). Despite the phasic motor activation, the ECG maintains a constant rate of approximately 60 / min (16). Respirations show the typical irregularity of REM sleep (17), and the EEG shows the typical desynchronization and activation of REM sleep (3-6). (B) In contrast to (A), there is no appreciable upper or lower EMG twitching (8-15) in conjunction with a burst of REMs (1-2), and there is partial suppression of the augmented submental EMG tone (7) during REM activity, followed by a burst of phasic submental EMG twitching that is closely linked with REM burst cessation.

The following neurologic tests were given after the polysomnographic studies documented RBD.

1. Neuropsychological tests ^[3,12] sensitive to cortical/subcortical dysfunction were completed by 89% (24/27) of patients, including 10 of 11 patients with eventual PD: Wechsler Adult Intelligence Scale-Revised, Wechsler Memory Scale, Buschke Selective Reminding Verbal Learning Test, Porteus Maze Test, and Controlled Word Association.

2. Brain scans: ^[1,3] MRI studies were completed in 89% (24/27) of patients; CT brain scans were completed in six patients; MRI and CT brain scans were regarded as clinically unremarkable in the presence of mild cerebral atrophy, nonspecific high-intensity deep white matter signals, ^[13] or other minor abnormalities.

3. Somatosensory and brainstem auditory evoked potential studies ^[3] were completed in 81% (22/27) of patients.

All patients satisfied recognized diagnostic criteria for RBD: ^[6,14]

1. Polysomnographic abnormality during REM sleep: elevated submental EMG tone, excessive phasic submental EMG or limb EMG twitching, or all of these.

2. Abnormal REM sleep behavior during polysomnographic studies (e.g., prominent limb or truncal jerking and complex, vigorous, or violent behavior) and/or a history of injurious or disruptive (i.e., clinically symptomatic) sleep behavior.

Patients with eventual PD satisfied Calne et al.'s ^[15] diagnostic criteria for PD at one of three levels of confidence: (1) Clinically possible PD: the presence of one of the following cardinal features: (resting or postural) tremor, rigidity, or bradykinesia. (2) Clinically probable PD: the presence of two of the following cardinal features: resting tremor, rigidity, bradykinesia, or impaired postural reflexes. Alternatively, the presence of one of the following is sufficient: asymmetrical resting tremor, asymmetrical rigidity, or asymmetrical bradykinesia. (3) Clinically definite PD: the presence of 3 or more of the following cardinal features: resting tremor, rigidity, bradykinesia, or impaired postural reflexes. Alternatively, two or more of these features are sufficient, provided that one of the first three features displays asymmetry.

Each patient had been diagnosed with RBD after completion of the clinical examinations, polysomnographic studies, MSLTs, and neurologic tests. All patients, less four, had follow-up examinations with one of the authors at least every 6 to 12 months. Four patients were followed by their physicians and by neurologists.

This study is focused on the longitudinal course of RBD in patients treated with nightly clonazepam.

Results.

Table 1 and Table 2 present the comparative findings for patients with RBD who subsequently developed PD, and those with RBD who, to date, have not developed PD. Age of RBD onset was the same for each group. In 73% (8/11) of cases, RBD had been present for 10 to 29 years before the emergence of PD. All patients had phasic EMG disturbances, and 78% (21/27) of patients had tonic EMG disturbances during REM sleep, as depicted in Figure 1, A and B. (It should be noted that compromise of REM-atonia is not a necessary feature of human RBD, since patients with RBD can have preserved REM-atonia in the face of excessive phasic motor activity and behavioral release in REM sleep. ^[16]) In each group, REM sleep latency was slightly abbreviated; the REM sleep percentage was normal but did differ between these two groups to a statistically significant, though clinically unremarkable, extent. Generalized sleep motor dysfunction was common, as 78% (21/27) of patients had periodic limb movements (PLM) of non-REM sleep, often with arousals, and approximately half the patients in each group had excessive nonperiodic EMG twitching during non-REM sleep.

The most significant difference between the two groups was in the PLM hourly index, as the group with eventual PD had an index more than twice that of the group with current idiopathic RBD. According to standard criteria, ^[14] the PLM disorder was severe in the eventual PD group and was moderately severe in the current idiopathic RBD group. Stage 3/4 sleep percentage was elevated for age in all patients, a common but enigmatic finding in RBD. ^[2,3,5] Episodes of sleepwalking or sleep terrors (i.e., non-REM sleep parasomnias that typically arise from stage 3/4 sleep) ^[14] were not recorded during any polysomnographic study. Customary REM/non-REM sleep cycling was well preserved in 93% (25/27) of patients. There were no significant differences between the two groups in sleep latency, total sleep time, sleep efficiency, and stages 1 and 2 sleep percentage.

All patients displayed behavior during REM sleep that was associated with dreaming: shouting, gesturing, throwing punches, kicking, sitting up, attempting to jump out of bed, crawling, and running in bed.

Minor sleep-disordered breathing, without oxygen desaturation or substantial sleep disruption, was present in 30% (8/27) of patients (n equals 5 with idiopathic RBD); one patient had mild obstructive sleep apnea.

MSLTs, completed in 73% (8/11) of patients with eventual PD and in 63% (10/16) of the current idiopathic RBD patients, did not document any daytime somnolence: Mean sleep latency was always longer than 5 minutes; mean sleep latency for each group was 18 plus minus 1.8 minutes and 10.6 plus minus 5.8 minutes; and REM sleep was not recorded during any nap. Narcolepsy was thus ruled out.

