This is the author’s version of a work that was submitted/accepted for pub-lication in the following source:

Morrison, S., Cortes, N., Newell, K.M., Silburn, P.A., & Kerr, G.K. (2013)Variability, regularity and coupling measures distinguish PD tremor fromvoluntary 5Hz tremor. Neuroscience Letters, 534, pp. 69-74.

This file was downloaded from: http://eprints.qut.edu.au/55344/

c© Copyright 2012 Elsevier

This is the author’s version of a work that was accepted for publica-tion in Neuroscience Letters. Changes resulting from the publishing pro-cess, such as peer review, editing, corrections, structural formatting, andother quality control mechanisms may not be reflected in this document.Changes may have been made to this work since it was submitted for pub-lication. A definitive version was subsequently published in NeuroscienceLetters, [VOL 534, (2012)] DOI: 10.1016/j.neulet.2012.11.040

Notice: Changes introduced as a result of publishing processes such ascopy-editing and formatting may not be reflected in this document. For adefinitive version of this work, please refer to the published source:

http://dx.doi.org/10.1016/j.neulet.2012.11.040

Accepted Manuscript

Title: Variability, Regularity and Coupling MeasuresDistinguish PD Tremor from Voluntary 5 Hz Tremor

Authors: S. Morrison, N. Cortes, K.M. Newell, P.A. Silburn,G. Kerr

PII: S0304-3940(12)01509-1DOI: doi:10.1016/j.neulet.2012.11.040Reference: NSL 29416

To appear in: Neuroscience Letters

Received date: 30-7-2012Revised date: 22-10-2012Accepted date: 17-11-2012

Please cite this article as: S. Morrison, K.M. Newell, P.A. Silburn, G. Kerr, Variability,Regularity and Coupling Measures Distinguish PD Tremor from Voluntary 5<ce:hspsp=0.25/>Hz Tremor, Neuroscience Letters (2010), doi:10.1016/j.neulet.2012.11.040

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 proofbefore it is published in its final form. Please note that during the production processerrors may be discovered which could affect the content, and all legal disclaimers thatapply to the journal pertain.

Page 1 of 21

Accep

ted

Man

uscr

ipt

Young adults were able to produce a tremor similar in frequency to PD tremor

PD tremor was more variable and irregular than voluntarily generated 5 Hz tremor

Stronger coupling between the arms was seen for voluntary tremor responses

Variability measures can be used to discern between different tremor forms

*Highlights (for review)

Page 2 of 21

Accep

ted

Man

uscr

ipt

Title: Variability, Regularity and Coupling Measures Distinguish PD Tremor from

Voluntary 5 Hz Tremor

Authors: S. Morrison1, N, Cortes

2, K.M. Newell

5, P.A. Silburn

3, 4, 6, G. Kerr3,4

Affiliations 1School of Physical Therapy, Old Dominion University, VA, USA

2Sports Medicine Assessment, Research & Testing Laboratory, George Mason

University, Manassas, VA, USA

3School of Exercise and Nutrition Sciences, Queensland University of Technology, QLD,

Australia

4Movement Neuroscience Program, Queensland University of Technology, QLD,

Australia

5Department of Kinesiology, Pennsylvania State University, PA, USA.

6Center for Clinical Research, University of Queensland, QLD, Australia

Resubmitted: 22nd

October 2012

Corresponding Author: Steven Morrison

Address: School of Physical Therapy

Old Dominion University

VA 23529

Phone: 757-683-6757

FAX: 757-683-4410

Email: smorriso @odu.edu

1

Page 3 of 21

Accep

ted

Man

uscr

ipt

2

Abstract 1

A characteristic of Parkinson’s disease (PD) is the development of tremor within the 4-6 Hz 2

range. One method used to better understand pathological tremor is to compare the responses to 3

tremor-type actions generated intentionally in healthy adults. This study was designed to 4

investigate the similarities and differences between voluntarily generated 4-6 Hz tremor and PD 5

tremor in regards to their amplitude, frequency and coupling characteristics. Tremor responses 6

for 8 PD individuals (on-and off-medication) and 12 healthy adults were assessed under postural 7

and resting conditions. Results showed that the voluntary and PD tremor were essentially 8

identical with regards to the amplitude and peak frequency. However, differences between the 9

groups were found for the variability (SD of peak frequency, proportional power) and regularity 10

