Quantitative REM Sleep without Atonia in Parkinson's Disease and Essential Tremor

ABSTRACT

Background

Rapid eye movement (REM) sleep behavior disorder (RBD) occurs occasionally in essential tremor (ET), but polysomnographic REM sleep without atonia (RSWA) analyses have been sparse.

Objective

To characterize the amount and distribution of polysomnographic RSWA, the electrophysiologic substrate of RBD, in patients with Parkinson's disease (PD) and ET.

Methods

We analyzed quantitative RSWA in 73 patients: PD (23), ET (23), and age‐sex‐matched controls (27). None had dream‐enactment behavior history or received antidepressants. Phasic, tonic, “any,” and phasic‐burst duration RSWA measures were calculated in the submentalis (SM) and anterior tibialis (AT) muscles. The automated REM atonia index (RAI) was also determined. Statistical analysis was performed by Kruskal‐Wallis rank‐sum and Mann‐Whitney tests.

Results

SM phasic RSWA was significantly greater for PD than ET patients and controls (12.5% ± 12.8% vs. 4.9% ± 6.7%, 3.9% ± 2.6%), as was SM “any” (13.54% ± 14.30% vs. 5.2% ± 7.6%, 4.2% ± 2.6%). RAI was significantly lower in PD than in ET and controls (0.78 ± 0.23 vs. 0.92 ± 0.09 vs. 0.90 ± 0.17, P ≤ 0.005), but no different between ET and controls. AT phasic and “any” RSWA was similar between the 3 groups. ET and control RSWA was similar in all measures. Two ET patients (8.7%) had SM RSWA similar to PD patients.

Conclusions

Elevated SM RSWA distinguished PD from ET in patients without dream‐enactment symptoms and occurs frequently in PD patients, and in isolated tremor suggests underlying synucleinopathy. Prospective studies will further validate these findings.

Isolated tremor syndromes such as early Parkinson's disease (PD) and essential tremor (ET) have overlapping features and can be difficult to clinically differentiate in the absence of other features such as bradykinesia, rigidity, or postural instability. Few objective tests are available for differentiating PD and ET. ¹ Additionally, there is growing evidence for a relationship between PD and ET with 1 study finding 20% of ET patients developing PD over a 5‐year time course ² and another determining the relative risk for PD in those with ET is 4.27. ³ Pathological evidence for Lewy body presence was found in 24% of ET patients, ⁴ and similar alterations in Purkinje cells were found in both ET and PD patients. ⁵ , ⁶ , ⁷ However, the significance of these findings remain uncertain, and future research is needed to examine the similarities and differences between these 2 disease pathologies.

The presence of rapid eye movement (REM) sleep‐behavior disorder (RBD) is strongly associated with alpha synucleinopathies, including PD, but its significance in ET is less clear. ⁸ , ⁹ , ¹⁰ , ¹¹ Idiopathic/isolated RBD can present decades before the onset of clinical PD symptoms of tremor, bradykinesia, and postural instability, and 70% to 90.9% of older RBD patients will ultimately “phenoconvert” to PD, dementia with Lewy bodies, or multiple system atrophy within 10 to 15 years following RBD diagnosis. ¹⁰ , ¹² , ¹³ , ¹⁴ A diagnosis of RBD requires loss of skeletal muscle paralysis, otherwise known as REM sleep without atonia (RSWA), in conjunction with a clinical history of dream enactment behaviors. Although RBD itself is a well‐recognized manifestation of prodromal synucleinopathy, RSWA without accompanying dream enactment likely occurs before RBD diagnosis and may be an even earlier biomarker for synucleinopathy neurodegenerative disease. ¹⁵ , ¹⁶

Questionnaire‐based studies of RBD prevalence in ET have shown a wide frequency, varying from 13% to 43.5% of ET patients who may have probable RBD, which is likely related to relatively poor specificity of RBD questionnaires in symptomatic patient groups. ¹⁷ , ¹⁸ , ¹⁹ , ²⁰ A prior study evaluated polysomnographic (PSG) findings in ET compared with PD; no ET patients met the International Classification of Sleep Disorders, 2nd edition criteria for a diagnosis of RBD, but 3 of 16 (19%) patients had qualitative RSWA. ²¹ Another study showed that 14 of 49 (29%) ET patients screened positive for RBD, but RBD was confirmed in only 5 (35%) of those patients. Interestingly, all 5 patients with ET + RBD in this study had rest tremor in addition to a postural and kinetic tremor, suggesting a probable underlying concomitant parkinsonian syndrome. ²²

We aimed to determine whether (1) quantitative RSWA amounts are greater in PD than ET, and (2) if RSWA is also present in ET, possibly representing patients who could have concomitant PD or risk for evolving subsequent PD.

