Reliability of single- and paired-pulse transcranial magnetic stimulation for the assessment of knee extensor muscle function

Accepted Manuscript Reliability of single- and paired-pulse transcranial magnetic stimulation for the assessment of knee extensor muscle function

John Temesi, Sandy N. Ly, Guillaume Y. Millet PII: DOI: Reference:

S0022-510X(17)30133-8 doi: 10.1016/j.jns.2017.02.037 JNS 15172

To appear in:

Journal of the Neurological Sciences

Received date: Revised date: Accepted date:

25 October 2016 25 January 2017 15 February 2017

Please cite this article as: John Temesi, Sandy N. Ly, Guillaume Y. Millet , Reliability of single- and paired-pulse transcranial magnetic stimulation for the assessment of knee extensor muscle function. The address for the corresponding author was captured as affiliation for all authors. Please check if appropriate. Jns(2017), doi: 10.1016/ j.jns.2017.02.037

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 proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT Reliability of single- and paired-pulse transcranial magnetic stimulation for the assessment of knee extensor muscle function

Human Performance Laboratory, Faculty of Kinesiology, University of Calgary, Calgary,

IP

a

T

John Temesia, Sandy N. Lya, Guillaume Y. Milleta

CR

Alberta, Canada

US

Corresponding author:

AN

Dr Guillaume Y. Millet, Human Performance Laboratory, Faculty of Kinesiology, University of Calgary, 2500 University Drive NW, Calgary, Alberta, CANADA, T2N 1N4. Tel. + 1 (403) 220-

AC

CE

PT

ED

M

3649. Fax. +1 (403) 220-0448. E-mail: [email protected]

1

ACCEPTED MANUSCRIPT Abstract This study examined inter-session and intra-session transcranial magnetic stimulation (TMS) reliability at two test stimulus intensities in the knee extensors. Strong and weak TMS was delivered via single- and paired- (3-ms and 100-ms inter-stimulus interval) pulses on the same

T

day and different days. All stimuli were delivered during isometric contractions of the knee

IP

extensors at 20% of maximal voluntary force. Motor-evoked potentials (MEP) were assessed in

CR

quadriceps femoris muscles. Relative (intra-class correlation coefficient, ICC) and absolute (standard error of measurement, SEM) reliability and variability (coefficient of variation) were

US

assessed. MEPs elicited by strong and weak single-pulse TMS had excellent relative reliability in

AN

all muscles as did weak short-interval and strong long-interval paired-pulse TMS (all ICC > 0.75). Conversely, relative reliability of strong short-interval and weak long-interval paired-pulse

M

TMS was lower (ICC: 0.34-0.83 and 0.22-0.97, respectively). MEP size variability was lower (P

ED

< 0.05) and SEM comparable or lower in strong compared to weak TMS conditions. These results suggest single- and paired-pulse TMS at both strong and weak intensities are generally

PT

reliable in the knee extensors. Strong (or both strong and weak) single-pulse TMS is

CE

recommended. The results indicate using weak test pulses for short-interval and strong test pulses for long-interval paired-pulse TMS are recommended. transcranial

AC

Keywords:

magnetic

stimulation,

corticospinal

excitability,

inhibition,

knee

extensor, reliability Abbreviations

AMT, active motor threshold; ANOVA, analysis of variance; CV, coefficient or variation; EMG, electromyography; I50 , TMS intensity to elicit MEPs at 50% of maximal MEP amplitude; I100 , TMS intensity to elicit maximal amplitude MEPs; ICC, intra-class correlation coefficient; k,

2

ACCEPTED MANUSCRIPT slope constant from Boltzmann modeling; LICIstrong,

paired-pulse TMS

delivered with

suprathreshold conditioning stimulus at I100 and test pulse at I100 and 100-ms inter-stimulus interval; LICIweak , paired-pulse TMS delivered with suprathreshold conditioning stimulus at I100 and test pulse at I50 and 100-ms inter-stimulus interval; MEPmax, estimated maximal MEP

T

amplitude from Boltzmann modeling; MEP strong, single-pulse TMS delivered at I100 ; MEPweak ,

IP

single-pulse TMS delivered at I50 ; Mmax, maximal M-wave amplitude during voluntary

CR

contractions at 20% of maximal voluntary force; MVC, maximal voluntary contraction; RF, rectus femoris; SEM, standard error of measurement; SICIstrong, paired-pulse TMS delivered with

US

subthreshold conditioning stimulus at 90% AMT and test pulse at I100 and 3-ms inter-stimulus

AN

interval; SICIweak , paired-pulse TMS delivered with subthreshold conditioning stimulus at 90% AMT and test pulse delivered at I100 and 3-ms inter-stimulus interval; TMS, transcranial

AC

CE

PT

ED

M

magnetic stimulation; VL, vastus lateralis; VM, vastus medialis.

3

ACCEPTED MANUSCRIPT 1. Introduction Transcranial magnetic stimulation (TMS) is a relatively painless and non-invasive technique used to investigate the properties of the corticospinal pathways. This technique is increasingly being employed in the assessment of inhibition of these pathways via paired-pulse paradigms.