(Approximate Entropy, ApEn) of the tremor signal. Additionally, coherence analysis revealed 11

strong inter-limb coupling during voluntary conditions while no bilateral coupling was seen for 12

the PD persons. Overall, healthy participants were able to produce a 5 Hz tremulous motion 13

indistinguishable to that of PD patients in terms of peak frequency and amplitude. However, 14

differences in the structure of variability and level of inter-limb coupling were found for the 15

tremor responses of the PD and healthy adults. These differences were preserved irrespective of 16

the medication state of the PD persons. The results illustrate the importance of assessing the 17

pattern of signal structure/variability to discriminate between different tremor forms, especially 18

where no differences emerge in standard measures of mean amplitude as traditionally defined. 19

20

Key Words: Tremor, Parkinson’s disease, Variability, Complexity, Bilateral Coupling 21

22

Page 4 of 21

Accep

ted

Man

uscr

ipt

3

Introduction 1

A general feature of pathological tremor forms is that they are characterized by increased 2

amplitude and a lower frequency in comparison to physiological tremor [1-3]. Typically, the 3

limb oscillations associated with neurological disorders like Parkinson’s disease (PD) and 4

essential tremor are restricted to a single dominant peak located between 4-6 Hz [2, 4]. In 5

contrast, tremulous motion in healthy adults is more broadband, with peaks commonly being 6

observed between 2-4 Hz, 8-12 Hz and, for finger tremor only, as high as 18-24 Hz [3, 5]. These 7

pronounced differences in frequency and amplitude facilitate the discrimination between 8

pathological and normal physiological tremor signals. However, how to accurately discriminate 9

between 4-6 Hz tremors arising from different neurological sources (e.g. PD, essential tremor) 10

from tremor generated voluntarily and/or related to psychogenic factors is a question of 11

continuing interest [6-8]. 12

One method used to assess the properties of tremor has been to compare the responses of 13

tremor-type actions generated voluntarily to those of pathologies [8-11]. While voluntarily 14

producing oscillatory motion within the 8-12 Hz range of physiological tremor is not achievable 15

for healthy individuals [12], producing rhythmical limb motion at or around the frequency of 16

many forms of pathological tremor (e.g. 4-6 Hz) is within the functional boundaries of the 17

voluntary movement [9-11]. Previous studies have reported that tremor generated voluntarily by 18

healthy adults under these conditions exhibits similar frequency characteristics, and is often 19

indistinguishable in regards to amplitude, from that observed in individuals with neurologic 20

disorders [8-11]. 21

Given the similarities in appearance between these tremor forms, there is an obvious interest 22

in how to discriminate between the different oscillatory outputs. One means by which to assess 23

Page 5 of 21

Accep

ted

Man

uscr

ipt

4

these differences is to identify the pattern of regularity and variability in the respective tremor 1

signals. Previous research has highlighted that, in addition to amplitude and frequency 2

differences, the tremor under pathological conditions is often less complex and variable than 3

physiological tremor assessed from healthy adults [13-15]. Another means by which to assess 4

tremor differences is to evaluate the level of coupling of tremor between limbs. Previous studies 5

have reported no coupling of tremor between the arms in healthy adults of differing ages[16-18] 6

or for persons with PD, irrespective of their medication state [19, 20]. However, increased 7

coupling between the limbs can be seen for other movement forms in healthy adults, persons 8

with PD[21, 22] and for patients with essential tremor [23]. One suggestion is that inter-limb 9

independence is an intrinsic feature of tremor production of healthy individuals and so 10

assessments of the degree of coupling can be used to distinguish between different tremor forms. 11