Methods

Patient Selection

Seventy‐three patients who completed a sleep study at the Mayo Clinic Center for Sleep Medicine from 2008 to 2018 were identified for retrospective analysis including 23 PD patients, 23 ET patients, and 27 controls without neurological disorders. Patients with a diagnosis of RBD, history of dream‐enactment behavior, or taking antidepressant medications at the time of PSG were excluded from analysis. All patients underwent a clinical interview regarding the presence or absence of dream‐enactment behaviors with a sleep physician during sleep consultation before PSG, and all video PSG studies were reviewed for the presence or absence of possible dream‐enactment behaviors by a board‐certified sleep medicine physician. Patients with ET met Movement Disorders Society criteria ²³ for ET at the time of PSG and did not have an accompanying diagnosis of PD. Patients with PD were diagnosed according to United Kingdom PD Society Brain Bank criteria and did not have a previous diagnosis of ET. Controls were matched to ET and PD patients for age and sex. Relevant clinical and demographic features were recorded for all patients.

Polysomnographic Recordings

Our methodology for video PSG recording and scoring has been described previously. ²⁴ , ²⁵ Submentalis (SM) and bipolar‐linked anterior tibialis (AT) electromyography (EMG) were recorded for all subjects.

Analysis of REM Sleep Muscle Activity

RSWA analysis methods were as previously described. ²⁴ , ²⁵ , ²⁶ , ²⁷ Visual/manual quantitative analysis of EMG activity was performed using HypnoLab sleep‐scoring software (ATES Medica Labs, Verona, Italy). Scorers were blinded to each group and individual patient data. All scorers had high inter‐rater reliability against a gold‐standard record, as previously described, ²⁴ , ²⁸ with a κ score of 0.89 for this study. Muscle activity was scored as phasic, tonic, or “any” (either form of muscle activity occurring within a 3‐second mini‐epoch) by previously published methods in the SM and AT individually, as well as for combined SM + AT muscles. ²⁵ Arm EMG was not recorded in this cohort of patients, given lack of clinical suspicion for parasomnia behaviors. Phasic muscle activity had 0.1 to 14.9‐second duration with amplitude of >4× the lowest background muscle activity. A 3‐second mini‐epoch was considered positive for phasic muscle activity when muscle activity met the criteria for phasic activity within that mini‐epoch. Tonic muscle activity was scored when muscle activity duration was >15 seconds with amplitude of >2× the lowest background muscle activity during REM sleep (<0.5 μV for all subjects). If tonic muscle activity occurred in >50% of a 30‐second epoch, the entire epoch was considered to be positive for tonic activity. “Any” muscle activity was scored if either phasic or tonic muscle activity was present in a particular 3‐second mini‐epoch. The duration of phasic‐muscle bursts was analyzed and compared between groups. For all breathing‐related arousals in REM sleep, the 3‐second epoch in which they occurred was excluded from analysis. The phasic and “any” percent muscle activity for each muscle was calculated as the number of positive mini‐epochs divided by the total number of REM sleep mini‐epochs throughout the study (excluding mini‐epochs with artifacts). The automated REM atonia index (RAI) was also analyzed in the SM muscle, yielding a value between 0 and 1 (1 = normal REM muscle atonia), derived measuring the average SM EMG amplitude for every 1‐second epoch during REM sleep. ²⁹

Statistical Analysis

The demographic and sleep data are presented as means with a range of 1 standard deviation. Kruskal‐Wallis rank‐sum test was performed to evaluate for significant differences between the 3 groups. If the Kruskal‐Wallis analysis rejected the null hypothesis, Mann‐Whitney tests with Bonferroni adjustment were performed to test for significance between the 3 groups. For each RSWA metric, an experiment‐wise Bonferroni correction factor for multiple comparisons (PD vs. ET, ET vs. control, PD vs. control) was applied, setting significance at an α of P < 0.017. All statistical analysis was done using R software.

Data Availability Statement

The raw de‐identified data that support the findings of this study are available to qualified investigators on request from the corresponding author. The data are not publicly available because of Mayo Clinic Policy and the Health Insurance Portability and Accountability Act. All the relevant data have been shared and published in this article.