T

Two paired-pulse TMS paradigms have frequently been employed to assess inhibition. A

IP

subthreshold conditioning pulse followed 1-6 ms later by a suprathreshold test pulse elicits an

CR

inhibited response in the relaxed muscle [1] that has been referred to as short-interval intracortical inhibition. This paradigm has also demonstrated inhibition during voluntary muscle

US

contractions [2] although as voluntary force level increases, the amount of inhibition is reduced

AN

via facilitatory mechanisms [3]. Inhibition has also been demonstrated in the relaxed muscle when a suprathreshold conditioning pulse is delivered and followed 50-200 ms later by a

M

suprathreshold test stimulus [4]. This was also observed during voluntary muscle contractions [5]

ED

and is commonly referred to a long-interval intracortical inhibition. McNeil et al. [6] investigated the effects of test-pulse TMS intensity and voluntary force level in the first dorsal interosseous

PT

muscle with long-interval (100-ms interstimulus interval) paired-pulses. They observed that

CE

inhibition was greatest at low voluntary force levels (i.e. 10-25% of maximal voluntary contraction (MVC) force) at all TMS intensities and that as TMS intensity and/or voluntary force

AC

level increased, facilitation was more likely to occur. Stimulus

intensity

is

an

important

factor

influencing

electromyographical (EMG)

parameters elicited by TMS. For example, increasing TMS intensity increases motor-evoked potential (MEP) amplitude and area and silent period duration until plateaus are reached [7, 8]. However, stronger TMS intensities may increase antagonist muscle recruitment and thus reduce the evoked force [9]. There is no consensus for the determination of TMS intensity and different

4

ACCEPTED MANUSCRIPT methods available (e.g. resting motor threshold, active motor threshold (AMT) or stimulus– response curves at different force levels) may result in the selection of different intensities [10], which may in turn affect study conclusions [11]. Increasing TMS intensity recruits greater numbers of both firing and non-firing

T

motoneurons [12]. Whether changes in the number of TMS-recruited firing and non-firing

IP

motoneurons (i.e. by changing TMS intensity) during voluntary muscle contractions influences

CR

the amount or reliability of either short- or long-interval inhibitory responses in the knee extensors is currently unknown.

importance,

especially

during

exercise

and

daily

living

activities

including

AN

functional

US

In most previous TMS studies, the upper limbs have been investigated. Due to their

locomotion, the lower limbs are increasingly being studied, particularly in relation to fatigue and

M

training interventions. An important component of fatigue is central fatigue, which includes

ED

supraspinal fatigue (i.e. within the brain). Since supraspinal fatigue results in force decrements, frequently without decreases (or even increases) in MEP size, inhibitory mechanisms are of

PT

interest. The possible role of intracortical inhibitory mechanisms on neuromuscular function in

CE

acute and chronic interventions in the lower limbs is an area that is poorly understood although both O'Leary et al. [13] and Weier et al. [14] assessed intracortical inhibition in the knee

AC

extensors, in response to fatiguing and training interventions, respectively. Only two recent studies [15, 16] have investigated the reliability of either single- or paired-pulse TMS in the lower limbs, yet the question of appropriate test stimulus intensities has not been addressed. During voluntary knee extension at 5% MVC and using only single TMS pulses, Luc et al. [15] observed generally excellent and good reliability for MEP amplitudes in response to single-pulse TMS over a range of TMS intensities. Subsequently, O'Leary et al. [16] showed good within-

5

ACCEPTED MANUSCRIPT and between-day reliability for MEP amplitudes and silent periods during voluntary 10% MVC isometric contractions. They also investigated short- and long-interval intracortical inhibition and observed both had good within-day reliability while the former and latter had moderate and poor between-day reliability, respectively. When investigating short-interval intracortical inhibition,

T

O'Leary et al. [16] studied the effects of changing the conditioning TMS pulse on the observed

IP

MEPs; however, the effects of stimulus intensity on the test TMS pulse are unknown. This is of

CR

particular interest, especially since their results [16] suggest that some subjects did not display inhibition, possibly due to an unquantifiable contribution of facilitatory factors [3].

US

As a result, the objective of this study was to investigate the reliability of single- and

AN

paired-pulse TMS measures elicited during submaximal voluntary isometric contractions of the

M

knee extensors at two different TMS test pulse intensities.

ED

2. Materials and methods 2.1 Subjects

PT

Twenty healthy adults (10 males and 10 females; 23 ± 5 years; 170 ± 11 cm; 63 ± 11 kg)

CE

participated in this study. All subjects were informed of the experimental protocol and all associated risks prior to giving written informed consent. This study was conducted according to

AC

the Declaration of Helsinki and approved by the University of Calgary Conjoint Health Research Ethics Board. Exclusion criteria included neurological disorders, lower-body injury in the previous 6 months and contraindications to TMS [17]. Subjects were instructed to avoid the consumption of caffeine on the day of the experiment and avoid performing any strenuous exercise during the 48 h prior to testing. 2.2 Experimental design

6

ACCEPTED MANUSCRIPT Subjects visited the laboratory on three occasions. The first session consisted of familiarization with the procedures used in the study and determination of MVC force. The MVC was determined as the highest recorded force output from three MVCs separated by 1 min and the MVC force from the familiarization session was used to calculate the force level for the

T

subsequent sessions. The two remaining sessions were randomized and counterbalanced and both

IP

sessions were conducted at the same time of day with 2 to 14 days between sessions.

CR

Experimental session A consisted of the experimental protocol as outlined below and in Figure 1. Experimental session B consisted of two parts. The first part (session B1) was identical to

US

session A. Session B2 was performed 30 min after the completion of session B1 and consisted of

AN

the 6 experimental TMS conditions and the femoral nerve electrical stimulation condition. 2.3 EMG and force recordings

M

Knee extensor force was measured by a force transducer (LC101-2K, Omegadyne, Sunbury,

ED

USA) with amplifier attached to the right leg by a noncompliant strap immediately proximal to the malleoli of the ankle joint. Subjects were seated in a custom-built isometric knee extension

PT

ergometer in an upright position with the hips and knees at 90 degrees of flexion. The force

CE

transducer was fixed to the ergometer so that force was measured in direct line with the applied force. The force was displayed on a computer screen and subjects received real-time visual

AC

feedback during all voluntary contractions. EMG activity of the right knee extensor (vastus lateralis (VL), rectus femoris (RF) and vastus medialis (VM)) and flexor (biceps femoris) muscles was recorded with pairs of self adhesive surface electrodes (10-mm recording diameter; Meditrace 100, Covidien, Mansfield, USA) in bipolar configuration with 30-mm inter-electrode distance and reference electrode on the patella. The skin where electrodes were placed was shaved, lightly abraded, and cleaned with

7

ACCEPTED MANUSCRIPT isopropyl alcohol in order to achieve a low impedance level (< 5 kΩ). Signals were converted from analog to digital at a sampling rate of 2000 Hz using a PowerLab data acquisition system (16/35, ADInstruments, Bella Vista, Australia) and octal bio-amplifier (ML138, ADInstruments; common mode rejection ratio = 85 dB, gain = 500) with bandpass filter (5-500 Hz)

AC

CE

PT

ED

M

AN

US

CR

IP

T

(ADInstruments). Data were then analyzed offline using LabChart 8 software (ADInstruments).