The aim of this study was to investigate the similarities and differences between voluntarily 12

generated 4-6 Hz tremulous movement and PD tremor in regards to their amplitude, frequency 13

and coupling characteristics. It was predicted that the voluntary tremor will exhibit similar 14

frequency and amplitude characteristics to that of the PD patients. Further, for the PD 15

individuals, these similarities will be preserved irrespective of their medication state [3, 24]. It 16

was also predicted that there will be notable differences in tremor regularity, variability and 17

coupling which will distinguish between the different individuals. 18

19

Methods 20

Subjects 21

Twelve young control subjects (six males, six females, age 24±5.1 yrs), and eight 22

Parkinsonian patients (two males, six females, 65.1±3.2 yrs) gave informed consent to participate 23

Page 6 of 21

Accep

ted

Man

uscr

ipt

5

in this study. All control participants were right-hand dominant, physically active, had normal 1

or corrected-to-normal vision, and reported no known neurological/cognitive disorders, or 2

history of neuromuscular injury that could influence performance. All Parkinson patients were 3

assessed using the UPDRS-motor section by a trained neurologist and all exhibited resting and 4

postural tremors bilaterally. Additional subject inclusion details and results have been reported 5

previously.[3] All participants provided written informed consent and completed a medical 6

history questionnaire prior to testing to determine health status. All experimental procedures 7

complied with University IRB guidelines and were in accordance with the Declaration of 8

Helsinki. 9

10

Experimental Design 11

For all participants, bilateral tremor responses were recorded under both resting and postural 12

conditions. For the PD patients, resting tremor was measured independently from both arms 13

which were allowed to hang relaxed and unsupported by their side [3, 25]. For the postural 14

tremor condition, participants performed a pointing task with both arms held parallel to the 15

ground [3]. For the upper limb, the shoulders were flexed 90° in the sagittal plane, elbows fully 16

extended, and forearms pronated. The index finger of each arm was extended at the 17

metacarpophalangeal joint with the thumb adducted and remaining fingers flexed to form a loose 18

fist. 19

For the simulated postural and resting tremor movements, healthy individuals were asked to 20

adopt the same postural or resting positions as described above. They were then instructed to 21

perform rapid alternating wrist flexion-extension movements with both arms. Prior to data 22

collection, subjects were given 4 trials of practice to generate the required frequency of tremor 23

Page 7 of 21

Accep

ted

Man

uscr

ipt

6

motion in each position. An auditory metronome set at 5 Hz was provided during this period to 1

allow each person to ensure they were aware of the required movement frequency. 2

3

Equipment 4

All testing procedures were performed while individuals adopted a standing position. Hand and 5

finger tremor were measured using four uniaxial Coulbourn accelerometers (V94-41) and 6

amplified through a Coulbourn transducer coupler (V75-25A, sample rate 100 Hz). The 7

accelerometers were attached to the hand (middle of third metacarpal) and index finger (dorsal 8

distal aspect) of each arm so the measurement axis was perpendicular to the ground during 9

pointing. Six 30 s trials were performed within each condition (total number of trials: 12). Rests 10

were provided between trials/conditions to reduce the effects of fatigue. 11

12

Data Analysis 13

The accelerometer data were filtered by a second-order Butterworth low-pass digital filter 14

(cutoff frequency 50 Hz). All data analysis was performed using custom software developed in 15

Matlab version 7.0 (Mathworks R14). 16

Time Series Analysis: Average tremor amplitude involved calculating the root mean square 17

(RMS) of the tremor signal (window size 100 ms). 18

Frequency Analysis: This was performed on the filtered tremor data between 0-40 Hz using 19

Welch's averaged, modified periodogram method (512 data point Hanning window). The 20

dependent measures calculated were; maximum amplitude of each signal (peak power), the 21

frequency at which the peak power was observed (peak power frequency, PPF), the variability of 22

PPF (SD PPF) and the distribution of power across the tremor signal (proportional power). The 23

Page 8 of 21

Accep

ted

Man

uscr

ipt

7

proportional power variable provides an estimate of the relative proportion of power (expressed 1

as a percentage of total power) that occurs within successive 1 Hz frequency bins. All dependent 2

measures were calculated for all trials per condition and across all subjects. 3

Signal Regularity: The degree of regularity of the accelerometer signals was assessed using 4

