Secondary Logo

Journal Logo

Original Article

Essential tremor-Parkinson's disease syndrome: clinical characteristics and subtypes using cluster analysis

Hou, Yanbing; Han, Qin; Ou, Ruwei; Liu, Kuncheng; Lin, Junyu; Yang, Tianmi; Shang, Huifang

Editor(s): Yin, Yanjie; Hao, Xiuyuan

Author Information
Chinese Medical Journal: August 01, 2022 - Volume - Issue - 10.1097/CM9.0000000000002229
doi: 10.1097/CM9.0000000000002229



Essential tremor (ET) and Parkinson's disease (PD) are common among movement disorders.[1] The prevalence of ET is estimated at 0.9% which increases to 4.6% in individuals >65 years of age[2]; the prevalence of PD is estimated at 0.3% which increases to 1% in individuals >60 years of age.[3] The relationship between ET and PD remains unclear; however, it is clear that patients with ET are at risk of developing PD.[4,5] ET-PD syndrome is characterized by the occurrence of PD in patients with a history of ET, which may be an independent phenotype distinct from PD.

Several studies have focused on the comparison of clinical features between ET-PD and PD patients[6-10]; however, the results for non-motor symptoms are controversial. Compared to PD patients, ET-PD patients exhibit milder rigidity and distinct tremor.[6] ET-PD patients may have a mild sleep disorder and hyposmia,[7] severe sleep and cognitive problems,[8] or impaired cognitive functions which could be correlated with the disease duration.[9] Both ET and PD are clinically heterogeneous; so, it is reasonable to speculate that ET-PD may have different clinical subtypes. To the best of our knowledge, no study investigates the clinical subtypes classification of ET-PD.

Therefore, the current study aimed to characterize the clinical features of ET-PD, evaluate the differences in clinical features between ET-PD and PD patients, and identify the clinical subtypes of ET-PD.


Ethical approval

The current study was approved by the Ethics Committee of West China Hospital, Sichuan University, and written informed consent was obtained from all participants.


Newly diagnosed patients from the movement disorders outpatient clinic, West China Hospital of Sichuan University, were enrolled during the period between November 2014 to July 2020. None of the patients had taken any anti-Parkinsonian medication before the enrollment. The PD diagnosis was based on the United Kingdom PD Society Brain Bank criteria.[11] The diagnosis of ET-PD was based on the reported clinical diagnostic criteria from previous studies[6-9] and included the following: (1) ET diagnosis was presented for more than 5 years before the PD diagnosis; (2) initial ET was characterized by moderate-or-large amplitude action tremor in the absence of any PD symptoms (bradykinesia, resting tremor, postural instability, etc.); and (3) initial ET diagnosis occurred without any red flags for possible PD emergence. The clinical history was obtained from patients or their relatives and other data were extracted from their medical records when available to retrospectively establish the initial ET diagnosis. A total of 93 ET-PD patients and 93 matched PD patients were recruited in the current study, and all patients were followed up at least once to reconfirm the diagnoses by face-to-face visits or telephone interview.

Clinical assessments

The complete medical history, neurological examination, and clinical assessment data were recorded through a series of structured questionnaires. A family history of ET was confirmed if the patients’ relatives had kinetic tremor in the absence of any other neurological symptoms for >5 years. A family history of PD was confirmed when patients’ relatives were diagnosed with PD as per their medical records. Motor symptoms were assessed using the unified PD rating scale (UPDRS) divided according to tremor (sum of items 16, 20, and 21), hypokinesia (sum of items 22 and 31), and postural instability (sum of items 13, 14, 15, 29, and 30).[12] Overall severity of non-motor symptoms was assessed using the non-motor symptoms scale (NMSS). Orthostatic hypotension symptom was assessed using the sum of NMSS D1 (sum of items 1 and 2). Cognitive functions were assessed through the Montreal cognitive assessment (MoCA).[13] Psychiatric manifestations were assessed using the Hamilton depression scale (HAMD)[14] and the Hamilton anxiety scale (HAMA).[15] Sleep disorders were assessed through the PD sleep scale-2 (PDSS-2).[16] The patient's quality of life was evaluated using the Chinese version of the PD questionnaire 39 (PDQ-39).[17]

Statistical analysis

Continuous variables were analyzed using analysis of variance analysis with Bonferroni comparisons when the data was normally distributed; Kruskal–Wallis test was used for non-normally distributed data. Categorical data were compared using the χ2 test or the Fisher exact test. The differences of clinical characteristics were compared between the ET-PD and PD groups. Considering the confounding effects of age, sex, education, and disease duration of PD on clinical characteristics, propensity scores matching analysis was used to balance the four variables.