Results

Demographic and Clinical Characteristics

The average age of our cohort was 70.6 ± 9.0 years old with 60 (82.2%) men. There were no significant differences of age or sex between the 3 groups of patients with ET, PD, or controls. Most patients were referred to the Sleep Center for evaluation of obstructive sleep apnea (n = 57) or snoring (n = 12) although 4 patients had insomnia as their primary reason for referral. None of the patients in the PD, ET, or control groups had reported dream enactment behaviors. Patients with PD experienced PD symptoms for an average of 31.8 (range = 1–144) months before PSG, whereas ET patients had been experiencing ET symptoms for an average of 56.4 (range = 1–312) months before PSG. There was no significant difference between PD and ET symptom duration (P = 0.538). At PSG, 15 ET patients (65.2%) were on medication for their tremor. One ET patient was taking levodopa (l‐dopa) (n = 1) and 1 was taking pramipexole (n = 1), although neither of these medications were determined to be effective in these 2 patients at follow‐up. Nineteen PD patients (82.6%) were on carbidopa‐l‐dopa with an average l‐dopa daily dose of 456 mg per day at PSG. One PD patient was on rotigotine and another PD patient was on ropinirole in addition to l‐dopa treatment. None of the control patients were taking dopaminergic therapies and 8 of them were on metoprolol. Only 2 patients in total (1 PD, 1 control) were taking benzodiazepines at PSG. All patients with PD had at least 1 characteristic symptom of PD (gait instability, bradykinesia, rest tremor, and muscle rigidity), although 4 ET (17.4%) and 2 control patients (7.4%) reported at least 1 of these symptoms. In patients diagnosed with ET, all reported both a kinetic or postural tremor, and 11 (47.8%) reported a family history of ET. Only 1 (4.4%) PD patient reported a postural tremor related to their PD diagnosis, whereas 2 (11.1%) control patients reported a kinetic tremor likely related to other etiologies (transient ischemic attack and syncope). Six ET patients received dopamine transporter uptake (DaT) scans that did not show evidence of PD. Clinical characteristics of PD and ET and related medications are described further in Table 1. There were also no significant differences in PSG parameters between groups, apart from significantly increased REM sleep time in controls compared to ET and total sleep time in controls compared to PD and ET patients, however, controls had a non‐significantly lower percentage of REM sleep time than PD and ET patients (Table S1).

TABLE 1.

Demographics and clinical characteristics of patients with PD, essential tremor, and controls

	PD(n = 23)	ET(n = 23)	Controls(n = 27)
Demographics
Sex (M/F)	18/5	20/3	22/5
Age at PSG	70.83 ± 9.85	70.65 ± 8.82	70.48 ± 8.79
Age at diagnosis	68.65 ± 9.75	70.17 ± 9.16	N/A
Duration of PD/ET symptoms (mon)	31.83 ± 42.57	56.43 ± 78.88	N/A
Mild cognitive impairment	4 (17.4%)	6 (26.0%)	2 (7.4%)
DaT scan	0 (0.0%)	6 (26.0%)	0 (0.0%)
PD characteristics
Bradykinesia	12 (52.2%)	1 (4.4%)	0 (0.0%)
Postural instability	11 (47.8%)	3 (13.0%)	0 (0.0%)
Rest tremor	17 (73.9%)	3 (13.0%)	1 (3.7%)
Rigidity	18 (78.3%)	2 (8.7%)	1 (3.7%)
Gait instability	17 (73.9%)	3 (13.0%)	2 (7.4%)
Any PD symptom	23 (100%)	4 (17.4%)	2 (7.4%)
Levodopa	19 (82.6%)	1 (4.4%)	0 (0.0%)
Pramipexole	1 (4.3%)	1 (4.4%)	0 (0.0%)
Essential tremor characteristics
Kinetic tremor	0 (0.0%)	20 (87.0%)	2 (7.4%)
Postural tremor	1 (4.3%)	13 (56.5%)	0 (0.0%)
Both (kinetic/postural)	0 (0.0%)	10 (43.4%)	0 (0.0%)
Family history of ET	1 (4.3%)	11 (47.8%)	0 (0.0%)
Propranolol	0 (0.0%)	5 (21.7%)	0 (0.0%)
Primidone	0 (0.0%)	3 (13.0%)	0 (0.0%)
Topiramate	0 (0.0%)	1 (4.4%)	0 (0.0%)
Gabapentin	0 (0.0%)'	3 (13.0%)	0 (0.0%)
Metoprolol	4 (17.4%)	3 (13.0%)	8 (29.6%)
Other tremor medication	2 (8.7%)	0 (0.0%)	2 (7.4%)
No tremor/PD medication	2 (8.7%)	8 (34.8%)	17 (63.0%)

Age and duration of symptoms are presented here as means and standard deviation, whereas the descriptive statistics are all presented as counts % of patients in subgroup.