Approximate Entropy (ApEn) analysis. ApEn measures the conditional probability of the signal 5

by providing a measure of the (logarithmic) likelihood that runs of patterns that are close for m 6

observations remain close on the next incremental comparisons (m+1). This analysis returns a 7

single value for the tremor signal within the range of 0-2. Typically, a signal with higher ApEn 8

values is described as being less regular (or more complex). Conversely, a signal with a lower 9

ApEn would be described as more regular (less complex) [26]. 10

Coupling Analysis: Estimation of the degree of coupling of tremor between arms was 11

determined by applying cross correlation (Pearson product moment) and coherence analyses. 12

The correlation analysis was conducted over multiple time lags (range 5 s). The peak 13

correlation coefficient and the lag at which this peak was observed were recorded. Coherence 14

analysis was performed to assess the degree of coupling in the frequency domain and the 15

maximum (peak) coherence value determined. This analysis was performed within the range 0-16

20 Hz (window size 512 data points, binwidth 0.1953 Hz). To assess whether the coherence 17

values were significantly different from zero, a 95% confidence interval was calculated [27]. 18

19

Statistical Analysis 20

A repeated measures mixed generalized linear model (GLM) was used to assess for 21

differences in the dependent measures. A two-way model was used to identify differences in the 22

tremor measures between groups (young, PDON/OFF medication) and across tremor conditions 23

Page 9 of 21

Accep

ted

Man

uscr

ipt

8

(resting, postural). Where significant effects were reported, post hoc evaluations were performed 1

using Tukey’s Honestly Significant Difference (HSD) test. All statistical analyses were 2

performed using SAS statistical software (SAS Institute Inc., NC), with the risk of Type I error 3

set at p <0.05. 4

5

Results 6

Tremor Similarities 7

Tremor RMS Amplitude: No significant differences in mean RMS amplitude were observed 8

for the tremor responses between the PD patients and the healthy adults simulating a 5 Hz 9

tremor. No significant differences were found, for the PD group, between the on/off medication 10

state. 11

Tremor Frequency: The resting and postural tremor forms for all groups were characterised 12

by a single dominant frequency peak, present between 4-6 Hz. For the simulated group, the 13

resting tremor peak was 5.16 Hz and, under postural conditions, 4.98 Hz. For the PD patients, 14

the resting tremor peaks were at 5.26 Hz (on-medication) and 5.34 Hz (off-medication). The 15

postural tremor peaks for these same individuals were at 5.03 Hz (on-medication) and 5.18 Hz 16

(off-medication). No significant difference was observed between groups or conditions (resting, 17

postural) in regards to the frequency of peak power (PPF). Furthermore, no differences in peak 18

power or PPF were observed between the two simulated conditions for the healthy adults. Figure 19

1 illustrates the typical tremor responses, power spectral and proportional power profiles for a 20

single PDOFF individual and a healthy adult producing a 5 Hz tremor. The tremor responses 21

across the postural and resting conditions are shown in this figure. Figure 2 depicts the major 22

Page 10 of 21

Accep

ted

Man

uscr

ipt

9

similarities in mean tremor RMS amplitude and PPF between the different groups and 1

conditions. 2

-------------------------------------------- 3

Insert Figures 1 & 2 4

-------------------------------------------- 5

Differences in Tremor 6

Tremor Frequency: A significant group by condition interaction effect was observed for the 7

SD of PPF (F2,26=14.33; p<0.001) and proportional power measures (F2,26=22.67; p<0.001). For 8

the SD PPF results, post hoc analysis revealed that the tremor for the PDON/OFF patients was more 9

variable (higher SD) than for the simulated conditions (all p’s<0.001, see figure 3). This effect 10

was found across the resting and postural conditions. For the proportional power analysis, a 11

greater percentage of the total power (>60%) was observed within the 5-6 Hz range under 12

simulated conditions (all p’s<0.001, see figure 3). This effect was found across both conditions. 13

Tremor Regularity: A significant group by condition effect was observed for the ApEn 14

values for finger tremor (F2,26=26.25; p<0.001). Post hoc analysis revealed that the tremor for 15

the PDON/OFF patients, under both resting and postural conditions, was more complex (higher 16