We selected variables involving different demographic data and clinical symptoms, in particular those that varied in the current comparisons between ET-PD and PD groups or were controversial based on the previous studies.[8,9] Further, a multi-collinearity diagnostic test was performed [Supplementary Table 1,]. Nine continuous variables including duration of ET before PD onset, UPDRS tremor score, UPDRS hypokinesia score, UPDRS postural instability score, NMSS orthostatic score, MoCA score, HAMD score, HAMA score, PDSS-2 score, and two categorical variables, including a family history of ET and hyposmia, were finally selected for subsequent cluster analysis.

Each variable was transformed into a z-score. The sum of squared error (SSE) was used to determine the optimal number of clusters. As the number of clusters increased, the “within sum of square” value concomitantly decreased. We focused on the fast change in the slope, namely the “elbow point”. It meant that when there was a further increase in the number of clusters, the WWS did not undergo significant enhancement. This “elbow point” was taken as the optimal number of clusters. After determining the number of clusters, the “k-means” non-hierarchical method was used for adjustment. To test the utility of this subgroup classification, we investigated the differences between the clusters. All statistical tests were two-tailed and a P value ≤0.05 was considered statistically significant. All statistical analyses were performed using the R software (version 3.6.3; R Foundation, Vienna, Austria).


Comparison between ET-PD and PD patients

A total of 93 newly diagnosed ET-PD patients and 93 newly diagnosed PD patients correspondingly matched for age, sex, education, and disease duration of PD were included in the study. The ET-PD group had a higher number of patients with a family history of ET, while the PD group had a higher number of patients with a family history of PD. ET-PD patients showed evident tremor, but milder hypokinesia and postural instability symptoms, relative to PD patients. The ET-PD group had a significantly lesser number of patients with hyposmia with lower NMSS scores. There were no significant differences in MoCA, HAMD, HAMA, PDSS-2, and PDQ-39 scores [Table 1].