8

ACCEPTED MANUSCRIPT Fig. 1. Schematic representation of the study protocol. (A) Order of testing components for sessions A and B (including B1 and B2) and (B) testing protocol for FNES condition and the 6 randomised TMS conditions for each session. These were comprised of brief (2-3 s) contractions at 20% MVC every 15 s for each block and with 60 s between blocks of TMS. FNES, femoral nerve electrical stimulation; MVC, maximal voluntary contraction; TMS, transcranial magnetic stimulation. 2.4 Femoral nerve electrical stimulation

IP

T

Single electrical stimuli of 1-ms duration were delivered by constant-current stimulator (DS7A,

CR

Digitimer, Welwyn Garden City, Hertfordshire, UK) to the right femoral nerve via cathode electrode (10-mm stimulating diameter; Meditrace 100, Covidien) in the inguinal triangle and 50

US

× 90 mm rectangular anode electrode (Durastick Plus, DJO Global, Vista, USA) in the gluteal fold. Single stimuli were delivered incrementally until plateaus in maximal M-wave and twitch

AN

amplitudes were reached. Supramaximal stimulation at 130% of the intensity to elicit maximal

2.5 Transcranial magnetic stimulation

M

M-waves (Mmax) was delivered in order to confirm supramaximality.

ED

Two magnetic stimulators (Magstim 2002 , The Magstim Company Ltd, Whitland, UK)

PT

connected via Bistim module were used to stimulate the motor cortex. All stimuli were delivered in Bistim mode with a 110-mm double-cone coil (maximum output of 1.4 T) to preferentially

CE

stimulate the left motor cortex. Subjects wore a cervical collar during TMS delivery in order to

AC

stabilize the neck and head. Single or paired monophasic TMS pulses of 1 ms duration with 0.1 ms rise time were manually delivered to elicit MEPs. Subjects wore a lycra swim cap where intersecting lines were drawn to identify the vertex using the preauricular points and the nasion and inion points. Every centimetre was demarcated from the vertex to 2 cm posterior to the vertex along the nasal-inion line and also to 1 cm over the left motor cortex. At each of these 6 points, a stimulus was delivered at 50% maximal stimulator 9

ACCEPTED MANUSCRIPT output during voluntary contractions at 10% MVC [10]. The TMS hotspot was defined as the site where the largest VL MEP amplitude was elicited, with consideration taken for maximising RF and VM MEP amplitudes and minimising biceps femoris MEP amplitude. The position of the coil was drawn directly onto the swim cap and kept constant throughout the protocol. The coil

T

position was verified by an experienced investigator before the delivery of each stimulus.

IP

A stimulus-response curve was performed at 20% MVC in order to determine stimulus

CR

intensities on each experimental day [10]. Subjects performed 4 brief consecutive contractions with single-pulse TMS delivered at each of the stimulus intensities of 20, 30, 40, 50, 60, 70, and

US

80% maximal stimulator output in randomized order. Additional stimuli were delivered at 90 and

AN

100% maximal stimulator output if a plateau in MEP amplitude was not reached. Single pulses were delivered at 15-s intervals after the subject had reached and stabilized at 20% MVC.

M

Following the stimulus-response curve, AMT was determined to the nearest 5% of maximal

ED

stimulator output by modified relative frequency method [18] based upon the stimulus response curve. The lowest intensity at which visually discernible VL MEPs were observed in at least half

PT

of 4 stimulations was determined to be AMT [18].

CE

MEP amplitude from the stimulus-response curve was modeled by Boltzmann sigmoidal

AC

function [8] using the equation: MEP (I) =

MEPmax I50 − S 1 + exp[ k ]

where MEPmax is the estimated maximal MEP amplitude, I is the stimulus intensity, I50 is the stimulus intensity to elicit MEPs of 50% of maximal MEP amplitude and k is the slope constant. From the equation, the intensity to elicit MEPmax (I100 ) and I50 in VL were determined and used as the test pulse stimulus intensities for the experimental protocol. 2.6 Experimental protocol 10

ACCEPTED MANUSCRIPT The experimental protocol is presented in Figure 1. Supramaximal femoral nerve electrical stimuli were delivered during four brief (~2-3 s) voluntary contractions at 20% MVC. The TMS conditions were performed as 6 randomized blocks of 10 contractions with 60 s between blocks to minimize fatigue. Brief (~2-3 s) 20% MVC voluntary contractions were performed every 15 s

T

during each stimulus condition. The 6 stimulus conditions were: (i) single-pulse TMS delivered

IP

at I50 (MEPweak ), (ii) single-pulse TMS delivered at I100 (MEPstrong), (iii) paired-pulse TMS

CR

delivered with a subthreshold conditioning stimulus at 90% AMT and test pulse at I50 separated by a 3-ms inter-stimulus interval (SICIweak ), (iv) paired-pulse TMS delivered with a subthreshold

US

conditioning stimulus at 90% AMT and test pulse delivered at I100 separated by a 3-ms inter-

AN

stimulus interval (SICIstrong), (v) paired-pulse TMS delivered with a suprathreshold conditioning stimulus at I100 and test pulse at I50 separated by a 100-ms inter-stimulus interval (LICIweak ) and

M

(vi) paired-pulse TMS delivered with a suprathreshold conditioning stimulus at I100 and test pulse

ED

at I100 separated by a 100-ms inter-stimulus interval (LICIstrong). Subjects were asked to recontract as quickly as possible following the delivery of all stimuli [19].

