- Email: [email protected]
The effect of contraction intensity on force fluctuations and motor unit entrainment in individuals with essential tremor
Clinical Neurophysiology 121 (2010) 233–239
Contents lists available at ScienceDirect
Clinical Neurophysiology journal homepage: www.elsevier.com/locate/clinph
The effect of contraction intensity on force fluctuations and motor unit entrainment in individuals with essential tremor M.E. Héroux a,*, G. Pari b,c, K.E. Norman a,d a
School of Human Kinetic, University of British Columbia, 210 - 6081 University Boulevard. BC, Canada V6T 1Z1 Movement Disorders Clinic, Kingston General Hospital, ON, Canada K7L 2V7 c School of Medicine, Queen’s University, 68 Barrie Street, ON, Canada K7L 3N6 d Centre for Neuroscience Studies, Queen’s University, Botterell Hall, ON, Canada K7L 3N6 b
a r t i c l e
i n f o
Article history: Accepted 28 October 2009 Available online 31 December 2009 Keywords: Force fluctuations Essential tremor Electromyography Power spectral analysis
a b s t r a c t Objectives: Quantify the effect of increasing contraction intensity on the amplitude of force fluctuations and neuromuscular and force tremor spectral power. Methods: Twenty-one subjects with essential tremor (ET) and 22 healthy controls applied isometric wrist extension contractions. Various sub-maximal contraction intensities were evaluated (5%-, 10%-, 20%- and 30%-MVC). Force fluctuations and wrist extensor neuromuscular activity were recorded using a load cell and electromyography (EMG). Results: Higher contraction intensities were associated with larger amplitude force fluctuations and greater neuromuscular activation. However, spectral power associated with tremor peaks remained relatively constant (EMG) or decreased (force) with increasing contraction intensity. Conclusions: Motor unit entrainment associated with centrally generated oscillatory inputs does not increase with greater levels of muscle activation. Significance: Rather than influencing a constant proportion of active motor units, abnormal oscillatory drive influences a relative constant number of total motor units. When combined with the findings from our previous study on postural tremor, the present results provide preliminary evidence that abnormal stretch reflex activity may contribute to this motor unit entrainment. Ó 2009 International Federation of Clinical Neurophysiology. Published by Elsevier Ireland Ltd. All rights reserved.
1. Introduction Recently we have shown that rhythmical motor unit entrainment does not significantly increase when individuals with ET support light to moderate loads despite an increase in overall neuromuscular activity (Héroux et al., 2009). With postural tremor, however, abnormal descending oscillatory activity may interact (synergistically or competitively) with short- and long-latency stretch reflexes caused by oscillatory motion (Elble et al., 1987; Matthews, 1993; Elble, 1996). Thus, we sought to investigate the influence of contraction intensity on the strength of motor unit entrainment in subjects with ET during closed-kinetic chain isometric contractions since the influence of stretch reflexes during such contractions is minimized (Burne et al., 1984; Doemges and Rack, 1992). Previously, Gillies (1994) noted an increase in force fluctuation amplitude with increasing contraction intensity in a report involv-
* Corresponding author. Address: School of Human Kinetics, University of British Columbia, 210-6081 University Blvd., Vancouver, BC, Canada V6T 1Z1. Tel.: +1 604 827 3372; fax: +1 604 822 6842. E-mail address: [email protected] (M.E. Héroux).
ing an unspecified number of ET subjects. Based on visual inspection, the author states that the tremor spectral peak represented a constant proportion of spectral power across a wide range of contraction intensities (5–95% maximal voluntary contraction (MVC)) and concludes that the central oscillator in ET affects a constant proportion of the central drive to the motor neuron pool. More recently, Bilodeau et al. (2000) reported an increase in force fluctuation amplitude with increasing contraction intensity in subjects with ET. Of particular interest was the anecdotal report that tremor spectral peak amplitude appeared to increase at the higher contraction intensity (2.5%- versus 20%-MVC), indicating that force tremor scaled with contraction intensity. Corroborating these findings, Burne et al. (2004) noted a linear increase in electromyography (EMG) tremor spectral peak amplitude when subjects with ET generated various low level isometric contractions (<5%-MVC). It remains unclear, however, whether the increase in tremor spectral power – both in force and EMG signals – increases proportionally to the overall level of muscle activation or whether, similar to what was previously reported for postural tremor, motor unit entrainment is proportionally greater at lower contraction intensities.
1388-2457/$36.00 Ó 2009 International Federation of Clinical Neurophysiology. Published by Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.clinph.2009.10.015
234
M.E. Héroux et al. / Clinical Neurophysiology 121 (2010) 233–239
The purpose of this study was to quantify the effect of increasing contraction intensity on the amplitude of neuromuscular and force tremor spectral power in order to determine whether or not the central oscillatory drive in ET affects a constant proportion of the motor neuron pool as suggested by Gillies (1994). In contrast to our previous study, closed-kinetic chain isometric contractions were used to minimize the influence of stretch reflex to the ongoing neuromuscular activity. 2. Methods 2.1. Subjects Twenty-one subjects with ET and 22 healthy control subjects of similar age participated. Subjects with ET were divided into two groups based on the strength and consistency of motor unit entrainment during isometric force production (see Section 2.5.2 for details; Héroux et al., 2009); two subjects with mild tremor from the previous study had highly variable and often absent motor unit entrainment at all target intensities and their data were not included. Subjects in the ET groups: (1) met the diagnosis criteria of the Consensus Statement of the Movement Disorder Society on Tremor (Deuschl et al., 1998), (2) did not present with factors associated with misdiagnosed ET or other red flag items (Elble, 2000; Jain et al., 2006) and (3) had spectral plots of hand postural tremor and forearm EMG recorded under various loading conditions that were consistent with ET (see Héroux et al. (2009) for details). All subjects provided informed consent to the protocol, which was approved by the local Research Ethics Board. 2.2. Apparatus The subject was seated in front of a computer monitor. The forearm of the side being tested was in full pronation and supported on a height-adjustable table with the wrist in neutral flexion/extension and neutral radial/ulnar deviation. Two straps secured the forearm to the table and a piece of opaque fabric prevented the subject from seeing their forearm and hand during testing. Voluntary isometric wrist extension contractions were made against a support device connected to a load cell (MLP100, Transducer Techniques) secured to the table. During MVC testing, the load cell signal was conditioned (Dataforth DSCA43–10 strain gage input conditioner), filtered using an eighth order analogue Butterworth filter (Max740, Maxim) and sampled at 1024 Hz by a 16-bit analog-to-digital converter (National Instrument PCI-MIO-16XE-10) controlled by a Pentium Xeon 2.66 GHz personal computer. During %-MVC trials, the load cell signal was also amplified with an amplifier gain of 8 (Burster Model 9243) prior to being sampled. For MVC testing the smallest detectable change in force was 0.231 N (range: 0–448 N), whereas it was 0.028 N (range: 0–56 N) for %-MVC testing. The computer monitor (21 in. Dell M992, 60 Hz refresh rate, set to 1024 768 pixels) was positioned 50 cm from the subject; the top of the viewing area was in line with the eyes. During %-MVC testing the subject viewed a vertically centered horizontal line that spanned the width of the viewing area and corresponded to the target force being assessed. A second line corresponding to the force exerted on the load cell scrolled left to right during the trial; the total width of the viewing area corresponded to 12 s. The vertical axis was scaled to ±6 N of the target force for all %-MVC testing resulting in a visual gain of 50 pixels/N. The visual angle of feedback, which is dependent on eye–monitor distance and visual gain, has been shown to influence the amplitude and structure of force fluctuations (Vaillancourt et al., 2006). Pilot testing on healthy subjects revealed that the steadiest trial (target intensity