Table 1 - Characteristics of ET-PD and PD patients.
Characteristicsa ET-PD (n = 93) ET-PD cluster 1 (n = 34) ET-PD cluster 2 (n = 59) PD (n = 93) P valueb P valuec P valued P valuee P valuef
Age, years 59.60 (12.29) 59.10 (14.22) 59.88 (11.15) 58.31 (11.63) 0.466 0.770 0.733
Female, n (%) 48 (51.61) 22 (64.71) 26 (44.07) 48 (51.61) 1.000 0.084 0.159
Education, years 10.41 (4.02) 10.32 (4.14) 10.47 (3.99) 10.67 (3.41) 0.645 0.870 0.886
Duration of PD, years 2.28 (1.93) 2.24 (1.97) 2.31 (1.92) 2.08 (1.47) 0.415 0.875 0.707
Duration of ET, years 17.93 (9.55) 15.13 (6.23) 19.54 (10.75) 0.014
Duration of ET prior to PD onset, years# 15.64 (9.33) 12.89 (5.85) 17.23 (10.57) 0.012
Family history of PD, n (%) 8 (8.60) 3 (8.82) 5 (8.47) 18 (19.35) 0.034 1.000 0.107
Family history of ET, n (%)# 52 (55.91) 12 (35.29) 40 (67.80) 0 (0) <0.001 0.004 <0.001 <0.001 <0.001
UPDRS tremor score# 7.54 (3.63) 8.03 (3.66) 7.25 (3.60) 5.11 (3.54) <0.001 0.323 <0.001 <0.001 0.001
UPDRS hypokinesia score# 5.68 (3.50) 6.41 (3.83) 5.25 (3.26) 8.82 (4.34) <0.001 0.125 <0.001 0.008 <0.001
UPDRS instability score# 0.41 (0.66) 0.68 (0.84) 0.25 (0.48) 1.32 (1.08) <0.001 0.010 <0.001 0.001 <0.001
Hyposmia, n (%)# 14 (15.05) 6 (17.65) 8 (13.56) 31 (33.33) 0.004 0.764 0.013 0.062 0.003
NMSS score 16.66 (16.81) 28.62 (20.43) 9.76 (8.85) 23.71 (25.11) 0.026 <0.001 <0.001 0.696 <0.001
D1. Cardiovascular# 0.41 (0.93) 0.88 (1.27) 0.14 (0.51) 0.68 (1.49) 0.142 0.002 0.007 1.000 0.025
D2. Sleep/fatigue 3.95 (5.28) 7.24 (6.55) 2.05 (3.13) 3.88 (4.58) 0.929 <0.001 <0.001 0.001 0.055
D3. Mood/apathy 3.67 (6.40) 7.82 (8.76) 1.27 (2.32) 6.73 (12.81) 0.041 <0.001 0.001 1.000 0.003
D4. Perceptual problems/hallucinations 0.15 (0.78) 0.24 (1.04) 0.10 (0.58) 0.03 (0.18) 0.155 0.429 0.200
D5. Attention/memory 2.77 (3.59) 4.18 (4.77) 1.97 (2.39) 3.37 (4.48) 0.322 0.015 0.024 0.939 0.111
D6. Gastrointestinal 1.19 (2.73) 1.50 (3.54) 1.02 (2.14) 1.44 (4.08) 0.628 0.414 0.722
D7. Urinary 2.19 (3.56) 2.79 (3.79) 1.85 (3.41) 2.80 (3.97) 0.278 0.219 0.283
D8. Sexual dysfunction 0.40 (1.64) 0.47 (1.83) 0.36 (1.54) 1.04 (3.44) 0.104 0.748 0.263
D9. Miscellaneous 1.89 (2.98) 3.35 (3.88) 1.05 (1.89) 3.74 (4.11) 0.001 0.002 <0.001 1.000 <0.001
MOCA score# 24.42 (3.94) 24.44 (4.06) 24.40 (3.91) 25.09 (3.35) 0.216 0.968 0.465
HAMD score# 6.15 (6.57) 12.12 (7.09) 2.71 (2.66) 7.26 (7.41) 0.282 <0.001 <0.001 <0.001 <0.001
HAMA score# 5.87 (6.83) 10.82 (8.55) 3.02 (3.11) 5.86 (6.16) 0.991 <0.001 <0.001 <0.001 0.014
PDSS-2 score# 9.86 (8.50) 16.12 (9.89) 6.25 (4.76) 8.13 (8.21) 0.159 <0.001 <0.001 <0.001 0.430
PDQ-39 score 20.50 (14.70) 28.79 (16.69) 15.71 (10.98) 22.67 (19.67) 0.395 <0.001 0.001 0.212 0.042
aAll data are presented as mean (standard deviation) values unless otherwise indicated.
bDifference between ET-PD and PD groups.
cDifference between the two ET-PD clusters.
dDifference among ET-PD Cluster 1, ET-PD Cluster 2 and PD groups.
ePost-hoc analysis: Difference between ET-PD Cluster 1 and PD groups.
fPost-hoc analysis: Difference between ET-PD Cluster 2 and PD groups.
#Variables included in the cluster analysis for ET-PD.
Statistical difference between two groups.ET: Essential tremor; HAMA: Hamilton anxiety scale; HAMD: Hamilton depression scale; MoCA: Montreal cognitive assessment; NMSS: Non-motor symptoms scale; PD: Parkinson's disease; PDQ-39: PD questionnaire 39; PDSS: PD sleep scale; UPDRS: Unified PD rating scale.

Cluster analysis

A two-cluster model of ET-PD patients was found to be the most appropriate [Figure 1A, and 1B]. The critical determinants of ET-PD subtypes were HAMD and HAMA scores [Figure 1C]. The ET-PD cluster 1 (n = 34) and the ET-PD cluster 2 (n = 59) had similar age, sex, education, and disease duration of PD. The ET-PD cluster 1 had a shorter duration of ET before PD onset, lesser number of patients with a family history of ET, higher UPDRS instability scores, higher non-motor symptoms scores (NMSS D1 scores, HAMD scores, HAMA scores, and PDSS-2 scores), and higher PDQ-39 scores relative to the ET-PD cluster 2. There were no significant differences in the UPDRS tremor scores, UPDRS hypokinesia scores, MoCA scores, and the number of patients with hyposmia between the two ET-PD clusters [Table 1 and Figure 1D and 1E].