PT

2.7 Data analysis

CE

Peak-to-peak MEP and M-wave amplitudes were measured for all conditions. Single-pulse MEP amplitudes were normalized to maximal M-waves elicited during contractions at 20% MVC. In

AC

paired-pulse conditions, MEP amplitude was considered to be 0 mV when MEP responses could not be clearly identified from involuntary EMG. For paired-pulse conditions, test MEP amplitudes were normalized to the mean of the 10 MEPs for the corresponding single-pulse condition (i.e. MEPweak or MEPstrong). Coefficient of variation (CV) was calculated from the 10 pulses delivered in each condition for each subject as CV = (standard deviation·mean-1 ) × 100. EMG root mean square was determined for the 500 ms prior to the delivery of TMS.

11

ACCEPTED MANUSCRIPT 2.8 Statistical analysis Statistical analyses were performed with IBM SPSS Statistics (version 23; IBM, Armonk, USA) for intra-class correlation coefficient (ICC) and Statistica (version 8; StatSoft, Tulsa, USA) for all other analyses. Shapiro-Wilk and Levene’s tests were used to verify data normality and

T

homogeneity of variances, respectively. Relative overall reliability was assessed by a two-way

IP

mixed intra-class correlation coefficients (ICC 3,1 ) in each muscle. ICC 3,1 was also determined

CR

between sessions A and B1 to assess relative inter-session reliability and between sessions B1 and B2 to assess relative intra-session reliability. ICCs are classified as poor (< 0.40), fair (0.40 –

US

0.59), good (0.60 – 0.74) and excellent (≥ 0.75) [20]. Absolute overall, inter-session and intra-

AN

session reliability was assessed by standard error of measurement (SEM) as calculated according to Weir [21] to account for between-trial noise. Two-way repeated-measures analyses of

M

variance (ANOVA) for session (A, B1 and B2) and TMS intensity were conducted for each

ED

muscle to compare test MEPs (actual values for single-pulse TMS and ratios for SICI TMS) and CVs between TMS intensities and sessions. Subjects demonstrating complete suppression of the

PT

test response in LICIstrong and LICIweak were eliminated from statistical analyses. Therefore, since

CE

subject numbers were different in LICIstrong and LICIweak , one-way repeated-measures ANOVAs for session (A, B1 and B2) were conducted for each muscle to compare test MEPs and CVs

AC

between sessions. A two-way repeated measures ANOVA for session and condition was also performed for EMG root mean square prior to TMS delivery. When the ANOVA revealed significant effects, Tukey’s post-hoc test was used to identify differences. Student’s paired t-tests were used to evaluate differences in stimulus intensities between sessions. Statistical significance was set at P < 0.05. All data are presented as means ± standard deviation.

12

ACCEPTED MANUSCRIPT 3. Results There were no differences in femoral nerve stimulation or TMS intensities employed between the two experimental days (sessions A and B) (P > 0.09 for all). The femoral nerve stimulation intensity for session A was 50 ± 16 mA and 56 ± 16 mA for session B. The TMS intensity to

T

elicit 90% AMT was 33 ± 7% (range: 18-45%) and 32 ± 7% (range: 18-45%) of maximal

IP

stimulator output for sessions A and B, respectively. The TMS intensity to elicit MEPweak was 51

CR

± 10% (range: 39-78%) and 50 ± 12% (range: 35-78%) of maximal stimulator output for sessions A and B, respectively. The TMS intensity to elicit MEP strong was 83 ± 13% (range: 53-

US

100%) and 80 ± 15% (50-100%) of maximal stimulator output for sessions A and B,

AN

respectively.

VL M-wave amplitudes were 15.7 ± 5.1, 15.7 ± 5.5 and 15.3 ± 5.4 mV, RF M-wave

M

amplitudes were 9.9 ± 4.3, 10.3 ± 4.8 and 9.9 ± 4.3 mV and VM M-wave amplitudes were 18.0 ±

ED

5.5, 17.9 ± 5.0 and 17.6 ± 4.7 mV for sessions A, B1 and B2, respectively. There were no differences in M-wave amplitudes between sessions for any muscle (P > 0.25 for all).

PT

There was no difference in EMG root mean square prior to TMS delivery across sessions

CE

or conditions for any muscle (P > 0.12 for all muscles) indicating that any differences between conditions are due to the stimulation paradigm and not the amount of background EMG.

AC

Figure 2 presents raw traces obtained from a representative subject for test MEPs elicited in all single- and paired-pulse TMS conditions. 3.1 Single-pulse TMS Normalized MEP amplitudes from single pulses are presented in Figure 3A-B. Smaller MEPs were elicited in MEPweak than MEPstrong for all muscles (all P < 0.001). SEM is presented in Table 1. ICCs for all three muscles were excellent for both MEPweak and MEPstrong (Table 2). The

13

ACCEPTED MANUSCRIPT variability for MEPstrong was less than for MEPweak (all P < 0.001; Table 3) and was similar

ED

M

AN

US

CR

IP

T

across sessions (all P > 0.05).