5%-MVC) resulted in a visual angle of 1.15°, which ensured that the potential influence of visual angle on force fluctuations was negligible (see Vaillancourt et al. (2006) for details). Surface EMG was used to measure the neuromuscular activity of the extensor carpi radialis brevis. Skin preparation, electrode placement and the EMG system used were the same as previously described (see Héroux et al. (2009) for details). The EMG signal was digitized using the previously mentioned 16-bit analog-to-digital converter; all EMG data were sampled at 1024 Hz. Load cell and EMG data were collected using custom-built Labview 8 software (National Instruments) and saved to the hard disk of the personal computer for subsequent analysis. 2.3. Procedure In subjects with ET, the hand reported to have the most severe tremor was measured. In the case of symmetrical tremor, the dominant hand was tested. The side tested in healthy controls was selected in order to obtain approximately the same dominant/nondominant ratio between ET and control groups. The MVC of the wrist extensors was first determined. Following a brief warm-up consisting of 3–4 sub-maximal contractions, the subject performed three successive maximal static contractions lasting 5 s with a 120 s rest period between contractions. Next, the subject was asked to produce and hold a series of constant force contractions at four sub-maximal intensities (5%, 10%, 20% and 30% of MVC). The subject was instructed to extend the wrist against the support device and superimpose the force line overtop of the target line on the monitor. After the initial rise in force and stabilization period, the subject held this level of force production with visual feedback for 8 s. Two blocked trials were recorded at each target intensity; the order of testing of the four target intensities was selected randomly. There was a 60 s rest period between trials of a given block and a minimum of 120 s rest between each block. 2.4. Data processing Prior to performing all data analyses, load cell and EMG data were conditioned by the following methods. The EMG data were digitally rectified and the offset value was removed. Next, the EMG and load cell data were digitally filtered using a dual-pass fourth order Butterworth filter with a low-pass cut-off frequency of 40 Hz. Furthermore, the 8 s window of force data from each trial was linearly detrended. All data processing and subsequent time and frequency analyses were performed using software written in Matlab 7 (The MathWorks, Inc.). 2.4.1. Amplitude of force fluctuations and EMG activity The maximum force value produced across all three trials was selected as the MVC value. Overall force fluctuation amplitude during %-MVC trials was determined by calculating the standard deviation of the force time series. This is a global measure that does not distinguish between force fluctuations associated with pathological tremor (4–12 Hz) from those related to visuomotor corrections (0–3 Hz) and normal physiological force variability (8–12 Hz) (Elble, 1996; Jones et al., 2002; Slifkin et al., 2000). For the EMG signal, the mean level of neuromuscular activity for each trial was computed. The mean of both trials from a given target force was computed for all measures and used for statistical analysis. 2.4.2. Spectral measures of isometric force tremor and neuromuscular activity Force and EMG auto-spectra were calculated for each target intensity and visual condition using the method of disjoint sections (Halliday et al., 1995). The spectra were estimated by averaging the
235
M.E. Héroux et al. / Clinical Neurophysiology 121 (2010) 233–239
finite Fourier transforms calculated for non-overlapping windows of 4096 points, which resulted in a frequency resolution of 0.25 Hz. As was the case in our previous study, visual inspection of individual EMG auto-spectra revealed reduced amplitude in, or total absence of, a prominent peak indicative of pathological tremor at one or sometimes two of the higher target intensities (i.e., 20%- or 30%-MVC) in subjects with clinically less marked ET. Thus, the spectrum with the most distinct and representative tremor spectral peak in the 4–12 Hz bandwidth was selected to calculate template tremor frequency values. This procedure was carried out for force and EMG spectra of each ET subject and the resulting template values were used to calculate spectral outcomes when peaks were difficult to discern. Subjects with ET were therefore divided into two groups based on the strength and consistency of motor unit entrainment (Héroux et al., 2009). Subjects in group 1 had prominent spectral peaks in EMG for all trials (relative EMG tremor spectral power > 10%; see below for details). Subjects included in group 2 had less prominent EMG spectral peaks at 5%and 10%-MVCs (relative EMG tremor spectral power 6 10%), which either (1) reduced in amplitude at either or both of the higher intensities or (2) was not present at one of the higher intensities and reduced at the other. The power of the main spectral peak associated with the central tremor component was calculated for each of the four target intensities and four EMG spectra for ET subjects based on the half-power bandwidth (Elble, 2003; Raethjen et al., 2004; Héroux et al., 2009). Furthermore, the percent of the total EMG power in the 0–40 Hz bandwidth accounted for by the central tremor component was computed ([tremor spectral power/total power] 100). Surface EMG reflects the activity of numerous motor units (i.e., interference pattern) and the entrainment associated with ET results in rhythmic bursts of EMG activity. Thus, this measure of the relative power associated with EMG tremor peaks has been used as an index of tremor severity and motor unit entrainment (Elble et al., 1994), the latter referring to the rhythmical synchronization of motor unit firing by abnormal descending oscillatory activity. 2.5. Statistical analysis Given the skewed distribution of tremor amplitude measures in ET (Héroux et al., 2006, 2009; Stacy et al., 2007), all data were examined to verify Gaussian distribution prior to performing statistical analysis. Measures which were not normally distributed were log-transformed (Bland and Altman, 1996) and analyzed again to ensure a Gaussian distribution. Summary statistics and figures present mean and 95% confidence interval values of backtransformed data using the antilog (Bland and Altman, 1996). The process of back-transforming log-transformed data results in asymmetrical 95% confidence intervals, reflecting the skewed distribution of the data. The dependent variables were each compared in a two-way ANOVA with a between subject factor for subject group and repeated
measures for target intensity. When relevant, post-hoc testing was performed using Tukey’s Honestly Significant Difference test. A value of 0.05 was chosen as the level of significance for all tests. Statistical analyses were carried out using SPSS statistical package (version 15.0.1). 3. Results Subject characteristics is presented in Table 1; 13 subjects were included in ET in group 1 and the eight others were included in group 2. There was no strength difference between subjects in the control group (MVC mean: 112.3 N; 95% CI: 97.7–126.9), ET group 1 (MVC mean: 111.6 N; 95% CI: 83.0–140.2) and ET group 2 (MVC mean: 106.5 N; 95% CI: 81.6–131.4) (p = 0.918). An example of force production at all four target intensities for a control subject is shown in Fig. 1A. Overall, the subject was very steady and able to maintain her mean force near the target. The power spectra from each trial after linear detrending are also plotted and illustrate the <3 Hz spectral power associated with visuomotor force corrections (Slifkin et al., 2000). Fig. 1B shows data from a subject with moderate ET of similar wrist extensor strength. The rhythmic tremor fluctuations and the <3 Hz visuomotor component are clearly visible in the power spectra of these data. The apparent trend for increasing tremor peaks was not borne out in the group statistics (see below). 3.1. Force fluctuation amplitude Overall force fluctuation amplitude – a global measure that does not distinguish between force fluctuations associated with pathological tremor from those related to visuomotor corrections and normal physiological force variability – was significantly different between target intensities (p = 0.001) and groups (p = 0.008) (Fig. 2A). There was no significant group by intensity interaction (p = 0.422). Overall force fluctuation amplitude increased significantly at each higher target intensities (p 6 0.001). In terms of the significant main effect for group, post-hoc analysis revealed that overall force fluctuations were significantly greater subjects in ET group 1 compared to ET group 2 and the control group (p 6 0.002). Interestingly, overall force fluctuation amplitude was not significantly different between the control group and ET group 2 (p = 0.677). 3.2. Frequency domain measures of force fluctuations The frequency of the spectral peak associated with tremor activity remained relatively constant across contraction intensities for all ET subjects (Table 2). In terms of the spectral power associated with these tremor peaks, there were significant main effects for group (p = 0.02) with subjects in ET group 1 having greater force tremor power than subjects in ET group 2 (Fig. 2B). There was also
Table 1 Subject characteristics.
Controls (n = 22) ET group 1 (n = 13) ET group 2 (n = 8) a b c d
Age (years)
Gender
Side testeda
Years with tremor
Tremor medications (#subjects: medication, useb)
FTM – A (24)c
FTM – B (36)
64.3 ± 4.3 (38–84) 57.2 ± 0.4 (38–72) 63.5 ± 0.5 (47–74)
9m 13 f 4m 9f 5m 3f
12 D 10 ND 7D 6 ND 6D 2 ND
–
4: b-blocker (H)
27.6 ± 5.2 (6–51) 22.1 ± 4.6 (7–40)
1: primidone (T) 6: b-blockerd (4 T, 3 T and H) 1: b-blocker (T and H)
1.0 ± 1.2 (0–5) 9.8 ± 4.2 (4–19) 7.3 ± 3.3 (2–13)
2.7 ± 1.6 (0–6) 6.0 ± 5.0 (3–13) 2.7 ± 2.1 (2–7)
D = dominant; ND = non-dominant. T = tremor reduction; H = anti-hypertension. Fahn–Tolosa–Marìn Tremor Rating Scale. Beta-adrenergic antagonist.