Figure 1:
(A) The selection of optimal cluster numbering based on WSS values. The slope changes quickly when the cluster numbering increases from 1 to 2. The elbow point is denoted as 2; (B) Graphical dispersion of PD-ET for the two main components through PCA. As indicated, the two clusters are completely separated; (C) As shown, HAMD score, HAMA score, and PDSS-2 score contribute largely to the Dim 1, while MoCA score, duration of ET before PD onset, UPDRS hypokinesia score, and postural instability score contribute largely to the Dim 2; (D) Significantly altered clinical characteristics between the two ET-PD clusters are indicated by an asterisk (∗P < 0.05; †P < 0.01; ‡P < 0.001); (E) The ET-PD cluster 1 had lesser number of patients with a family history of ET relative to the ET-PD cluster 2. There was no significant difference in the number of patients with hyposmia between the two ET-PD clusters. Dim: Dimension; ET: Essential tremor; HAMA: Hamilton anxiety scale; HAMD: Hamilton depression scale; MoCA: Montreal cognitive assessment; NMSS: Non-motor symptoms scale; PCA: Principal component analysis; PD: Parkinson's disease; PDQ-39: PD questionnaire 39; PDSS: PD sleep scale; PDSS-2: PD sleep scale-2; UPDRS: Unified PD rating scale; WSS: Within sum of square.

Comparisons among two ET-PD clusters and the PD group

As compared to the PD group, both ET-PD clusters showed prominent tremor, but slight hypokinesia and postural instability symptoms; the ET-PD cluster 1 had higher HAMD scores, HAMA scores, and PDSS-2 scores, while the ET-PD cluster 2 had lower HAMD scores, lower HAMA scores, lesser number of patients with hyposmia, lower NMSS scores, and lower PDQ-39 scores [Table 1].


We found that ET-PD could be classified into two significantly different subtypes through cluster analysis using a relatively large sample of newly diagnosed ET-PD patients, namely the ET-PD simple type and the ET-PD complex type. The ET-PD simple type represented a longer duration of ET before PD onset, milder non-motor symptoms, and postural instability symptoms; patients in this group were more likely to have a family history of ET. The ET-PD complex type represented a shorter duration of ET before PD onset, outstanding non-motor symptoms, and relatively prominent postural instability symptoms; patients in this group were less likely to have a family history of ET. Notably, patients of the ET-PD complex type were tightly coupled with worse quality of life as compared to those belonging to the ET-PD simple type. This was the first study on the classification of clinical subtypes of ET-PD and may have profound clinical significance to provide a better understanding of ET-PD and its clinical characteristics. Patients of the ET-PD simple type usually show mild clinical symptoms and have a better quality of life, while patients of the ET-PD complex type have relatively severe clinical symptoms and poor quality of life; they need active intervention.

Current literature indicates that the mean latency between onset of ET and onset of PD symptoms is 14 years, and >30% of ET-PD patients have a long latency (>20 years) from the onset of ET to the first sign of PD base on neurological examination.[18] The mean latency for the ET-PD simple type is 17.23 years, while the mean latency for the ET-PD complex type is 12.89 years. A previous study using 123I-Ioflupane SPECT investigated the striatal dopamine transporter (DAT) binding in ET patients, tremor dominant PD patients, and healthy controls (HCs), and reported that the ET group had higher uptake values relative to the PD group but lower uptake values relative to HCs (trend of HCs > ET group > PD group).[19] Notably, a DAT imaging study further analyzed the striatal binding ratio (SBR) in ET patients, ET-PD patients, and HCs, and showed that the SBR was slightly lower in the ET group and evidently lower in the ET-PD group as compared to HCs (trend of HCs > ET group > ET-PD group).[20] We speculated that a dopaminergic deficit may be sustained in the duration of ET before PD onset, and longer/shorter latency may correspond to slower/faster dopaminergic deficit, respectively. The ET-PD complex type with shorter latency may show a faster dopaminergic deficit before PD onset, and the ET-PD simple type with longer latency may exhibit a slower dopaminergic deficit before PD onset.