AC

CE

PT

Fig. 2. Representative electromyographical traces from the 6 experimental TMS conditions for the vastus lateralis from a single subject. (A) Single-pulse TMS at I50 (MEPweak ), (B) single-pulse TMS at I100 (MEPstrong), (C) paired-pulse TMS delivered with a subthreshold conditioning stimulus at 90% AMT and test pulse at I50 and separated by a 3-ms inter-stimulus interval (SICIweak ), (D) paired-pulse TMS delivered with a subthreshold conditioning stimulus at 90% AMT and test pulse at I100 separated by a 3-ms inter-stimulus interval (SICIstrong), (E) pairedpulse TMS delivered with a suprathreshold conditioning stimulus at I100 and test pulse at I50 separated by a 100-ms inter-stimulus interval (LICIweak ) and (F) paired-pulse TMS delivered with a suprathreshold conditioning stimulus at I100 and test pulse at I100 separated by a 100-ms inter-stimulus interval (LICIstrong). AMT, active motor threshold; I50 , TMS intensity to elicit MEPs at 50% of maximal MEP amplitude; I100 , TMS intensity to elicit maximal amplitude MEPs; MEP, motor-evoked potential; TMS, transcranial magnetic stimulation. The arrow indicates the timing of TMS delivery in each panel.

3.2 Paired-pulse TMS with 3-ms inter-stimulus interval Normalized MEP amplitudes elicited by strong and weak test pulses with a 3-ms inter-stimulus interval are presented in Figure 3C-D. There was less inhibition in SICIstrong than SICIweak for all

14

ACCEPTED MANUSCRIPT muscles (all P < 0.001). Inhibition was not always observed in SICIweak or SICIstrong during single sessions for at least one muscle (11 and 17%, respectively for SICIweak and SICIstrong). This lack

AC

CE

PT

ED

M

AN

US

CR

IP

T

of inhibition was observed in all three muscles.

15

ACCEPTED MANUSCRIPT

CR

IP

T

Fig. 3. Mean normalized MEP amplitudes for VL, RF and VM for the three experimental sessions. (A) Single-pulse TMS at I50 (MEPweak ), (B) single-pulse TMS at I100 (MEPstrong), (C) paired-pulse TMS delivered with a subthreshold conditioning stimulus at 90% AMT and test pulse at I50 and separated by a 3-ms inter-stimulus interval (SICIweak ), (D) paired-pulse TMS delivered with a subthreshold conditioning stimulus at 90% AMT and test pulse at I 100 separated by a 3-ms inter-stimulus interval (SICIstrong), (E) paired-pulse TMS delivered with a suprathreshold conditioning stimulus at I100 and test pulse at I50 separated by a 100-ms interstimulus interval (LICIweak , n = 10) and (F) paired-pulse TMS delivered with a suprathreshold conditioning stimulus at I100 and test pulse at I100 separated by a 100-ms inter-stimulus interval (LICIstrong, n = 20 for RF and n = 19 for VL and VM). AMT, active motor threshold; I50 , TMS intensity to elicit MEPs at 50% of maximal MEP amplitude; I100 , TMS intensity to elicit maximal amplitude MEPs; Mmax, maximal M-wave amplitude; MEP, motor-evoked potential; RF, rectus femoris; TMS, transcranial magnetic stimulation; VL, vastus lateralis; VM, vastus medialis. Values are presented as means ± standard deviation.

US

SEM for SICI are presented in Table 1. ICCs for all muscles were excellent for SICI weak

AN

and fair to good for SICIstrong (Table 2). The variability in SICIstrong was lower than in SICIweak (all P < 0.001; Table 3) and there were no session effects for CV in either SICI weak or SICIstrong

M

(all P > 0.05).

ED

3.3 Paired-pulse TMS with 100-ms inter-stimulus interval Normalized MEP amplitudes elicited by strong and weak test pulses with a 100-ms inter-

PT

stimulus interval are presented in Figure 3E-F. In LICIweak , the test MEP was completely

CE

supressed in 10 subjects in all sessions and in LICIstrong, the test MEP was completely suppressed in VL and VM in one subject in all sessions. Conversely, in LICIstrong inhibition of the test MEP

AC

was not observed in all subjects (Figure 4). SEM for LICI are presented in Table 1. ICC was excellent for LICIstrong in all muscles and for RF LICIweak while VL and VM ICC in LICIweak were poor to fair (Table 2). The variability in LICIstrong was lower than in LICIweak (P < 0.01 for n = 10; Table 3) and there were no session effects for CV in either LICIweak or LICIstrong for any muscle (all P > 0.05 for n = 10).

16

AN

US

CR

IP

T

ACCEPTED MANUSCRIPT

AC

4. Discussion

CE

PT

ED

M

Fig. 4. EMG traces from the two 100-ms inter-stimulus interval paired-pulse TMS conditions for the vastus lateralis. (A) In one subject, complete inhibition of the test MEP with paired-pulse TMS delivered with a suprathreshold conditioning stimulus at I100 and test pulse at I50 separated by a 100-ms inter-stimulus interval (LICIweak ) and (B) in another subject, facilitation of the test MEP with paired-pulse TMS delivered with a suprathreshold conditioning stimulus at I100 and test pulse at I100 separated by a 100-ms inter-stimulus interval (LICIstrong). Five subjects consistently demonstrated facilitation in vastus lateralis and vastus medialis and one subject demonstrated facilitation in rectus femoris. The observed facilitation remained when MEP area was considered rather than amplitude. I50 , TMS intensity to elicit MEPs at 50% of maximal MEP amplitude; I100 , TMS intensity to elicit maximal amplitude MEPs; MEP, motor-evoked potential; MEPstrong, maximal-amplitude MEP; MEPweak , half maximal-amplitude MEP; TMS, transcranial magnetic stimulation. The arrow indicates the timing of TMS delivery in each panel.