236
M.E. Héroux et al. / Clinical Neurophysiology 121 (2010) 233–239
Fig. 1. Force data from a (A) control and (B) ET subject. The upper graph shows force production at the four target intensities (from bottom to top: 5%-, 10%-, 20%- and 30%MVC). The bottom graphs correspond to the power spectra of the pictured force time series after linearly detrending.
a significant main effect for contraction intensity (p = 0.012). Posthoc testing revealed that this main effect was due to force tremor power being greater at 5%-MVC compared to both 10%-MVC (p = 0.004) and 20%-MVC (p = 0.015). The group by intensity interaction was not significant (p = 0.121). 3.3. Overall neuromuscular activity amplitude As expected, overall neuromuscular activity increased with increasing contraction intensity (p < 0.001). Post-hoc analysis confirmed that this increase in EMG amplitude was significant for all comparisons (p < 0.001) except for 5%- versus 10%-MVC (p = 0.178). As can be seen in Fig. 3A, the increase in EMG activity was similar in all three groups, which was confirmed by the non-
significant group effect (p = 0.409) and group by intensity interaction (p = 0.272). 3.4. Frequency domain measures of tremor related neuromuscular activity Motor unit entrainment characteristic of ET resulted in EMG spectral peaks at the tremor frequency (Table 2). Fig. 4 shows EMG power spectra at all four target intensities for a control subject and an ET subject in each of group 1 and group 2. The power spectra for the subject from ET group 2 have peaks in power at approximately 7 Hz (Fig. 4B). The peaks at 5%- and 10%-MVC dominate the power spectra, whereas they are proportionally smaller at 20%- and 30%-MVC as the power of neighboring
M.E. Héroux et al. / Clinical Neurophysiology 121 (2010) 233–239
237
Fig. 2. Force fluctuation time series and spectral amplitude results. (A) Overall force fluctuation amplitude increased with increasing force production. Subjects in ET group 1 had greater amplitude force fluctuations than both healthy controls and subjects in ET group 2, whereas subjects in ET group 2 were not significantly different than healthy controls. (B) Tremor spectral power (half-power bandwidth on either side of tremor spectral peak) was greater in ET group 1 compared to those in ET group 2. Tremor spectral power was significantly greater at 5%-MVC compared to 10%- and 20%-MVC. Values are mean ± 95% CI.
Table 2 Force and neuromuscular activity spectral peak frequencies. Peak frequency 5%-MVCa ET group 1 (n = 13) Force 6.3 ± 1.3 (4.25–9.50) EMG 6.2 ± 1.1 (4.25–9.25)
10%-MVC
20%-MVC
30%-MVC
6.3 ± 1.3 (4.50–9.00) 6.3 ± 1.3 (4.00–10.25)
6.5 ± 1.2 (4.25–9.00) 6.5 ± 1.2 (4.50–8.75)
6.6 ± 1.3 (4.25–9.00) 6.6 ± 1.2 (4.50–9.00)
ET group 2 (n = 8) Force 6.6 ± 1.0 (5.00–8.00) EMG 6.6 ± 1.0 (5.00–8.25)
6.5 ± 1.2 (5.25–8.50) 6.5 ± 1.0 (5.50–8.25)
6.8 ± 1.2 (5.00–8.75) 6.75 ± 0.8 (5.75–8.50)
6.8 ± 1.2 (5.00–9.00) 6.9 ± 0.8 (5.25–8.75)
Values are mean ± standard deviation (minimum–maximum). a MVC = maximal voluntary contraction.
frequencies increased. Most striking are the power spectra from the subject in ET group 1 with their large narrow tremor peaks at all intensities.
Fig. 3. Neuromuscular activity (EMG) amplitude and EMG tremor spectral power. (A) Neuromuscular activity levels increased with increasing contraction intensity similarly in all three groups. (B) The amplitude of EMG tremor power (half-power bandwidth on either side of EMG tremor spectral peak) was significantly greater in subjects in ET group 1 compared to those in ET group 2. Interestingly, EMG tremor power did not differ significantly with increasing contraction intensities. (C) Similarly, relative EMG tremor power ([EMG tremor spectral power/total 0–40 Hz power] 100) was not significantly influenced by increasing contraction intensities. Values are mean ± 95% CI.
Given that the two ET groups were largely determined based on the amplitude and consistency of EMG tremor spectral peaks, it was not surprising to find a significant group effect (p = 0.001), with subjects in group 1 having greater EMG tremor power values than those in group 2 (see Fig. 3B). Of greater interest was the lack of a significant main effect for intensity (p = 0.211) or intensity by group interaction (p = 0.399), indicating that, despite the significantly higher overall neuromuscular activity associated with higher target intensities, the strength of oscillatory muscle activity at the tremor frequency remained relative constant at contraction intensities ranging from 5%- to 30%-MVC.
238
M.E. Héroux et al. / Clinical Neurophysiology 121 (2010) 233–239
mained relatively constant at contraction intensities ranging from 10%- to 30%-MVC. Furthermore, while subjects in ET group 1 had greater amplitude force fluctuations compared to subjects in ET group 2 and the control group, the amplitude of the force fluctuations of subjects in ET group 2 was similar to that of the healthy controls at all intensities tested. Our findings are in contrast to the previous report by Gillies (1994) where, based on visual inspection, force tremor spectral power was found to represent a constant proportion of the mean force being produced over a large range of intensities (5–95%MVC). Based on this observation, Gillies concluded that the output of the central oscillator affects a constant proportion of the central drive to the motor neuron pool. Unfortunately, the author does not provide any information regarding the number and severity of the subjects studied. Furthermore, force recordings were obtained by having subjects push against a strain gage with their index and middle fingers with no upper extremity support or stabilization. More recently, Burne et al. (2004) reported a linear increase in the amplitude of EMG tremor spectral peaks when subjects with ET generated various low level forces. The amplitude and range of contraction intensities tested by Burne et al. (2004) were extremely small given that they were selected to simulate the changes in muscular demand associated with varying arm and hand orientation and the associated change in gravitational force. Thus the generalizability of their results is limited to very low level contractions. 4.2. Factors influencing the strength of motor unit entrainment
Fig. 4. Example of power spectra of wrist extensor neuromuscular activity (EMG) from (A) a control subjects, (B) a subject from ET group 2 and (C) a subject from ET group. Each graph consists of four EMG power spectra, one for each target intensity. Note the sharp spectral peak in the subject from ET group 2 (B) at lower intensities, which becomes less distinct at higher contraction intensities as power in neighboring frequencies increases. The subject from ET group 1 had a marked peak at all contraction intensities, with only a slight increase in neighboring frequencies at higher contraction intensities.