A previous study comprising >400 families of probands with ET, PD, or ET-PD assessed the family history information of ET and PD. A higher proportion of PD probands’ relatives with PD than relatives with ET was found, while a higher proportion of ET or ET-PD probands’ relatives with ET than relatives with PD was also reported.[21] ET-PD probands had similar patterns of familial risk as ET probands rather than PD probands, which implies that ET-PD may have higher genetic similarity to ET as compared to PD. Our data were consistent with the published findings on familial aggregation. We found a greater proportion of ET-PD probands’ relatives with ET, and even significantly more probands with the ET-PD simple type having a family history of ET. The ET-PD simple type with a higher prevalence of ET family history suggested that the ET-PD simple type and ET had greater similarities than between the ET-PD complex type and ET.

With similar duration as Parkinsonian symptoms, the two ET-PD subtypes showed significantly different motor symptoms, which were characterized by higher tremor scores and lower hypokinesia/instability scores in comparison to PD patients. These findings were partially in agreement with previous studies.[6-8] The ET-PD complex type exhibited higher instability than the ET-PD simple type, albeit to a lesser extent than the PD group. The hypokinesia/instability in PD may be attributed to basal ganglia dysfunction, which suggested greater involvement of this system in patients with the ET-PD complex type.

The current data showed that hyposmia, often as a prodromal symptom of PD, was far less common in ET-PD patients. A previous study also reported an interesting trend for the proportion of patients with hyposmia (PD group > ET-PD group > ET group).[6] The hyposmia did not differ significantly between the two ET-PD subtypes. However, the ET-PD complex type exhibited remarkable non-motor symptoms including mood disturbance, sleep disorder, and worse quality of life. These impairments suggested widespread neurological involvement in ET-PD with a significant overlap in changes associated with PD. The different proportions in these two ET-PD subtypes in the cohort of ET-PD patients may explain the controversy surrounding the differences in non-motor symptoms between ET-PD and PD groups in previous studies.

There are some limitations in this study. First, olfaction and orthostatic symptoms were obtained from questionnaires instead of more objective examinations. Second, we used clinical criteria for the diagnoses and the diagnoses were not confirmed by postmortem. However, all patients were carefully evaluated by a professional movement specialist having >20 years of experience in movement disorders. Finally, the current study did not include genetic information for each subject and we intend to perform genetic analysis in our subsequent research.


The ET-PD patients had significantly different characteristics for motor symptoms as compared to the PD patients and may be divided further into two clinical subtypes, namely, the ET-PD simple type and the ET-PD complex type.