ApEn) than the tremor produced under simulated conditions (all p’s<0.001, see figure 3). No 17

differences in ApEn scores were observed within each group between the resting and postural 18

tremor conditions. 19

-------------------------------------------- 20

Figure 3 21

------------------------------------------- 22

23

Coupling Relations 24

Page 11 of 21

Accep

ted

Man

uscr

ipt

10

For the correlation analysis, there was a significant group-by-condition effect for the strength 1

of the coupling relation between arms (F2,26=7.88; p<0.001). Post hoc analysis revealed that the 2

strength of the bilateral coupling was significantly greater under simulated conditions than for 3

the PDON/OFF individuals (all r values < 0.12; p <0.05; see figure 4). There was no evidence of 4

coupling of tremor between the arms for the PD individuals irrespective of the tremor form 5

assessed or their medication state (on/off). All peak correlation values were found at 0 s lag. 6

The coherence analysis provided a similar pattern of results with a significant group-by-7

condition effect being found (F2,26=23.15; p<0.001). Subsequent analysis highlighted significant 8

differences between the simulated and PDON/OFF tremor (all p’s<0.05). For the PDON/OFF tremor, 9

there was no evidence of any coupling of tremor between limbs or any differences as a function 10

of medication state. However, a significantly higher level of coupling was observed under 11

simulated conditions (figure 4). All significant coherence peaks were between 4-6 Hz. 12

-------------------------------------------- 13

Figure 4 14

------------------------------------------- 15

16

Discussion 17

This study was designed to examine the similarities and differences between Parkinsonian 18

tremor and that generated voluntary at 5 Hz by healthy adults. There were no differences in 19

modal frequency or overall amplitude of the tremor between the healthy participants and the 20

PDON/OFF persons. However, there were significant differences in the structure of variability of 21

the tremor responses between groups. The simulated tremor was more regular (lower ApEn), 22

less variable in frequency (lower SD of PPF, more power within a narrower frequency range) 23

Page 12 of 21

Accep

ted

Man

uscr

ipt

11

and was more strongly coupled between limbs than the tremor recorded from PD individuals. 1

These differences were preserved irrespective of the medication state of the PD persons. 2

3

Similarities between Voluntary and PD Tremor 4

One of the major distinguishing neurological features of Parkinson disease is the presence of 5

enhanced tremor under resting and/or postural conditions [28]. These tremulous oscillations are 6

typically observed between 4-6 Hz and reflect changes in the underlying neural mechanisms 7

driving tremorgenesis. Since PD tremor differs dramatically in both frequency and amplitude 8

from the physiological tremor observed in healthy adults, various studies have compared the 9

responses of tremor-type actions generated voluntarily to that of pathologies [8-11]. Such 10

comparisons are employed to provide additional information about the dynamics of pathological 11

tremor. For example, as the voluntary tremulous action can be generated within a similar 12

frequency range, comparisons can be made between two comparable oscillatory outputs. A 13

central finding of the current study is that, in terms of terms of peak frequency and amplitude, 14

there was no difference in the tremor for healthy adults and individuals with Parkinson’s disease. 15

The tremor response of all individuals was characterized by a single dominant frequency peak 16

between 4-6 Hz, irrespective of whether the tremor was recorded under postural or resting 17

conditions. 18

An additional finding of note was the lack of tremor differences for the PD patients between 19

their on and off medication state. Generally, the use of anti-parkinsonian medication produces a 20

decrease in the amplitude of resting/postural tremor with little concurrent change in tremor 21

frequency or inter-limb coupling [20, 24, 29]. However, the extent of any amplitude decline can 22

Page 13 of 21

Accep

ted

Man

uscr

ipt

12

be highly variable between PD individuals [3, 19, 28], an observation which may contribute to 1

the lack of any tremor differences between the on-off medication state in the current study. 2