When considering the relative power associated with these EMG tremor peaks, an index of tremor severity and motor unit entrainment (Elble et al., 1994), subjects in ET group 1 had higher levels of motor unit entrainment than those in ET group 2 (p < 0.001; Fig. 3C). Although relative EMG tremor power did tend to decrease with increasing contraction intensity, especially in ET group 1, the main effect for intensity (p = 0.105) and the intensity by group interaction (p = 0.335) were not significant. 4. Discussion The key findings from the present study were that: (1) overall force fluctuation amplitude and neuromuscular activity scaled in a signal-dependent manner with increasing contraction intensity in all subjects, whereas tremor spectral power of both force and EMG signals remained constant (EMG) or were reduced (force) with increasing contraction intensity and (2) overall force fluctuation amplitude in subjects with less pronounced tremor spectral peaks (i.e., ET group 2) was similar to healthy controls. 4.1. The effect of contraction intensity on force fluctuations in subjects with ET Despite the presence of greater amplitude force fluctuations that were scaled in a signal-dependent manner in subjects with ET, the portion of the force fluctuations associated with tremor (i.e., force tremor spectral power) was greatest at 5%-MVC and re-
Although the exact mechanisms and time-course of the progression of tremor severity remains unclear (Elble et al., 1994, 2005; Elble, 1995, 2000), results from the present study provide further support that one of the effects of increasing tremor severity is greater motor unit entrainment, especially at higher contraction intensities. Indirectly, this implies that stronger descending oscillatory activity entrains a greater number of small motor units and/or entrains additional higher threshold motor units, which supports our previous findings related to postural tremor (Héroux et al., 2009). In this previous study, however, it was not possible to determine to what extent abnormal descending oscillatory activity interacted with short- and long-latency stretch reflexes elicited by oscillatory motion to produce motor unit entrainment (Elble et al., 1987; Matthews, 1993; Elble, 1996). In the present experiment, we specifically used closed-kinetic chain isometric contractions to minimize the influence of stretch reflexes on EMG activity (Burne et al., 1984; Doemges and Rack, 1992) and thus gain a better understanding of the strength of descending oscillatory drive across a range of light to moderate contraction intensities. Relative EMG tremor power was considerably lower in present study, which involved 21 of the same ET subjects and similar contraction intensities (Héroux et al., 2009). Specifically, the tremor peak in subjects in ET group 1 accounted for, on average, 22–25% of the overall EMG activity when supporting a range of light to moderate loads. When these same subjects were producing closed-kinetic chain isometric contractions, relative EMG tremor power was 16–19% at the two lower contraction intensities and approximately 10% at the two higher intensities. A similar pattern was also present in subjects in ET group 2, but with smaller relative EMG tremor amplitude values (i.e., <8%). Although differences exist between the two tasks, these observations provide indirect evidence that abnormal stretch reflex activity may increase the level of motor unit entrainment when the hand is free to tremor. 4.3. Study limitations Previous studies, including our own, have attempted to group ET subjects based on tremor severity using different criteria (Elble,
M.E. Héroux et al. / Clinical Neurophysiology 121 (2010) 233–239
1986; Louis et al., 2001; Stolze et al., 2001; Koster et al., 2002; Héroux et al., 2009). Although there exists no widely accepted and valid criterion for determining tremor severity, we believe the strength of motor unit entrainment is an appropriate choice for grouping ET subjects because it is a measure of the strength of descending oscillatory drive onto the motor neurons, the key sign associated with ET, and it is correlated with tremor amplitude (Elble, 1986, 1996; Elble et al., 1994). The cut-off value (i.e., 10% relative EMG tremor power) was based on the experience we gained in a previous study involving 21 of the same ET subjects (Héroux et al., 2009); however, it had to be lowered in the present study because relative EMG tremor power was lower in isometric force production than in postural holding. We acknowledge the limitations of a dichotomous grouping criterion on what is likely a complex and continuous spectrum of clinical and neurophysiological status. We felt, however, that the separation of ET subjects into two groups was warranted as it improves the interpretability and generalizability of our findings, as well as allows for comparison with our previous findings. Another issue that occurs in many ET studies is the high variability of tremor data (Elble et al., 1994, 2006; Louis et al., 2001; Héroux et al., 2006, 2009). In the present study, dividing ET subjects into two groups using an index of tremor severity reduced the level of within-group variability. Furthermore, all data were examined to verify for a Gaussian distribution and a log transformation of the data was performed prior to statistical analysis when this criterion was not met. Despite our ability to detect numerous group and group by intensity effects in the present study, the high level of variability in ET groups may have limited our ability to identify more subtle differences between groups. 5. Conclusion The present study has shown that the level of motor unit entrainment resulting from abnormal descending oscillatory activity in ET remains relatively constant across a range of light to moderate contraction intensities. Compared to our previous results, however, relative EMG tremor amplitude was lower during isometric contractions produced in a closed-kinetic chain set-up providing preliminary evidence that abnormal stretch reflex activity may contribute to motor unit entrainment in subjects with ET. References Bilodeau M, Keen DA, Sweeney PJ, Shields RW, Enoka RM. Strength training can improve steadiness in persons with essential tremor. Muscle Nerve 2000;23:771–8. Bland JM, Altman DG. Transformations, means, and confidence intervals. BMJ 1996;312:1079.
239
Burne JA, Blanche T, Morris JJ. Muscle loading as a method to isolate the underlying tremor components in essential tremor and Parkinson’s disease. Muscle Nerve 2004;30:347–55. Burne JA, Lippold OC, Pryor M. Proprioceptors and normal tremor. J Physiol 1984;348:559–72. Deuschl G, Bain P, Brin M. Consensus statement of the Movement Disorder Society on Tremor. Ad Hoc Scientific Committee. Mov Disord 1998;13(Suppl. 3):2–23. Doemges F, Rack PM. Changes in the stretch reflex of the human first dorsal interosseous muscle during different tasks. J Physiol 1992;447:563–73. Elble RJ. Physiologic and essential tremor. Neurology 1986;36:225–31. Elble RJ. The role of aging in the clinical expression of essential tremor. Exp Gerontol 1995;30:337–47. Elble RJ. Central mechanisms of tremor. J Clin Neurophysiol 1996;13:133–44. Elble RJ. Diagnostic criteria for essential tremor and differential diagnosis. Neurology 2000a;54:S2–86. Elble RJ. Essential tremor frequency decreases with time. Neurology 2000b;55:1547–51. Elble RJ. Characteristics of physiologic tremor in young and elderly adults. Clin Neurophysiol 2003;114:624–35. Elble RJ, Higgins C, Elble S. Electrophysiologic transition from physiologic tremor to essential tremor. Mov Disord 2005;20:1038–42. Elble RJ, Higgins C, Leffler K, Hughes L. Factors influencing the amplitude and frequency of essential tremor. Mov Disord 1994;9:589–96. Elble RJ, Higgins C, Moody CJ. Stretch reflex oscillations and essential tremor. J Neurol Neurosurg Psychiatry 1987;50:691–8. Elble RJ, Pullman SL, Matsumoto JY, Raethjen J, Deuschl G, Tintner R. Tremor amplitude is logarithmically related to 4- and 5-point tremor rating scales. Brain 2006;129:2660–6. Gillies JD. Measurement of physiological and essential tremor. In: Gandevia SC, Burke D, Anthony S, editors. Science and practice in clinical neurology. Cambridge University Press; 1994. p. 191–201. Halliday DM, Rosenberg JR, Amjad AM, Breeze P, Conway BA, Farmer SF. A framework for the analysis of mixed time series/point process data – theory and application to the study of physiological tremor, single motor unit discharges and electromyograms. Prog Biophys Mol Biol 1995;64:237–78. Héroux M, Pari G, Norman KE. The effect of inertial loading on postural tremor in essential tremor. J Clin Neurophysiol 2009;120:1020–9. Héroux ME, Parisi SL, Larocerie-Salgado J, Norman KE. Upper-extremity disability in essential tremor. Arch Phys Med Rehabil 2006;87:661–70. Jain S, Lo SE, Louis ED. Common misdiagnosis of a common neurological disorder: how are we misdiagnosing essential tremor? Arch Neurol 2006;63:1100–4. Jones KE, Hamilton AF, Wolpert DM. Sources of signal-dependent noise during isometric force production. J Neurophysiol 2002;88:1533–44. Köster B, Deuschl G, Lauk M, Timmer J, Guschlbauer B, Lücking CH. Essential tremor and cerebellar dysfunction: abnormal ballistic movements. J Neurol Neurosurg Psychiatry 2002;73:400–5. Louis ED, Barnes L, Albert SM, Cote L, Schneier FR, Pullman SL, et al. Correlates of functional disability in essential tremor. Mov Disord 2001;16:914–20. Matthews PB. Interaction between short- and long-latency components of the human stretch reflex during sinusoidal stretching. J Physiol 1993;462:503–27. Raethjen J, Lauk M, Koster B, Fietzek U, Friege L, Timmer J, et al. Tremor analysis in two normal cohorts. Clin Neurophysiol 2004;115:2151–6. Slifkin AB, Vaillancourt DE, Newell KM. Intermittency in the control of continuous force production. J Neurophysiol 2000;84:1708–18. Stacy MA, Elble RJ, Ondo WG, Wu SC, Hulihan J. Assessment of interrater and intrarater reliability of the Fahn–Tolosa–Marin Tremor Rating Scale in essential tremor. Mov Disord 2007;22:833–8. Stolze H, Petersen G, Raethjen J, Wenzelburger R, Deuschl G. The gait disorder of advanced essential tremor. Brain 2001;124:2278–86. Vaillancourt DE, Haibach PS, Newell KM. Visual angle is the critical variable mediating gain-related effects in manual control. Exp Brain Res 2006;173:742–50.