Conflicts of interest



1. Haubenberger D, Hallett M. Essential tremor. N Engl J Med 2018;378:1802–1810. doi: 10.1056/NEJMcp1707928.
2. Louis ED, Ferreira JJ. How common is the most common adult movement disorder? Update on the worldwide prevalence of essential tremor. Mov Disord 2010;25:534–541. doi: 10.1002/mds.22838.
3. de Lau LML, Breteler MMB. Epidemiology of Parkinson's disease. Lancet Neurol 2006;5:525–535. doi: 10.1016/s1474-4422(06)70471-9.
4. Tallon-Barranco A, Vazquez A, Jimenez-Jimenez FJ, Orti-Pareja M, Gasalla T, Cabrera-Valdivia F, et al. Clinical features of essential tremor seen in neurology practice: a study of 357 patients. Parkinsonism Relat Disord 1997;3:187–190. doi: 10.1016/s1353-8020(97)00031-x.
5. Benito-León J, Louis ED, Bermejo-Pareja F. Neurological disorders in central Spain study group. Risk of incident Parkinson's disease and parkinsonism in essential tremor: a population based study. J Neurol Neurosurg Psychiatry 2009;80:423–425. doi: 10.1136/jnnp.2008.147223.
6. Arabia G, Lupo A, Manfredini LI, Vescio B, Nistico R, Barbagallo G, et al. Clinical, electrophysiological, and imaging study in essential tremor-Parkinson's disease syndrome. Parkinsonism Relat Disord 2018;56:20–26. doi: 10.1016/j.parkreldis.2018.06.005.
7. Ryu DW, Lee SH, Oh YS, An JY, Park JW, Song IU, et al. Clinical characteristics of Parkinson's disease developed from essential tremor. J Parkinsons Dis 2017;7:369–376. doi: 10.3233/JPD-160992.
8. Louis ED, Wise A, Alcalay RN, Rao AK, Factor-Litvak P. Essential tremor-Parkinson's disease: a double whammy. J Neurol Sci 2016;366:47–51. doi: 10.1016/j.jns.2016.04.040.
9. Louis ED, Rohl B, Collins K, Cosentino S. Poorer cognitive performance in patients with essential tremor-Parkinson's disease vs. patients with Parkinson's disease. Front Neurol 2015;6:106. doi: 10.3389/fneur.2015.00106.
10. Simoes RM, Constantino A, Gibadullina E, Houghton D, Louis ED, Litvan I. Examining the motor phenotype of patients with both essential tremor and Parkinson's disease. Tremor Other Hyperkinet Mov (N Y) 2012;2. tre-02-47-149-3. doi: 10.7916/D8CN72N0.
11. Hughes AJ, Daniel SE, Kilford L, Lees AJ. Accuracy of clinical diagnosis of idiopathic Parkinson's disease: a clinico-pathological study of 100 cases. J Neurol Neurosurg Psychiatry 1992;55:181–184. doi: 10.1136/jnnp.55.3.181.
12. Reijnders JSAM, Ehrt U, Lousberg R, Aarsland D, Leentjens AFG. The association between motor subtypes and psychopathology in Parkinson's disease. Parkinsonism Relat Disord 2009;15:379–382. doi: 10.1016/j.parkreldis.2008.09.003.
13. Nasreddine ZS, Phillips NA, Bedirian V, Charbonneau S, Whitehead V, Collin I, et al. The Montreal Cognitive Assessment, MoCA: a brief screening tool for mild cognitive impairment. J Am Geriatr Soc 2005;53:695–699. doi: 10.1111/j.1532-5415.2005.53221.x.
14. Hamilton M. Development of a rating scale for primary depressive illness. Br J Soc Clin Psychol 1967;6:278–296. doi: 10.1111/j.2044-8260.1967.tb00530.x.
15. Clark DB, Donovan JE. Reliability and validity of the Hamilton Anxiety Rating Scale in an adolescent sample. J Am Acad Child Adolesc Psychiatry 1994;33:354–360. doi: 10.1097/00004583-199403000-00009.
16. Trenkwalder C, Kohnen R, Hogl B, Metta V, Sixel-Doring F, Frauscher B, et al. Parkinson's disease sleep scale – validation of the revised version PDSS-2. Mov Disord 2011;26:644–652. doi: 10.1002/mds.23476.
17. Jenkinson C, Peto V, Fitzpatrick R, Greenhall R, Hyman N. Self-reported functioning and well-being in patients with Parkinson's disease: comparison of the short-form health survey (SF-36) and the Parkinson's Disease Questionnaire (PDQ-39). Age Ageing 1995;24:505–509. doi: 10.1093/ageing/24.6.505.
18. Tarakad A, Jankovic J. Essential tremor and Parkinson's disease: exploring the relationship. Tremor Other Hyperkinet Mov (N Y) 2018;8:589. doi: 10.7916/D8MD0GVR.
19. Isaias IU, Canesi M, Benti R, Gerundini P, Cilia R, Pezzoli G, et al. Striatal dopamine transporter abnormalities in patients with essential tremor. Nucl Med Commun 2008;29:349–353. doi: 10.1097/MNM.0b013e3282f4d307.
20. Waln O, Wu Y, Perlman R, Wendt J, Van AK, Jankovic J. Dopamine transporter imaging in essential tremor with and without parkinsonian features. J Neural Transm (Vienna) 2015;122:1515–1521. doi: 10.1007/s00702-015-1419-z.
21. Louis ED, Clark L, Ottman R. Familial aggregation and co-aggregation of essential tremor and Parkinson's disease. Neuroepidemiology 2016;46:31–36. doi: 10.1159/000442021.

Cluster analysis; Essential tremor; Parkinson's disease; Subtypes

Supplemental Digital Content

Copyright © 2022 The Chinese Medical Association, produced by Wolters Kluwer, Inc. under the CC-BY-NC-ND license.