RSWA Metrics

Kruskal‐Wallis analysis between the 3 groups showed a significant difference in 6 of 9 RSWA metrics analyzed (Table 2). PD patients had significantly higher SM phasic activity than ET (12.5 ± 12.8 vs. 4.9 ± 6.7, P = 0.002) and controls (12.5 ± 12.8 vs. 4.0 ± 2.6, P = 0.001). SM “any” was also significantly higher in PD than ET (13.5 ± 14.3 vs. 5.2 ± 7.6, P = 0.002) and controls (13.5 ± 14.3 vs. 4.2 ± 2.6, P = 0.002). Automated RAI was significantly lower (consistent with higher amounts of RSWA) in PD than in ET (0.78 ± 0.23 vs. 0.92 ± 0.09 P = 0.002) and controls (0.78 ± 0.23 vs. 0.90 ± 0.17, P = 0.005) but no different between ET and healthy controls in the SM (Fig. 1). Combined SM + AT phasic and “any” muscle activity did not differ between the 3 groups (Fig. S1). SM tonic and AT phasic and “any” muscle activity were no different between any of the groups. The number of patients meeting previously defined PD‐RBD diagnostic cutoffs are listed in Table 3. ²⁴

TABLE 2.

REM sleep without atonia metrics for patients with PD, essential tremor, and controls

	PD(n = 23)A	ET(n = 23)B	Controls(n = 27)C	P‐value(<0.017)
SM duration (s)	0.73 ± 0.39	0.63 ± 0.32	0.54 ± 0.27
AT duration (s)	0.97 ± 0.57	0.74 ± 0.49	0.48 ± 0.22	A > C
SM phasic %	12.53 ± 12.83	4.95 ± 6.68	3.99 ± 2.64	A > B, A > C
SM any %	13.54 ± 14.30	5.17 ± 7.60	4.15 ± 2.58	A > B, A > C
AT phasic %	19.00 ± 15.88	17.21 ± 12.90	13.04 ± 10.49
AT any %	19.30 ± 16.28	17.40 ± 13.24	13.04 ± 10.49
Tonic %	1.78 ± 5.00	0.36 ± 1.74	0.22 ± 0.77
SM + AT phasic %	27.47 ± 18.57	20.85 ± 14.28	16.37 ± 10.36
SM + AT any %	28.52 ± 19.95	21.19 ± 14.97	16.53 ± 10.25
RAI	0.78 ± 0.23	0.92 ± 0.09	0.90 ± 0.17	A < B, A < C

Means and standard deviations are displayed for each RSWA metric. Group comparisons used 3‐way Kruskal–Wallis analysis followed by a Mann–Whitney U test with Bonferroni adjustment, with significant differences listed in final column; each group represented by PD = A, ET = B, and control = C.

SM, submentalis muscle; AT, anterior tibialis muscle; RAI, REM atonia index.

FIG 1.

RSWA metrics for patients with Parkinson's disease, essential tremor, and controls. Boxplots of RSWA metrics including (A) SM phasic and (B) “any,” (C) AT phasic and (D) “any,” as well as (E) RAI and (F) tonic are displayed. Significant group differences (*P < 0.017, **P < 0.01, ***P < 0.001) are marked by asterisks above the plots

TABLE 3.

Number of subjects that met previously defined diagnostic cutoffs for PD‐RBD ²⁴

	PD‐RBD cutoff	PD(n = 23)	ET(n = 23)	Controls(n = 27)
SM duration (s)	0.65	56.5% (13)	34.8% (8)	22.2% (6)
AT duration (s)	0.79	56.5% (13)	34.8% (8)	7.4% (2)
SM phasic %	15.5	30.4 (7)	8.7 (2)	0.0 (0)
SM any %	21.6	17.4 (4)	4.3 (1)	0.0 (0)
AT phasic %	30.2	21.7 (5)	21.7 (5)	11.1 (3)
AT any %	30.2	21.7 (5)	21.7 (5)	11.1 (3)
Tonic %	1.2	17.4 (4)	4.3 (1)	3.7 (1)
SM + AT phasic %	37.8	21.7 (5)	13.0 (3)	3.7 (1)
SM + AT any %	43.4	21.7 (5)	4.3 (1)	0.0 (0)
RAI	0.88	48.8% (11)	21.7% (5)	14.8% (4)
No. of subjects ≥1 cutoff	–	12 (52.17%)	8 (34.78%)	9 (33.33%)
No. of subjects ≥2 cutoffs	–	9 (39.13%)	5 (21.74%)	0 (0.0%)

The number and percentage of subjects meeting each RSWA individual diagnostic threshold metric, and those meeting 1 or 2 thresholds (excluding duration) are shown.