Contents lists available at ScienceDirect
Clinical Neurophysiology journal homepage: www.elsevier.com/locate/clinph
The effect of contraction intensity on force fluctuations and motor unit entrainment in individuals with essential tremor M.E. Héroux a,*, G. Pari b,c, K.E. Norman a,d a
School of Human Kinetic, University of British Columbia, 210 - 6081 University Boulevard. BC, Canada V6T 1Z1 Movement Disorders Clinic, Kingston General Hospital, ON, Canada K7L 2V7 c School of Medicine, Queen’s University, 68 Barrie Street, ON, Canada K7L 3N6 d Centre for Neuroscience Studies, Queen’s University, Botterell Hall, ON, Canada K7L 3N6 b
a r t i c l e
i n f o
Article history: Accepted 28 October 2009 Available online 31 December 2009 Keywords: Force fluctuations Essential tremor Electromyography Power spectral analysis
a b s t r a c t Objectives: Quantify the effect of increasing contraction intensity on the amplitude of force fluctuations and neuromuscular and force tremor spectral power. Methods: Twenty-one subjects with essential tremor (ET) and 22 healthy controls applied isometric wrist extension contractions. Various sub-maximal contraction intensities were evaluated (5%-, 10%-, 20%- and 30%-MVC). Force fluctuations and wrist extensor neuromuscular activity were recorded using a load cell and electromyography (EMG). Results: Higher contraction intensities were associated with larger amplitude force fluctuations and greater neuromuscular activation. However, spectral power associated with tremor peaks remained relatively constant (EMG) or decreased (force) with increasing contraction intensity. Conclusions: Motor unit entrainment associated with centrally generated oscillatory inputs does not increase with greater levels of muscle activation. Significance: Rather than influencing a constant proportion of active motor units, abnormal oscillatory drive influences a relative constant number of total motor units. When combined with the findings from our previous study on postural tremor, the present results provide preliminary evidence that abnormal stretch reflex activity may contribute to this motor unit entrainment. Ó 2009 International Federation of Clinical Neurophysiology. Published by Elsevier Ireland Ltd. All rights reserved.
1. Introduction Recently we have shown that rhythmical motor unit entrainment does not significantly increase when individuals with ET support light to moderate loads despite an increase in overall neuromuscular activity (Héroux et al., 2009). With postural tremor, however, abnormal descending oscillatory activity may interact (synergistically or competitively) with short- and long-latency stretch reflexes caused by oscillatory motion (Elble et al., 1987; Matthews, 1993; Elble, 1996). Thus, we sought to investigate the influence of contraction intensity on the strength of motor unit entrainment in subjects with ET during closed-kinetic chain isometric contractions since the influence of stretch reflexes during such contractions is minimized (Burne et al., 1984; Doemges and Rack, 1992). Previously, Gillies (1994) noted an increase in force fluctuation amplitude with increasing contraction intensity in a report involv-
* Corresponding author. Address: School of Human Kinetics, University of British Columbia, 210-6081 University Blvd., Vancouver, BC, Canada V6T 1Z1. Tel.: +1 604 827 3372; fax: +1 604 822 6842. E-mail address: [email protected] (M.E. Héroux).
ing an unspecified number of ET subjects. Based on visual inspection, the author states that the tremor spectral peak represented a constant proportion of spectral power across a wide range of contraction intensities (5–95% maximal voluntary contraction (MVC)) and concludes that the central oscillator in ET affects a constant proportion of the central drive to the motor neuron pool. More recently, Bilodeau et al. (2000) reported an increase in force fluctuation amplitude with increasing contraction intensity in subjects with ET. Of particular interest was the anecdotal report that tremor spectral peak amplitude appeared to increase at the higher contraction intensity (2.5%- versus 20%-MVC), indicating that force tremor scaled with contraction intensity. Corroborating these findings, Burne et al. (2004) noted a linear increase in electromyography (EMG) tremor spectral peak amplitude when subjects with ET generated various low level isometric contractions (<5%-MVC). It remains unclear, however, whether the increase in tremor spectral power – both in force and EMG signals – increases proportionally to the overall level of muscle activation or whether, similar to what was previously reported for postural tremor, motor unit entrainment is proportionally greater at lower contraction intensities.
1388-2457/$36.00 Ó 2009 International Federation of Clinical Neurophysiology. Published by Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.clinph.2009.10.015
234
M.E. Héroux et al. / Clinical Neurophysiology 121 (2010) 233–239
The purpose of this study was to quantify the effect of increasing contraction intensity on the amplitude of neuromuscular and force tremor spectral power in order to determine whether or not the central oscillatory drive in ET affects a constant proportion of the motor neuron pool as suggested by Gillies (1994). In contrast to our previous study, closed-kinetic chain isometric contractions were used to minimize the influence of stretch reflex to the ongoing neuromuscular activity. 2. Methods 2.1. Subjects Twenty-one subjects with ET and 22 healthy control subjects of similar age participated. Subjects with ET were divided into two groups based on the strength and consistency of motor unit entrainment during isometric force production (see Section 2.5.2 for details; Héroux et al., 2009); two subjects with mild tremor from the previous study had highly variable and often absent motor unit entrainment at all target intensities and their data were not included. Subjects in the ET groups: (1) met the diagnosis criteria of the Consensus Statement of the Movement Disorder Society on Tremor (Deuschl et al., 1998), (2) did not present with factors associated with misdiagnosed ET or other red flag items (Elble, 2000; Jain et al., 2006) and (3) had spectral plots of hand postural tremor and forearm EMG recorded under various loading conditions that were consistent with ET (see Héroux et al. (2009) for details). All subjects provided informed consent to the protocol, which was approved by the local Research Ethics Board. 2.2. Apparatus The subject was seated in front of a computer monitor. The forearm of the side being tested was in full pronation and supported on a height-adjustable table with the wrist in neutral flexion/extension and neutral radial/ulnar deviation. Two straps secured the forearm to the table and a piece of opaque fabric prevented the subject from seeing their forearm and hand during testing. Voluntary isometric wrist extension contractions were made against a support device connected to a load cell (MLP100, Transducer Techniques) secured to the table. During MVC testing, the load cell signal was conditioned (Dataforth DSCA43–10 strain gage input conditioner), filtered using an eighth order analogue Butterworth filter (Max740, Maxim) and sampled at 1024 Hz by a 16-bit analog-to-digital converter (National Instrument PCI-MIO-16XE-10) controlled by a Pentium Xeon 2.66 GHz personal computer. During %-MVC trials, the load cell signal was also amplified with an amplifier gain of 8 (Burster Model 9243) prior to being sampled. For MVC testing the smallest detectable change in force was 0.231 N (range: 0–448 N), whereas it was 0.028 N (range: 0–56 N) for %-MVC testing. The computer monitor (21 in. Dell M992, 60 Hz refresh rate, set to 1024 768 pixels) was positioned 50 cm from the subject; the top of the viewing area was in line with the eyes. During %-MVC testing the subject viewed a vertically centered horizontal line that spanned the width of the viewing area and corresponded to the target force being assessed. A second line corresponding to the force exerted on the load cell scrolled left to right during the trial; the total width of the viewing area corresponded to 12 s. The vertical axis was scaled to ±6 N of the target force for all %-MVC testing resulting in a visual gain of 50 pixels/N. The visual angle of feedback, which is dependent on eye–monitor distance and visual gain, has been shown to influence the amplitude and structure of force fluctuations (Vaillancourt et al., 2006). Pilot testing on healthy subjects revealed that the steadiest trial (target intensity