The present study investigated the absolute and relative reliability of knee extensor MEPs elicited by single- and paired-pulse TMS at two TMS test stimulus intensities during submaximal voluntary isometric contractions. The principal findings are that i) that the excellent relative reliability of SICIweak MEPs was better than the fair to good relative reliability of SICIstrong MEPs despite more variability, ii) test MEPs elicited in LICIstrong had excellent relative reliability while reliability for LICIweak ranged from poor to excellent and iii) in LICIweak condition, half the 17

ACCEPTED MANUSCRIPT subjects had completely suppressed test MEPs. Our results also confirm that single-pulse MEPs elicited by strong and weak single pulses had excellent relative reliability and variability for strong single-pulse TMS was less than for weak single-pulse TMS. Altogether, this study is important when considering the testing intensity of single- and paired-pulse TMS protocols.

T

Some practical recommendations are provided below.

IP

The present study contrasts other related knee extensor studies [15, 16] in that the test

CR

stimulus was not related to a percentage of either resting or active motor thresholds and also that two test pulse intensities were employed in paired-pulse paradigms. Given the previously

US

observed sigmoidal relationship between TMS intensity and MEP size [8, 10], the selection of

AN

the test-pulse intensity from the resting or active motor threshold may cause it to correspond to the rising part of the sigmoidal curve or the plateau (i.e. maximal MEP size), depending on the

M

subject. In the present study, TMS intensities were determined at 20% MVC and then evaluated

ED

for reliability at an identical force level. As such, comparisons between studies using thresholds

study.

CE

4.1 Single-pulse TMS

PT

as a basis for TMS intensity determination may not be directly comparable to the results of this

Single-pulse TMS at both strong and weak TMS intensities was shown to have excellent relative

AC

reliability for MEPs elicited in the knee extensor muscles both within a single session and between sessions. This is in agreement with previous studies of the knee extensors during light isometric contractions (i.e. 5-10% MVC) [15, 16] and indicates that MEPs elicited by singlepulse TMS, regardless of how stimulus intensity is determined, are reliable. The intra-session reliability suggests that the reliability of these measures will be maintained during longer

18

ACCEPTED MANUSCRIPT experimental sessions where testing may last several hours. However, absolute reliability (SEM) was better for MEPstrong. Greater variability was observed for MEP weak than MEPstrong and this may be explained by the well-established sigmoidal relationship between TMS intensity and MEP size [8, 10]. The

T

weak TMS intensity was selected to elicit a MEP that was half maximal MEP amplitude and, as

IP

such, this corresponds to the rising phase of the sigmoidal curve. Since even small changes in

CR

corticospinal excitability will shift the response to the left or right of the curve, responses to weak TMS are potentially susceptible to increased variability.

US

4.2 Paired-pulse TMS with 3-ms inter-stimulus interval

AN

SICI is believed to be mediated by GABAA, as demonstrated by the increased SICI response following the administration of positive GABAA modulators [22]. SICIweak demonstrated

M

excellent relative reliability while SICIstrong responses had lower relative reliability (i.e. generally

ED

good). Absolute reliability (SEM) was comparable between strong and weak test pulse conditions (Table 1) and the variability in SICIweak was approximately double SICIstrong despite

PT

the latter having lower relative reliability.

CE

Previous research in the first dorsal interosseous muscle reported greater inhibition with weaker TMS test pulses, provided they were greater than resting motor threshold, and overall

AC

little inhibition during voluntary isometric contractions at 10% MVC [23]. Similarly, the present study observed greater inhibition with weak TMS. The lack of inhibition, and even facilitation, in many subjects in the current study, particularly in SICIstrong, was maintained when MEP area was considered. The lack of observed inhibition is likely due to the combination of inhibitory (i.e.

short-interval

intracortical inhibition)

and

19

facilitatory

(i.e.

short-interval intracortical

ACCEPTED MANUSCRIPT facilitation) mechanisms contributing to paired-pulse paradigms composed of a subthreshold conditioning pulse and suprathreshold test pulse with an inter-stimulus interval of 2-3 ms [3, 24]. 4.3 Paired-pulse TMS with 100-ms inter-stimulus interval LICI is believed to be mediated by GABAB, as demonstrated by the increased LICI response

T

following the administration of positive GABAB modulators [25]. Strong TMS test pulses in the

IP

LICI condition (LICIstrong) demonstrated excellent relative reliability. However, in LICIweak only

CR

RF had excellent relative reliability. The much lower relative reliability in both VL and VM occurred despite TMS intensity being determined in VL. The reliability of LICIweak could also

US

only be assessed in half the subjects because the test MEP was completely suppressed in the

AN

others. The SEM was similar in both LICI conditions for all muscles (Table 1); however, the ratio between the test MEP and corresponding MEP is much larger in LICIstrong. The test MEP

M

being frequently completely suppressed and thus limiting interpretation in LICIweak , in addition

ED

to LICIweak having greater variability (i.e. CV), lower relative reliability and similar SEM despite smaller ratios, indicate that LICIstrong is a more reliable measure.

PT

4.4 Validity

CE

In addition to the question of whether elicited responses to paired-pulse TMS are reliable is whether these measures are valid. As previously mentioned, the SICI paradigm elicits responses

AC

that are comprised of both short-interval inhibitory and facilitatory components [3, 24]. The results of the present study and a current inability to separate inhibitory and facilitatory mechanisms

from

this

understanding changes to

paired-pulse

paradigm

suggest

important

limits,

especially

in

inhibitory corticospinal mechanisms with an intervention (e.g.

fatiguing exercise, training program), as there is currently no way to parse out changes in the inhibitory and facilitatory components. In the LICI paradigm, the presence of facilitation when

20

ACCEPTED MANUSCRIPT TMS is delivered during the silent period has previously been reported in the first dorsal interosseous muscle [6]. McNeil et al. [6] observed that inhibition was reduced with higher TMS test pulse intensity and that inhibition was maximized at low voluntary contraction levels (i.e. 10 and 25% MVC). They suggested that the maintenance of or increase in test MEP size may be due

T

to i) TMS activating pathways that are activated prior to the neurons that are inhibited, ii)

IP

different pathways being activated by TMS than by voluntary drive to muscles or iii) input from

CR

TMS being sufficiently greater than that of the strongest voluntary drive and able to overcome the induced inhibition. It is unclear how possible facilitation with a paradigm that is classically

US

interpreted as inhibitory should be interpreted in either cross-sectional or intervention studies.