3

Differences between Voluntary and PD Tremor 4

While there were no significant differences between the PDON/OFF and voluntary 5 Hz 5

tremulous movements with regards to peak frequency and mean RMS amplitude, fundamental 6

differences between the tremor in the two groups were observed for the pattern of regularity of 7

the variability. Of particular note, the voluntary generated tremor was more regular (lower 8

ApEn) and more tightly restricted around 5 Hz (lower SD of PPF, greater proportion of power 9

within 5-6 Hz) than the tremor for the PDON/OFF group, irrespective of their medication state. 10

This finding provides an interesting addition to reports which have focused on comparing PD 11

tremor to physiological tremor [13-15]. In these studies, PD tremor is typically less variable and 12

more regular than the tremor for healthy adults of a similar age. Taken together, these results 13

illustrate that even though the neuromotor driving signal for individuals with PD is constrained 14

to a very narrow frequency range, their tremor response still exhibits a moderate degree of 15

variability. Indeed, the PD tremor exhibited a higher degree of variability and was more 16

irregular (more complex) than the voluntary tremor responses. 17

One additional difference in tremor between the groups relates to the degree of coupling 18

between limbs. For the PD individuals, the results of the specific analyses (i.e., cross correlation, 19

coherence) showed no evidence of any inter-limb coupling. In contrast, under simulated 20

(voluntary) tremor conditions, there was a significantly high level of coupling between the arms. 21

For the coherence results, this strong bilateral relation was within 4-6 Hz, which coincided with 22

the frequency range at which the simulated tremor was performed. The bilateral differences in 23

Page 14 of 21

Accep

ted

Man

uscr

ipt

13

coupling strength probably reflect underlying variation in the neural mechanisms driving each 1

oscillatory output. In Parkinson’s disease (and assessments of tremor in healthy adults), the 2

prevailing view is the tremor within each limb is generated independently from uncoupled or 3

parallel oscillators within the CNS [3, 20, 30]. In contrast, during rhythmical actions generated 4

voluntarily, the level of coupling between contralateral limb segments tends to be strong [21, 5

31]. One hypothesis is that this coupling reflects the act of voluntarily increasing the neural drive 6

to selected muscle groups [31]. 7

Overall, our findings reveal that tremor in PD patients was more irregular, variable and 8

characterized by a lack of bilateral coupling compared to voluntary 5 Hz tremulous motion 9

performed by healthy adults. The results reveal the sensitivity of measures of signal 10

structure/variability to discriminate between groups and conditions in tremor analysis, beyond 11

that provided by standard measures of (mean) tremor amplitude. Further, comparisons between 12

healthy adults and patients with neurological disorders could incorporate assessments of 13

voluntary oscillatory 4-6 Hz movement in order to distinguish between various tremor forms. 14

Page 15 of 21

Accep

ted

Man

uscr

ipt

14

References

1. Elble, R.J., Origins of tremor. The Lancet, 2000. 355(9210): p. 1113-1114. 1

2. McAuley, J.H. and C.D. Marsden, Physiological and pathalogical tremors and rhythmic 2

central motor control. Brain, 2000. 123: p. 1545-1567. 3

3. Kerr, G., S. Morrison, and P. Silburn, Coupling between limb tremor and postural sway in 4

Parkinson's disease. Movement Disorders, 2008. 23(3): p. 386-394. 5

4. Deuschl, G., P. Bain, and M. Brin, Consensus statement of the Movement Disorder Society 6

on Tremor. Ad Hoc Scientific Committee. Movement Disorders, 1998. 13(Suppl 3): p. 2-23. 7

5. Elble, R.J., Characteristics of physiologic tremor in young and elderly adults. Clinical 8

Neurophysiology, 2003. 114(4): p. 624-635. 9

6. Lang, A.E. and V. Voon, Psychogenic movement disorders: Past developments, current 10

status, and future directions. Movement Disorders, 2011. 26(6): p. 1175-1186. 11

7. McAuley, J. and J. Rothwell, Identification of psychogenic, dystonic, and other organic 12

tremors by a coherence entrainment test. Movement Disorders, 2004. 19(3): p. 253-267. 13