5%-MVC) resulted in a visual angle of 1.15°, which ensured that the potential influence of visual angle on force fluctuations was negligible (see Vaillancourt et al. (2006) for details). Surface EMG was used to measure the neuromuscular activity of the extensor carpi radialis brevis. Skin preparation, electrode placement and the EMG system used were the same as previously described (see Héroux et al. (2009) for details). The EMG signal was digitized using the previously mentioned 16-bit analog-to-digital converter; all EMG data were sampled at 1024 Hz. Load cell and EMG data were collected using custom-built Labview 8 software (National Instruments) and saved to the hard disk of the personal computer for subsequent analysis. 2.3. Procedure In subjects with ET, the hand reported to have the most severe tremor was measured. In the case of symmetrical tremor, the dominant hand was tested. The side tested in healthy controls was selected in order to obtain approximately the same dominant/nondominant ratio between ET and control groups. The MVC of the wrist extensors was first determined. Following a brief warm-up consisting of 3–4 sub-maximal contractions, the subject performed three successive maximal static contractions lasting 5 s with a 120 s rest period between contractions. Next, the subject was asked to produce and hold a series of constant force contractions at four sub-maximal intensities (5%, 10%, 20% and 30% of MVC). The subject was instructed to extend the wrist against the support device and superimpose the force line overtop of the target line on the monitor. After the initial rise in force and stabilization period, the subject held this level of force production with visual feedback for 8 s. Two blocked trials were recorded at each target intensity; the order of testing of the four target intensities was selected randomly. There was a 60 s rest period between trials of a given block and a minimum of 120 s rest between each block. 2.4. Data processing Prior to performing all data analyses, load cell and EMG data were conditioned by the following methods. The EMG data were digitally rectified and the offset value was removed. Next, the EMG and load cell data were digitally filtered using a dual-pass fourth order Butterworth filter with a low-pass cut-off frequency of 40 Hz. Furthermore, the 8 s window of force data from each trial was linearly detrended. All data processing and subsequent time and frequency analyses were performed using software written in Matlab 7 (The MathWorks, Inc.). 2.4.1. Amplitude of force fluctuations and EMG activity The maximum force value produced across all three trials was selected as the MVC value. Overall force fluctuation amplitude during %-MVC trials was determined by calculating the standard deviation of the force time series. This is a global measure that does not distinguish between force fluctuations associated with pathological tremor (4–12 Hz) from those related to visuomotor corrections (0–3 Hz) and normal physiological force variability (8–12 Hz) (Elble, 1996; Jones et al., 2002; Slifkin et al., 2000). For the EMG signal, the mean level of neuromuscular activity for each trial was computed. The mean of both trials from a given target force was computed for all measures and used for statistical analysis. 2.4.2. Spectral measures of isometric force tremor and neuromuscular activity Force and EMG auto-spectra were calculated for each target intensity and visual condition using the method of disjoint sections (Halliday et al., 1995). The spectra were estimated by averaging the
235
M.E. Héroux et al. / Clinical Neurophysiology 121 (2010) 233–239
finite Fourier transforms calculated for non-overlapping windows of 4096 points, which resulted in a frequency resolution of 0.25 Hz. As was the case in our previous study, visual inspection of individual EMG auto-spectra revealed reduced amplitude in, or total absence of, a prominent peak indicative of pathological tremor at one or sometimes two of the higher target intensities (i.e., 20%- or 30%-MVC) in subjects with clinically less marked ET. Thus, the spectrum with the most distinct and representative tremor spectral peak in the 4–12 Hz bandwidth was selected to calculate template tremor frequency values. This procedure was carried out for force and EMG spectra of each ET subject and the resulting template values were used to calculate spectral outcomes when peaks were difficult to discern. Subjects with ET were therefore divided into two groups based on the strength and consistency of motor unit entrainment (Héroux et al., 2009). Subjects in group 1 had prominent spectral peaks in EMG for all trials (relative EMG tremor spectral power > 10%; see below for details). Subjects included in group 2 had less prominent EMG spectral peaks at 5%and 10%-MVCs (relative EMG tremor spectral power 6 10%), which either (1) reduced in amplitude at either or both of the higher intensities or (2) was not present at one of the higher intensities and reduced at the other. The power of the main spectral peak associated with the central tremor component was calculated for each of the four target intensities and four EMG spectra for ET subjects based on the half-power bandwidth (Elble, 2003; Raethjen et al., 2004; Héroux et al., 2009). Furthermore, the percent of the total EMG power in the 0–40 Hz bandwidth accounted for by the central tremor component was computed ([tremor spectral power/total power] 100). Surface EMG reflects the activity of numerous motor units (i.e., interference pattern) and the entrainment associated with ET results in rhythmic bursts of EMG activity. Thus, this measure of the relative power associated with EMG tremor peaks has been used as an index of tremor severity and motor unit entrainment (Elble et al., 1994), the latter referring to the rhythmical synchronization of motor unit firing by abnormal descending oscillatory activity. 2.5. Statistical analysis Given the skewed distribution of tremor amplitude measures in ET (Héroux et al., 2006, 2009; Stacy et al., 2007), all data were examined to verify Gaussian distribution prior to performing statistical analysis. Measures which were not normally distributed were log-transformed (Bland and Altman, 1996) and analyzed again to ensure a Gaussian distribution. Summary statistics and figures present mean and 95% confidence interval values of backtransformed data using the antilog (Bland and Altman, 1996). The process of back-transforming log-transformed data results in asymmetrical 95% confidence intervals, reflecting the skewed distribution of the data. The dependent variables were each compared in a two-way ANOVA with a between subject factor for subject group and repeated
measures for target intensity. When relevant, post-hoc testing was performed using Tukey’s Honestly Significant Difference test. A value of 0.05 was chosen as the level of significance for all tests. Statistical analyses were carried out using SPSS statistical package (version 15.0.1). 3. Results Subject characteristics is presented in Table 1; 13 subjects were included in ET in group 1 and the eight others were included in group 2. There was no strength difference between subjects in the control group (MVC mean: 112.3 N; 95% CI: 97.7–126.9), ET group 1 (MVC mean: 111.6 N; 95% CI: 83.0–140.2) and ET group 2 (MVC mean: 106.5 N; 95% CI: 81.6–131.4) (p = 0.918). An example of force production at all four target intensities for a control subject is shown in Fig. 1A. Overall, the subject was very steady and able to maintain her mean force near the target. The power spectra from each trial after linear detrending are also plotted and illustrate the <3 Hz spectral power associated with visuomotor force corrections (Slifkin et al., 2000). Fig. 1B shows data from a subject with moderate ET of similar wrist extensor strength. The rhythmic tremor fluctuations and the <3 Hz visuomotor component are clearly visible in the power spectra of these data. The apparent trend for increasing tremor peaks was not borne out in the group statistics (see below). 3.1. Force fluctuation amplitude Overall force fluctuation amplitude – a global measure that does not distinguish between force fluctuations associated with pathological tremor from those related to visuomotor corrections and normal physiological force variability – was significantly different between target intensities (p = 0.001) and groups (p = 0.008) (Fig. 2A). There was no significant group by intensity interaction (p = 0.422). Overall force fluctuation amplitude increased significantly at each higher target intensities (p 6 0.001). In terms of the significant main effect for group, post-hoc analysis revealed that overall force fluctuations were significantly greater subjects in ET group 1 compared to ET group 2 and the control group (p 6 0.002). Interestingly, overall force fluctuation amplitude was not significantly different between the control group and ET group 2 (p = 0.677). 3.2. Frequency domain measures of force fluctuations The frequency of the spectral peak associated with tremor activity remained relatively constant across contraction intensities for all ET subjects (Table 2). In terms of the spectral power associated with these tremor peaks, there were significant main effects for group (p = 0.02) with subjects in ET group 1 having greater force tremor power than subjects in ET group 2 (Fig. 2B). There was also
Table 1 Subject characteristics.
Controls (n = 22) ET group 1 (n = 13) ET group 2 (n = 8) a b c d
Age (years)
Gender
Side testeda
Years with tremor
Tremor medications (#subjects: medication, useb)
FTM – A (24)c
FTM – B (36)
64.3 ± 4.3 (38–84) 57.2 ± 0.4 (38–72) 63.5 ± 0.5 (47–74)
9m 13 f 4m 9f 5m 3f
12 D 10 ND 7D 6 ND 6D 2 ND
–
4: b-blocker (H)
27.6 ± 5.2 (6–51) 22.1 ± 4.6 (7–40)
1: primidone (T) 6: b-blockerd (4 T, 3 T and H) 1: b-blocker (T and H)
1.0 ± 1.2 (0–5) 9.8 ± 4.2 (4–19) 7.3 ± 3.3 (2–13)
2.7 ± 1.6 (0–6) 6.0 ± 5.0 (3–13) 2.7 ± 2.1 (2–7)
D = dominant; ND = non-dominant. T = tremor reduction; H = anti-hypertension. Fahn–Tolosa–Marìn Tremor Rating Scale. Beta-adrenergic antagonist.