Two patients (8.7%) with ET were identified as outliers, with their SM phasic and “any” RSWA higher than the mean values for the PD group as well as meeting cutoffs for RBD diagnosis (Fig. 1A,B). The first patient had a 26‐year history of familial ET at the time of the PSG. Initially, the tremor was noted to be more prominent on the right side, but at the time of the last neurologic examination at the age of 85 years, 2 years following the PSG, the tremor was reported as being symmetric with no features to suggest parkinsonism. The second patient was diagnosed with an ET at the time of the PSG but 3 months later was noted to also have a tremor in the right thumb at rest with reduced swing of the right arm. When last seen at the age of 73 years, 4 years following the PSG, he continued to show features compatible with mild parkinsonism, but a DaT scan was normal.

Discussion

We found that isolated SM RSWA is more frequent in PD than in patients with ET providing additional evidence for SM RSWA as a biomarker for underlying synucleinopathy. Only 2 patients (8.7%) with ET met our previously published SM phasic RSWA cutoffs for RBD diagnosis. Interestingly, both of these patients had very mild evidence of parkinsonism (asymmetric tremor in 1, bradykinesia, mild rest tremor yet a negative DaT scan in the other, which was more suggestive of an ET than PD diagnosis in that case at the time of last follow‐up). These findings are suggestive of a mild comingling synucleinopathy with possible early or prodromal parkinsonism to explain their elevated SM RSWA. This percentage of ET patients with quantitatively elevated RSWA is lower than identified in some previous studies that qualitatively examined the prevalence of RBD or subclinical RBD, ¹⁷ , ¹⁸ , ¹⁹ , ²¹ but confirms that a minority of ET patients have abnormal REM sleep muscle activity. Concurrent Lewy body pathology has been described in 13% to 24% of ET patients, with rare patients showing preferential or even selective focal Lewy body pathology deposits in the locus coeruleus. ⁴ , ³⁰ , ³¹ Extension of Lewy body pathology to neighboring REM sleep atonia control centers, such as the subcoeruleus nucleus, could explain the findings of RSWA in some ET patients in previous cohorts. Although additional studies are required for further validation, overall, the presence of elevated SM RSWA is suggestive of underlying syuncleinopathy in patients with tremor disorders. Longitudinal studies of RSWA in patients with ET in parallel to serial evaluation for development of parkinsonism and other objective markers of striatonigral degeneration such as DaT scanning are needed to better understand the relationship between RSWA and concurrent synucleinopathy in ET patients.

We did not find a significant difference in AT RSWA between all 3 patient groups. This is consistent with our prior work showing that SM but not AT RSWA is specific for underlying synucleinopathy in parkinsonism and dementia. ³² , ³³ The reason is unclear, but may be related to somatotopic organization of REM sleep atonia control, with the motor control of AT being vulnerable to multiple processes in addition to α‐synuclein neurodegeneration including aging. Alternatively, because we do not exclude scorable REM periodic limb movements (PLMs) when quantifying AT RSWA (given the difficulty of distinguishing true PLMs during REM sleep, and the lack of any previously published methodology to distinguish PLMs from aperiodic RSWA), our AT metrics could be consequently inflated. Yet, our AT RSWA metrics closely parallel those of other approaches that exclude PLMs from RSWA calculations and our findings are consistent with other studies that found weak discrimination between RBD and controls in the lower limbs. ²⁸ , ³⁴ The similarities in AT RSWA levels between our patient and control groups despite the differences in PLM index make it unlikely that PLMs influenced our results.