8. Voon, V, et al., The involuntary nature of conversion disorder. Neurology, 2011. 74: p223-28 14

9. Pollok, B., et al., The cerebral oscillatory network of voluntary tremor. The Journal of 15

Physiology, 2004. 554(3): p. 871-878. 16

10. Burkhard, P., J. Langston, and J. Tetrud, Voluntarily simulated tremor in normal subjects. 17

Clinical Neurophysiology, 2002. 32(2): p. 119-26. 18

11. Kumru, H., et al., Transient arrest of psychogenic tremor induced by contralateral ballistic 19

movements. Neuroscience Letters, 2004. 370(2-3): p. 135-139. 20

12. Morrison, S., L. Hong, and K.M. Newell, Upper frequency limits of bilateral coordination 21

patterns. Neurosci Lett, 2009. 454: p. 233-238. 22

13. Deuschl, G., M. Lauk, and J. Timmer, Tremor classification and tremor time series analysis. 23

Chaos, 1995. 5: p. 48-51. 24

14. Morrison, S., et al., Differential time- and frequency-dependent structure of postural sway 25

and finger tremor in Parkinson’s disease. Neuroscience Letters, 2008. 443: p. 123-128. 26

15. Vaillancourt, D.E. and K.M. Newell, The dynamics of resting and postural tremor in 27

Parkinson's disease. Clinical Neurophysiology, 2000. 111: p. 2046-2056. 28

16. Daneault, J.-F., B. Carignan, and C. Duval, Bilateral effect of a unilateral voluntary 29

modulation of physiological tremor. Clinical Neurophysiology, 2010. 121: p. 734-743. 30

Page 16 of 21

Accep

ted

Man

uscr

ipt

15

17. Hwang, I.-S., et al., Postural fluctuations during pointing from a unilateral or bilateral stance. 1

Human Movement Science, 2006. 25(2): p. 275-291. 2

18. Marsden, C., et al., The relation between physiological tremor of the two hands in healthy 3

subjects. Electroencephalogr Clin Neurophys, 1969. 27: p. 179-185. 4

19. Ben-Pazi, H., et al., Synchrony of rest tremor in multiple limbs in parkinson's disease: 5

evidence for multiple oscillators. Journal Of Neural Transmission. 2001. 108(3): p. 287-296. 6

20. Morrison, S., G. Kerr, and P. Silburn, Bilateral tremor relations in Parkinson's disease: 7

Effects of mechanical coupling and medication. Parkinsonism & Related Disorders, 2008. 8

14: p. 298-308. 9

21. Byblow, W.D., et al., Bimanual coordination in Parkinson's disease: Deficits in movement 10

frequency, amplitude, and pattern switching. Movement Disorders, 2002. 17(1): p. 20-29. 11

22. Johnson, K.A., et al., Bimanual co-ordination in Parkinson's disease. Brain, 1998. 121(4): p. 12

743-753. 13

23. Lauk, M., et al., Side-to-side correlation of muscle activity in physiological and pathological 14

human tremors. Clinical Neurophysiology, 1999. 110(10): p. 1774-1783. 15

24. Henderson, J.M., et al., Postural Tremor of Parkinsons Disease. Clinical 16

Neuropharmacology, 1994. 17(3): p. 277-285. 17

25. Raethjen, J., et al., Determinants of physiologic tremor in a large normal population. Clinical 18

Neurophysiology, 2000. 111: p. 1825-1837. 19

26. Pincus, S.M., Approximate entropy as a measure of system complexity. Proceedings of the 20

National Academy of Sciences of the United States of America, 1991. 88: p. 2297-2301. 21

27. Halliday, A.M., et al., A framework for the analysis of mixed time series/point process data - 22

theory and application to the study of physiological tremor, single motor unit discharges and 23

electromyograms. Progress in Biophysics and Molecular Biology, 1995. 64: p. 237-278. 24