236
M.E. Héroux et al. / Clinical Neurophysiology 121 (2010) 233–239
Fig. 1. Force data from a (A) control and (B) ET subject. The upper graph shows force production at the four target intensities (from bottom to top: 5%-, 10%-, 20%- and 30%MVC). The bottom graphs correspond to the power spectra of the pictured force time series after linearly detrending.
a significant main effect for contraction intensity (p = 0.012). Posthoc testing revealed that this main effect was due to force tremor power being greater at 5%-MVC compared to both 10%-MVC (p = 0.004) and 20%-MVC (p = 0.015). The group by intensity interaction was not significant (p = 0.121). 3.3. Overall neuromuscular activity amplitude As expected, overall neuromuscular activity increased with increasing contraction intensity (p < 0.001). Post-hoc analysis confirmed that this increase in EMG amplitude was significant for all comparisons (p < 0.001) except for 5%- versus 10%-MVC (p = 0.178). As can be seen in Fig. 3A, the increase in EMG activity was similar in all three groups, which was confirmed by the non-
significant group effect (p = 0.409) and group by intensity interaction (p = 0.272). 3.4. Frequency domain measures of tremor related neuromuscular activity Motor unit entrainment characteristic of ET resulted in EMG spectral peaks at the tremor frequency (Table 2). Fig. 4 shows EMG power spectra at all four target intensities for a control subject and an ET subject in each of group 1 and group 2. The power spectra for the subject from ET group 2 have peaks in power at approximately 7 Hz (Fig. 4B). The peaks at 5%- and 10%-MVC dominate the power spectra, whereas they are proportionally smaller at 20%- and 30%-MVC as the power of neighboring
M.E. Héroux et al. / Clinical Neurophysiology 121 (2010) 233–239
237
Fig. 2. Force fluctuation time series and spectral amplitude results. (A) Overall force fluctuation amplitude increased with increasing force production. Subjects in ET group 1 had greater amplitude force fluctuations than both healthy controls and subjects in ET group 2, whereas subjects in ET group 2 were not significantly different than healthy controls. (B) Tremor spectral power (half-power bandwidth on either side of tremor spectral peak) was greater in ET group 1 compared to those in ET group 2. Tremor spectral power was significantly greater at 5%-MVC compared to 10%- and 20%-MVC. Values are mean ± 95% CI.
Table 2 Force and neuromuscular activity spectral peak frequencies. Peak frequency 5%-MVCa ET group 1 (n = 13) Force 6.3 ± 1.3 (4.25–9.50) EMG 6.2 ± 1.1 (4.25–9.25)
10%-MVC
20%-MVC
30%-MVC
6.3 ± 1.3 (4.50–9.00) 6.3 ± 1.3 (4.00–10.25)
6.5 ± 1.2 (4.25–9.00) 6.5 ± 1.2 (4.50–8.75)
6.6 ± 1.3 (4.25–9.00) 6.6 ± 1.2 (4.50–9.00)
ET group 2 (n = 8) Force 6.6 ± 1.0 (5.00–8.00) EMG 6.6 ± 1.0 (5.00–8.25)
6.5 ± 1.2 (5.25–8.50) 6.5 ± 1.0 (5.50–8.25)
6.8 ± 1.2 (5.00–8.75) 6.75 ± 0.8 (5.75–8.50)
6.8 ± 1.2 (5.00–9.00) 6.9 ± 0.8 (5.25–8.75)
Values are mean ± standard deviation (minimum–maximum). a MVC = maximal voluntary contraction.
frequencies increased. Most striking are the power spectra from the subject in ET group 1 with their large narrow tremor peaks at all intensities.
Fig. 3. Neuromuscular activity (EMG) amplitude and EMG tremor spectral power. (A) Neuromuscular activity levels increased with increasing contraction intensity similarly in all three groups. (B) The amplitude of EMG tremor power (half-power bandwidth on either side of EMG tremor spectral peak) was significantly greater in subjects in ET group 1 compared to those in ET group 2. Interestingly, EMG tremor power did not differ significantly with increasing contraction intensities. (C) Similarly, relative EMG tremor power ([EMG tremor spectral power/total 0–40 Hz power] 100) was not significantly influenced by increasing contraction intensities. Values are mean ± 95% CI.
Given that the two ET groups were largely determined based on the amplitude and consistency of EMG tremor spectral peaks, it was not surprising to find a significant group effect (p = 0.001), with subjects in group 1 having greater EMG tremor power values than those in group 2 (see Fig. 3B). Of greater interest was the lack of a significant main effect for intensity (p = 0.211) or intensity by group interaction (p = 0.399), indicating that, despite the significantly higher overall neuromuscular activity associated with higher target intensities, the strength of oscillatory muscle activity at the tremor frequency remained relative constant at contraction intensities ranging from 5%- to 30%-MVC.
238
M.E. Héroux et al. / Clinical Neurophysiology 121 (2010) 233–239
mained relatively constant at contraction intensities ranging from 10%- to 30%-MVC. Furthermore, while subjects in ET group 1 had greater amplitude force fluctuations compared to subjects in ET group 2 and the control group, the amplitude of the force fluctuations of subjects in ET group 2 was similar to that of the healthy controls at all intensities tested. Our findings are in contrast to the previous report by Gillies (1994) where, based on visual inspection, force tremor spectral power was found to represent a constant proportion of the mean force being produced over a large range of intensities (5–95%MVC). Based on this observation, Gillies concluded that the output of the central oscillator affects a constant proportion of the central drive to the motor neuron pool. Unfortunately, the author does not provide any information regarding the number and severity of the subjects studied. Furthermore, force recordings were obtained by having subjects push against a strain gage with their index and middle fingers with no upper extremity support or stabilization. More recently, Burne et al. (2004) reported a linear increase in the amplitude of EMG tremor spectral peaks when subjects with ET generated various low level forces. The amplitude and range of contraction intensities tested by Burne et al. (2004) were extremely small given that they were selected to simulate the changes in muscular demand associated with varying arm and hand orientation and the associated change in gravitational force. Thus the generalizability of their results is limited to very low level contractions. 4.2. Factors influencing the strength of motor unit entrainment
Fig. 4. Example of power spectra of wrist extensor neuromuscular activity (EMG) from (A) a control subjects, (B) a subject from ET group 2 and (C) a subject from ET group. Each graph consists of four EMG power spectra, one for each target intensity. Note the sharp spectral peak in the subject from ET group 2 (B) at lower intensities, which becomes less distinct at higher contraction intensities as power in neighboring frequencies increases. The subject from ET group 1 had a marked peak at all contraction intensities, with only a slight increase in neighboring frequencies at higher contraction intensities.