AN

4.5 Limitations

The present study investigated the reliability of single- and paired-pulse TMS responses at 20%

M

MVC. However, the results reported from the methodology employed here may not apply to

ED

contractions at different intensities. Given the work of McNeil et al. [6], which indicates that inhibition in LICI paradigm is reduced or eliminated at stronger voluntary contraction levels in

PT

the first dorsal interosseous, further studies are required to elucidate these effects in the knee

CE

extensors. For example, fatigue and intervention studies often employ stronger voluntary contraction levels to investigate changes over time such as with the measure of voluntary

AC

activation (e.g. voluntary contractions at 50, 75 and 100% MVC) and voluntary force levels >20% MVC may be more relevant to many fatiguing protocols. Another limitation is that the timing of the testing of female subjects across the menstrual cycle was not controlled in this study. However, the random distribution of female subjects across their menstrual cycles would likely have negated or limited any possible influences on the results. Finally, the aforementioned differences in methodology for determining TMS stimulus intensities (i.e. 50 and 100% of

21

ACCEPTED MANUSCRIPT maximal MEP amplitude in this study and a percentage of resting or active motor thresholds in other studies) make comparisons between previous studies difficult. 4.6 Recommendations In studies investigating cortical voluntary activation, a strong TMS intensity is necessary to elicit

T

maximal force responses. Therefore, if only one single-pulse TMS intensity is employed, it is

IP

recommended that this TMS intensity elicit maximal responses (i.e. strong). This is reinforced by

CR

both lower SEM and variability with strong TMS. However, due to the differential changes in MEP responses over time previously observed with fatiguing protocols [11, 26], the use of both

US

strong and weak TMS pulses is optimal although this may not always be possible. The results of

AN

the present study indicate that in terms of reliability and variability, SICIweak and SICIstrong are similar, although each has weaknesses. In order to observe inhibition, the use of weak TMS test

M

pulses are recommended with the SICI paradigm. This however must be balanced with greater

ED

variability and SEM with weak test pulses. In the LICI paradigm, it is essential that if change is being quantified (e.g. fatiguing exercise) or comparisons are being made, TMS intensities must

PT

be selected such that the test MEP is not completely suppressed at any point in the protocol.

4.7 Conclusions

CE

Therefore, it is recommended that the TMS test pulse is strong.

AC

This study demonstrates excellent overall, inter-session and intra-session relative reliability of weak short-interval paired-pulse TMS and strong long-interval paired-pulse TMS. Both strong short-interval paired-pulse TMS and weak long-interval paired-pulse TMS have lower relative reliability. This study also reiterated the excellent reliability of single-pulse TMS at both strong and weak intensities. When selecting TMS intensities, strong TMS test pulses are recommended for single-pulse TMS if only one stimulus is being employed, especially when voluntary

22

ACCEPTED MANUSCRIPT activation is being assessed. The use of a weak test pulse to assess short-interval paired-pulse TMS and a strong test pulse for long-interval paired-pulse TMS are also recommended.

Funding

T

This research did not receive any specific grant from funding agencies in the public, commercial,

CR

IP

or not-for-profit sectors.

Acknowledgements

US

The authors gratefully thank all subjects that participated in this study for the generous

AN

contribution of their time. They also sincerely acknowledge the assistance of John Holash for

M

technical support and Gianluca Vernillo for critical review.

ED

References

[1] T. Kujirai, M.D. Caramia, J.C. Rothwell, B.L. Day, P.D. Thompson, A. Ferbert, et al.,

PT

Corticocortical inhibition in human motor cortex, J Physiol. 471 (1993) 501-519.

CE

[2] A. Berardelli, S. Rona, M. Inghilleri, M. Manfredi, Cortical inhibition in Parkinson's disease. A study with paired magnetic stimulation, Brain. 119 (1996) 71-77.

AC

[3] E. Ortu, F. Deriu, A. Suppa, E. Tolu, J.C. Rothwell, Effects of volitional contraction on intracortical inhibition and facilitation in the human motor cortex, J Physiol. 586 (2008) 51475159. [4] J. Valls-Sole, A. Pascual-Leone, E.M. Wassermann, M. Hallett, Human motor evoked responses to paired transcranial magnetic stimuli, Electroencephalogr Clin Neurophysiol. 85 (1992) 355-364.

23

ACCEPTED MANUSCRIPT [5] D. Claus, M. Weis, U. Jahnke, A. Plewe, C. Brunholzl, Corticospinal conduction studied with magnetic double stimulation in the intact human, J Neurol Sci. 111 (1992) 180-188. [6] C.J. McNeil, P.G. Martin, S.C. Gandevia, J.L. Taylor, Long-interval intracortical inhibition in a human hand muscle, Exp Brain Res. 209 (2011) 287-297.

T

[7] V.K. Kimiskidis, S. Papagiannopoulos, K. Sotirakoglou, D.A. Kazis, A. Kazis, K.R. Mills,

IP

Silent period to transcranial magnetic stimulation: construction and properties of stimulus-

CR

response curves in healthy volunteers, Exp Brain Res. 163 (2005) 21-31.

[8] H. Devanne, B.A. Lavoie, C. Capaday, Input-output properties and gain changes in the

US

human corticospinal pathway, Exp Brain Res. 114 (1997) 329-338.

AN

[9] G. Todd, J.L. Taylor, S.C. Gandevia, Measurement of voluntary activation of fresh and fatigued human muscles using transcranial magnetic stimulation, J Physiol. 551 (2003) 661-671.