RBD was fully or substantially controlled with nightly clonazepam treatment ^[2,3,5] in 89% (24/27) of all patients, for a mean follow-up interval of 5.9 plus minus 2.1 years.

Two patients (one with eventual PD) had previously been treated intermittently with neuroleptics (i.e., central dopamine antagonists) for psychotic depression. One patient had a remote history of alcohol abuse. There was no identified family history of either RBD or PD.

PD was eventually diagnosed with the following levels of confidence: definite PD, 73% (8/11); probable PD, 18% (2/11); possible PD, 9% (1/11). A resting tremor, typical of PD, was present in 73% (8/11) of patients. All patients with PD had 3 or more ancillary signs of PD: handwriting deterioration with micrographia, hypophonia, seborrheic dermatitis, sialorrhea, masklike facies, gait disturbance (e.g., shuffling, loss of arm swing), dementia, and more. Clinical examinations did not detect pyramidal tract or cerebellar signs, gaze palsies, dyspraxia, or autonomic disturbances. There was no known exposure to toxic substances, nor any history of head injury or encephalitis. All MRI and CT brain scans were unremarkable, and heavy iron deposition in the basal ganglia was never detected. Serial neuropsychologic tests in the eventual PD group detected moderately severe dementia in 18% (2/11) of patients; dementia has not been detected in the current idiopathic RBD group.

The diagnosis of PD was independently confirmed in 82% (9/11) of patients on the basis of several examinations by at least two neurologists (one from our center) who were aware of the RBD diagnosis. Inter-rater reliability was not assessed.

Data on PD treatment were available in 64% (7/11) of patients: 7 (88%) patients had prompt and substantial improvement of bradykinesia, rigidity, tremor, and ancillary signs and symptoms with carbidopa/L-dopa (n equals 5) or selegiline (n equals 3) treatment, or both; 1 patient could not tolerate treatment.

Discussion.

This report establishes that RBD can be the sole heralding clinical manifestation of a parkinsonian disorder (presumably PD) ^[15,17] in a substantial percentage of older male patients with RBD. PD comprised 85% (11/13) of all neurologic disorders emerging after RBD had been diagnosed. This apparently specific, newly identified link between RBD and PD may become even stronger as additional cases of PD eventually emerge in our RBD series. However, some patients may subsequently develop multiple system atrophy (MSA) ^[17] (striatonigral degeneration subtype). There are three published cases (all men) of MSA initially manifesting as RBD, involving the Shy-Drager syndrome, ^[18] striatonigral degeneration, ^[19] and olivopontocerebellar atrophy ^[19] subtypes. There is another published case of autopsy-diagnosed ``Lewy body disease,'' with marked loss of monoaminergic neurons in the substantia nigra and locus ceruleus, affecting an elderly man with longstanding ``idiopathic'' RBD. ^[20]

Because females comprise only one-fifth of all RBD cases in the world literature, ^[5] the sample size of RBD cases is currently too small to ascertain whether the strong association between RBD and PD (which affects men and women equally) is restricted to men only.

Could chronic treatment of RBD with clonazepam (which appears to exert its beneficial effect by suppressing excessive phasic motor activity in REM sleep ^[21]) have contributed to the emergency of PD? We think not. First, a computerized literature search did not identify any reports on adverse extrapyramidal effects related to clonazepam, although there are various reports on the antiparkinsonian and related beneficial effects of the drug. ^[22-27] Second, clonazepam promotes central dopaminergic activity, ^[28-30] although there may also be complex agonistic-antagonistic effects. ^[31-33] Third, a large group of adults with injurious sleepwalking and sleep terrors, treated with clonazepam on a chronic, nightly basis, have not developed extrapyramidal signs or other stigmata of PD. ^[34]

PD is associated with disturbed motor activity during REM sleep (reviewed in ^[6]), and recent work suggests that both preclinical and clinical RBD may be common in PD. ^[35-38] Thus, patients with PD should be periodically questioned, with their spouses, concerning the presence or emergence of RBD. Although the mechanisms by which the REM-atonia circuitry, which involves specific neuronal groups in the pons and medulla, ^[8,39-41] can be perturbed by PD are currently unknown, the pedunculopontine nucleus (PPN) ^[42] is likely to be prominently involved, for at least three reasons. First, there is a strong reciprocal connectivity between the PPN and the substantia nigra, ^[42] the main site of pathology accounting for the cardinal signs of PD. ^[17] Second, the neuropathology of PD includes prominent neuronal loss within the PPN. ^[43] Third, the PPN has strong connections with both the REM-atonia circuitry ^[42,44] and the REM-phasic generator circuitry. ^[44,45] Furthermore, the substantia nigra also has a strong connection to the REM-phasic generator circuitry and may have an important role in the genesis of pontogeniculooccipital waves, ^[46] a hallmark REM sleep phasic event. The retrorubral nucleus, which may be involved in the REM-atonia circuitry ^[44] and is located near the substantia nigra, may also be involved in the linked PD-RBD pathology. Postmortem brain analyses may elucidate the major mechanisms underlying the RBD-PD disease association.

Acknowledgements

The contributions of David C. Anderson, MD; Connie Ullevig, RN; Edgar Garcia-Rill, PhD; Jerome M. Siegel, PhD; Allan Callies, BA; and our Sleep Lab Technologists are gratefully acknowledged.