When considering the relative power associated with these EMG tremor peaks, an index of tremor severity and motor unit entrainment (Elble et al., 1994), subjects in ET group 1 had higher levels of motor unit entrainment than those in ET group 2 (p < 0.001; Fig. 3C). Although relative EMG tremor power did tend to decrease with increasing contraction intensity, especially in ET group 1, the main effect for intensity (p = 0.105) and the intensity by group interaction (p = 0.335) were not significant. 4. Discussion The key findings from the present study were that: (1) overall force fluctuation amplitude and neuromuscular activity scaled in a signal-dependent manner with increasing contraction intensity in all subjects, whereas tremor spectral power of both force and EMG signals remained constant (EMG) or were reduced (force) with increasing contraction intensity and (2) overall force fluctuation amplitude in subjects with less pronounced tremor spectral peaks (i.e., ET group 2) was similar to healthy controls. 4.1. The effect of contraction intensity on force fluctuations in subjects with ET Despite the presence of greater amplitude force fluctuations that were scaled in a signal-dependent manner in subjects with ET, the portion of the force fluctuations associated with tremor (i.e., force tremor spectral power) was greatest at 5%-MVC and re-
Although the exact mechanisms and time-course of the progression of tremor severity remains unclear (Elble et al., 1994, 2005; Elble, 1995, 2000), results from the present study provide further support that one of the effects of increasing tremor severity is greater motor unit entrainment, especially at higher contraction intensities. Indirectly, this implies that stronger descending oscillatory activity entrains a greater number of small motor units and/or entrains additional higher threshold motor units, which supports our previous findings related to postural tremor (Héroux et al., 2009). In this previous study, however, it was not possible to determine to what extent abnormal descending oscillatory activity interacted with short- and long-latency stretch reflexes elicited by oscillatory motion to produce motor unit entrainment (Elble et al., 1987; Matthews, 1993; Elble, 1996). In the present experiment, we specifically used closed-kinetic chain isometric contractions to minimize the influence of stretch reflexes on EMG activity (Burne et al., 1984; Doemges and Rack, 1992) and thus gain a better understanding of the strength of descending oscillatory drive across a range of light to moderate contraction intensities. Relative EMG tremor power was considerably lower in present study, which involved 21 of the same ET subjects and similar contraction intensities (Héroux et al., 2009). Specifically, the tremor peak in subjects in ET group 1 accounted for, on average, 22–25% of the overall EMG activity when supporting a range of light to moderate loads. When these same subjects were producing closed-kinetic chain isometric contractions, relative EMG tremor power was 16–19% at the two lower contraction intensities and approximately 10% at the two higher intensities. A similar pattern was also present in subjects in ET group 2, but with smaller relative EMG tremor amplitude values (i.e., <8%). Although differences exist between the two tasks, these observations provide indirect evidence that abnormal stretch reflex activity may increase the level of motor unit entrainment when the hand is free to tremor. 4.3. Study limitations Previous studies, including our own, have attempted to group ET subjects based on tremor severity using different criteria (Elble,
M.E. Héroux et al. / Clinical Neurophysiology 121 (2010) 233–239
1986; Louis et al., 2001; Stolze et al., 2001; Koster et al., 2002; Héroux et al., 2009). Although there exists no widely accepted and valid criterion for determining tremor severity, we believe the strength of motor unit entrainment is an appropriate choice for grouping ET subjects because it is a measure of the strength of descending oscillatory drive onto the motor neurons, the key sign associated with ET, and it is correlated with tremor amplitude (Elble, 1986, 1996; Elble et al., 1994). The cut-off value (i.e., 10% relative EMG tremor power) was based on the experience we gained in a previous study involving 21 of the same ET subjects (Héroux et al., 2009); however, it had to be lowered in the present study because relative EMG tremor power was lower in isometric force production than in postural holding. We acknowledge the limitations of a dichotomous grouping criterion on what is likely a complex and continuous spectrum of clinical and neurophysiological status. We felt, however, that the separation of ET subjects into two groups was warranted as it improves the interpretability and generalizability of our findings, as well as allows for comparison with our previous findings. Another issue that occurs in many ET studies is the high variability of tremor data (Elble et al., 1994, 2006; Louis et al., 2001; Héroux et al., 2006, 2009). In the present study, dividing ET subjects into two groups using an index of tremor severity reduced the level of within-group variability. Furthermore, all data were examined to verify for a Gaussian distribution and a log transformation of the data was performed prior to statistical analysis when this criterion was not met. Despite our ability to detect numerous group and group by intensity effects in the present study, the high level of variability in ET groups may have limited our ability to identify more subtle differences between groups. 5. Conclusion The present study has shown that the level of motor unit entrainment resulting from abnormal descending oscillatory activity in ET remains relatively constant across a range of light to moderate contraction intensities. Compared to our previous results, however, relative EMG tremor amplitude was lower during isometric contractions produced in a closed-kinetic chain set-up providing preliminary evidence that abnormal stretch reflex activity may contribute to motor unit entrainment in subjects with ET. References Bilodeau M, Keen DA, Sweeney PJ, Shields RW, Enoka RM. Strength training can improve steadiness in persons with essential tremor. Muscle Nerve 2000;23:771–8. Bland JM, Altman DG. Transformations, means, and confidence intervals. BMJ 1996;312:1079.
239
Burne JA, Blanche T, Morris JJ. Muscle loading as a method to isolate the underlying tremor components in essential tremor and Parkinson’s disease. Muscle Nerve 2004;30:347–55. Burne JA, Lippold OC, Pryor M. Proprioceptors and normal tremor. J Physiol 1984;348:559–72. Deuschl G, Bain P, Brin M. Consensus statement of the Movement Disorder Society on Tremor. Ad Hoc Scientific Committee. Mov Disord 1998;13(Suppl. 3):2–23. Doemges F, Rack PM. Changes in the stretch reflex of the human first dorsal interosseous muscle during different tasks. J Physiol 1992;447:563–73. Elble RJ. Physiologic and essential tremor. Neurology 1986;36:225–31. Elble RJ. The role of aging in the clinical expression of essential tremor. Exp Gerontol 1995;30:337–47. Elble RJ. Central mechanisms of tremor. J Clin Neurophysiol 1996;13:133–44. Elble RJ. Diagnostic criteria for essential tremor and differential diagnosis. Neurology 2000a;54:S2–86. Elble RJ. Essential tremor frequency decreases with time. Neurology 2000b;55:1547–51. Elble RJ. Characteristics of physiologic tremor in young and elderly adults. Clin Neurophysiol 2003;114:624–35. Elble RJ, Higgins C, Elble S. Electrophysiologic transition from physiologic tremor to essential tremor. Mov Disord 2005;20:1038–42. Elble RJ, Higgins C, Leffler K, Hughes L. Factors influencing the amplitude and frequency of essential tremor. Mov Disord 1994;9:589–96. Elble RJ, Higgins C, Moody CJ. Stretch reflex oscillations and essential tremor. J Neurol Neurosurg Psychiatry 1987;50:691–8. Elble RJ, Pullman SL, Matsumoto JY, Raethjen J, Deuschl G, Tintner R. Tremor amplitude is logarithmically related to 4- and 5-point tremor rating scales. Brain 2006;129:2660–6. Gillies JD. Measurement of physiological and essential tremor. In: Gandevia SC, Burke D, Anthony S, editors. Science and practice in clinical neurology. Cambridge University Press; 1994. p. 191–201. Halliday DM, Rosenberg JR, Amjad AM, Breeze P, Conway BA, Farmer SF. A framework for the analysis of mixed time series/point process data – theory and application to the study of physiological tremor, single motor unit discharges and electromyograms. Prog Biophys Mol Biol 1995;64:237–78. Héroux M, Pari G, Norman KE. The effect of inertial loading on postural tremor in essential tremor. J Clin Neurophysiol 2009;120:1020–9. Héroux ME, Parisi SL, Larocerie-Salgado J, Norman KE. Upper-extremity disability in essential tremor. Arch Phys Med Rehabil 2006;87:661–70. Jain S, Lo SE, Louis ED. Common misdiagnosis of a common neurological disorder: how are we misdiagnosing essential tremor? Arch Neurol 2006;63:1100–4. Jones KE, Hamilton AF, Wolpert DM. Sources of signal-dependent noise during isometric force production. J Neurophysiol 2002;88:1533–44. Köster B, Deuschl G, Lauk M, Timmer J, Guschlbauer B, Lücking CH. Essential tremor and cerebellar dysfunction: abnormal ballistic movements. J Neurol Neurosurg Psychiatry 2002;73:400–5. Louis ED, Barnes L, Albert SM, Cote L, Schneier FR, Pullman SL, et al. Correlates of functional disability in essential tremor. Mov Disord 2001;16:914–20. Matthews PB. Interaction between short- and long-latency components of the human stretch reflex during sinusoidal stretching. J Physiol 1993;462:503–27. Raethjen J, Lauk M, Koster B, Fietzek U, Friege L, Timmer J, et al. Tremor analysis in two normal cohorts. Clin Neurophysiol 2004;115:2151–6. Slifkin AB, Vaillancourt DE, Newell KM. Intermittency in the control of continuous force production. J Neurophysiol 2000;84:1708–18. Stacy MA, Elble RJ, Ondo WG, Wu SC, Hulihan J. Assessment of interrater and intrarater reliability of the Fahn–Tolosa–Marin Tremor Rating Scale in essential tremor. Mov Disord 2007;22:833–8. Stolze H, Petersen G, Raethjen J, Wenzelburger R, Deuschl G. The gait disorder of advanced essential tremor. Brain 2001;124:2278–86. Vaillancourt DE, Haibach PS, Newell KM. Visual angle is the critical variable mediating gain-related effects in manual control. Exp Brain Res 2006;173:742–50.