28. Elble, R.J. and W.C. Koller, Tremor.1990, Baltimore: Johns Hopkins. 25

29. Duval, C., Rest and postural tremors in patients with Parkinson's disease. Brain Research 26

Bulletin, 2006. 70(1): p. 44-48. 27

30. Hurtado, J.M., et al., Inter- and intralimb oscillator coupling in parkinsonian tremor. 28

Movement Disorders, 2000. 15(4): p. 683-691. 29

31. Swinnen, S.P., et al., The effect of movement speed on upper-limb coupling strength. Human 30

Movement Science, 1992. 11(5): p. 615-636. 31

32

Page 17 of 21

Accep

ted

Man

uscr

ipt

16

Figure Captions 1

2

Figure 1 Representative examples of the tremor signal for a single PDOFF participant and a 3

healthy subject voluntarily generating a 5 Hz tremor. The respective power spectral 4

and proportional power profiles for each tremor signal are also shown. Tremor 5

responses were recorded under postural and resting conditions. 6

7

Figure 2 Similarities in tremor amplitude (mean RMS) and modal frequency (peak power 8

frequency, PPF) between the PDON/OFF and healthy subjects. Overall group mean 9

results for the resting and postural conditions are shown. Error bars shown in each 10

figure represent one SE of the mean. 11

12

Figure 3 Differences in tremor regularity (mean ApEn) and variation (SD of PPF) between the 13

PDON/OFF and healthy subjects. Results are shown for both the resting and postural 14

conditions. Error bars represent one SE of the mean. Significant group differences 15

between the PD and the voluntary tremor responses are denoted with an asterisk (*). 16

17

Figure 4 Significant differences in level of inter-limb coupling between groups based upon 18

changes in mean peak correlation and coherence values. Error bars represent one SE 19

of the mean. Significant differences in the level of coupling between the PDON/OFF 20

and healthy groups are denoted with an asterisk (*). 21

Page 18 of 21

Accep

ted

Man

uscr

ipt

Frequency (Hz)

0 2 4 6 8 10

Tre

mor

Pow

er (

m.s

-2)2

0

2000

4000

6000

8000

10000

12000

PDOFF Tremor (postural)

-40

-20

0

20

40

Time (s)

0 1 2 3 4 5 6 7 8 9 10

Tre

mor

Am

pli

tude

(m.s

-2)

-40

-20

0

20

40PD

OFF Tremor (resting)

0

2000

4000

6000

8000

10000

12000

0

2000

4000

6000

8000

10000

12000

0

2000

4000

6000

8000

10000

12000

-40

-20

0

20

40

-40

-20

0

20

40

Simulated Tremor (resting)

Simulated Tremor (postural)

Pro

port

iona

l P

ow

er (

%)

0

10

20

30

40

50

60

0

10

20

30

40

50

60

Frequency (Hz)0

-1

1-2

2-3

3-4

4-5

5-6

6-7

7-8

9-1

0

10

-11

11

-12

0

10

20

30

40

50

60

0

10

20

30

40

50

60

Figure 1

Page 19 of 21

Accep

ted

Man

uscr

ipt

Simulated PDon PDoff

Mea

n R

MS

Tre

mo

r A

mp

litu

de

(m.s

-2)

0

1

2

3

4

5

6

Simulated PDon PDoff

0

1

2

3

4

5

6Postural Conditions

Simulated PDon PDoff

1

2

3

4

5

6

7 Right Finger

Left Finger

Resting Conditions

Simulated PDon PDoff

Fre

qu

ency

of

Pea

k P

ow

er (

Hz)

1

2

3

4

5

6

7

Figure 2

Page 20 of 21

Accep

ted

Man

uscr

ipt

Postural Conditions

Voluntary PDon PDoff

Mea

n A

pE

n

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

Right Finger

Left Finger

Resting Conditions

Voluntary PDon PDoff

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

SD

of

PP

F (

Hz)

0

1

2

3

4

0

1

2

3

4

*

*

*

*

*

*

*

*

Figure 3

Page 21 of 21

Accep

ted

Man

uscr

ipt

Voluntary PDon PDoff

Mea

n P

eak

Co

her

ence

0.0

0.2

0.4

0.6

0.8

1.0 Resting Condition

Postural Condition

Voluntary PDon PDoff

Mea

n P

eak

Co

rrel

atio

n

0.0

0.2

0.4

0.6

0.8

1.0

*

*

Figure 4