Our study has several limitations, including probable sampling and referral biases. Whereas our predominantly male cohort may limit the generalizability of our results, a male sex majority is also typical of RBD and PD patient populations so it is naturalistic for these patient populations. Our cohort had well characterized ET and PD on clinical examination, thus our data does not support use of PSG in differentiating clinical tremor syndromes, however, it is possible that clinical diagnostic inaccuracy led to misclassification of ET and PD, or that some patients had overlapping PD and ET. Future prospective studies should also sample arm EMG muscles, which were not available in our retrospective data set. Arm EMG analyses for RSWA would be of particular interest because most PD and ET tremor impacts the arms and arm EMG muscle activity has been demonstrated to have optimal specificity and sensitivity compared to other EMG channels in discriminating between RBD and controls. ²⁸ It is also important to continue to examine RSWA and its relationship with other common biomarkers of PD. ³⁵ Further large‐scale studies with ante mortem diagnosis of PD and ET confirmed by DaT scanning as well as movement disorder laboratory defined tremor types are needed to confirm and expand the potential utility and significance of our findings.

Conclusions

Elevated SM REM sleep without atonia is common in PD patients even in the absence of clinical RBD, but is rare in ET. These data provide further evidence that quantitative RSWA is an ante mortem marker of underlying synucleinopathy in patients with clinical motor or cognitive disorders. ³² , ³³ Further prospective studies with definitively diagnosed ET and PD (with pathology or DaT scan‐supported diagnoses with electrophysiological tremor analysis) are needed to confirm and extend our findings.

Author Roles

(1) Research project: A. Conception, B. Organization, C. Execution; (2) Statistical Analysis: A. Design, B. Execution, C. Review and Critique; (3) Manuscript Preparation: A. Writing of the First Draft, B. Review and Critique.

T.R.G.: 1A, 1B, 1C, 2B, 3A

S.J.M.: 1A, 1B, 1C, 2B, 2C, 3B

E.G.: 1C, 2A, 2B, 3B

J.C.F.: 1C, 2B, 3B

P.C.T.: 1C, 2C, 3B

L.N.T.: 1C, 2B, 2C, 3B

C.L.R.: 1C, 2B, 2C, 3B

S.M.W.: 1C, 2B, 2C, 3B

J.P.C.: 1C, 2B, 3B

J.T.J.: 1C, 2B, 2C, 3B

C.D.O.: 1C, 2B, 2C, 3B

L.J.E.: 1C, 2B, 2C, 3B

E.L.V.: 1C, 2B, 2C, 3B

R.S.: 1C, 2B, 2C, 3B

B.F.B.: 1C, 2B, 2C, 3B

M.H.S.: 1C, 2B, 2C, 3B

E.K.S.: 1A, 1B, 1C, 2A, 2C, 3B

Disclosure

Ethical Compliance Statement: The Mayo Clinic Institutional Review Board approved this study. Participating patients (or their legally authorized representatives) provided written consent to use their medical information. We confirm that we have read the Journal's position on issues involved in ethical publication and affirm that this work is consistent with those guidelines.

Funding Sources and Conflict of Interest: The project described was supported by a Mayo Clinic Alzheimer's Disease Research Center Grant Award from the National Institute on Aging (P50 AG016574), and the National Center for Research Resources, and the National Center for Advancing Translational Sciences, NIH (1 UL1 RR024150‐01). The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIH. The authors declare that there are no conflicts of interest relevant to this work.

Financial Disclosures for the Previous 12 Months: R.S. receives research support from the National Institute on Aging, the National Institute of Neurological Disorders and Stroke, and the Mayo Clinic Small Grants Program National Center for Advancing Translational Sciences (NCATS) and unrestricted research grant from Acadia Pharmaceutical, Inc. B.F.B. receives funding from the National Institute on Aging, the National Institute of Neurological Disorders and Stroke, the Lewy Body Dementia Association, GE Healthcare, Axovant Sciences, Inc., and Biogen. E.K.S. reports research support from the National Center for Research Resources, the National Center for Advancing Translational Sciences, and the National Institute on Aging. All other authors have no disclosures to report.

Supporting information

Figure S1 Additional RSWA metrics for patients with Parkinson's disease, essential tremor, and controls. Boxplots of RSWA metrics, including (A) combined SM and AT phasic, (B) “any,” (C) both SM duration, and (D) AT duration. Significant group differences (*P > 0.017, **P > 0.01, ***P > 0.001) are marked by asterisks above the plots

Table S1 Polysomnographic data for patients with Parkinson's disease, essential tremor, and controls. Means and standard deviations for each sleep metric are displayed. Group comparisons used 3‐way Kruskal‐Wallis analysis followed by a Mann‐Whitney U test with Bonferroni adjustment, with significant differences listed in the final column; groups represented by PD = A, ET = B, and Control = C