M

[10] J. Temesi, M. Gruet, T. Rupp, S. Verges, G.Y. Millet, Resting and active motor thresholds

ED

versus stimulus-response curves to determine transcranial magnetic stimulation intensity in quadriceps femoris, J Neuroeng Rehabil. 11 (2014) 40.

PT

[11] J. Temesi, T. Rupp, V. Martin, P.J. Arnal, L. Feasson, S. Verges, et al., Central fatigue

CE

assessed by transcranial magnetic stimulation in ultratrail running, Med Sci Sports Exerc. 46 (2014) 1166-1175.

AC

[12] P.G. Martin, S.C. Gandevia, J.L. Taylor, Output of human motoneuron pools to corticospinal inputs during voluntary contractions, J Neurophysiol. 95 (2006) 3512-3518. [13] T.J. O'Leary, M.G. Morris, J. Collett, K. Howells, Central and peripheral fatigue following non-exhaustive and exhaustive exercise of disparate metabolic demands, Scand J Med Sci Sports. 26 (2016) 1287-1300.

24

ACCEPTED MANUSCRIPT [14] A.T. Weier, A.J. Pearce, D.J. Kidgell, Strength training reduces intracortical inhibition, Acta Physiol. 206 (2012) 109-119. [15] B.A. Luc, A.S. Lepley, M.A. Tevald, P.A. Gribble, D.B. White, B.G. Pietrosimone, Reliability of corticomotor excitability in leg and thigh musculature at 14 and 28 days, J Sport

T

Rehabil. 23 (2014) 330-338.

IP

[16] T.J. O'Leary, M.G. Morris, J. Collett, K. Howells, Reliability of single and paired-pulse

CR

transcranial magnetic stimulation in the vastus lateralis muscle, Muscle Nerve. 52 (2015) 605615.

AN

an update, Clin Neurophysiol. 122 (2011) 1686.

US

[17] S. Rossi, M. Hallett, P.M. Rossini, A. Pascual-Leone, Screening questionnaire before TMS:

[18] S. Groppa, A. Oliviero, A. Eisen, A. Quartarone, L.G. Cohen, V. Mall, et al., A practical

ED

Neurophysiol. 123 (2012) 858-882.

M

guide to diagnostic transcranial magnetic stimulation: report of an IFCN committee, Clin

[19] J. Mathis, D. de Quervain, C.W. Hess, Dependence of the transcranially induced silent

PT

period on the 'instruction set' and the individual reaction time, Electroencephalogr Clin

CE

Neurophysiol. 109 (1998) 426-435.

[20] D.V. Cicchetti, S.A. Sparrow, Developing criteria for establishing interrater reliability of

127-137.

AC

specific items: applications to assessment of adaptive behavior, Am J Ment Defic. 86 (1981)

[21] J.P. Weir, Quantifying test-retest reliability using the intraclass correlation coefficient and the SEM, J Strength Cond Res. 19 (2005) 231-240.

25

ACCEPTED MANUSCRIPT [22] V. Di Lazzaro, A. Oliviero, E. Saturno, M. Dileone, F. Pilato, R. Nardone, et al., Effects of lorazepam on short latency afferent inhibition and short latency intracortical inhibition in humans, J Physiol. 564 (2005) 661-668. [23] M.I. Garry, R.H. Thomson, The effect of test TMS intensity on short-interval intracortical

T

inhibition in different excitability states, Exp Brain Res. 193 (2009) 267-274.

inhibition

(SICI)

and

short-interval

Neurophysiol. 119 (2008) 2291-2297.

intracortical

CR

intracortical

IP

[24] S.H. Peurala, J.F. Muller-Dahlhaus, N. Arai, U. Ziemann, Interference of short-interval facilitation

(SICF),

Clin

US

[25] M.N. McDonnell, Y. Orekhov, U. Ziemann, The role of GABA(B) receptors in intracortical

AN

inhibition in the human motor cortex, Exp Brain Res. 173 (2006) 86-93. [26] D. Bachasson, J. Temesi, M. Gruet, K. Yokoyama, T. Rupp, G.Y. Millet, et al., Transcranial stimulation

intensity

affects

exercise-induced

M

magnetic

changes

in

corticomotoneuronal

AC

CE

PT

ED

excitability and inhibition and voluntary activation, Neuroscience. 314 (2016) 125-133.

26

ACCEPTED MANUSCRIPT Table 1

Standard Error of Measurement (SEM) for normalized motor-evoked potential

amplitudes elicited

by strong and weak single- and paired-pulse transcranial magnetic

stimulation in the vastus lateralis, rectus femoris, and vastus medialis muscles

T

MEPweak MEPstrong SICIweak SICIstrong LICIweak a LICIstrongb

0.155

0.137

0.109

0.158

0.151

0.080

0.132

0.103

0.126

0.072

0.090

0.116

0.144

0.093

0.094

0.098

0.085

0.036

0.042

0.121

4.91

0.134

0.088

0.160

0.136

4.97

6.43

0.121

0.098

0.147

0.135

4.58

2.73

0.118

0.090

0.139

0.101

4.16

2.86

0.088

Inter-session

3.93

3.34

0.079

Intra-session

3.19

1.99

0.080

Overall

8.01

5.79

Inter-session

8.92

7.16

Intra-session

4.38

3.58

US

Overall

Inter-session

b

M

analysis of non-zero values only (n = 10) analysis of non-zero values only (n = 19 for vastus lateralis and vastus medialis; n = 20 for

AC

a

CE

Intra-session

ED

5.17

PT

Overall

AN

Rectus femoris

Vastus medialis

CR

0.092

IP

Vastus lateralis

rectus femoris)

27

ACCEPTED MANUSCRIPT

Table 2

Two-way mixed intra-class correlation coefficients (ICC 3, 1 ) for normalized motor-

evoked potential amplitudes elicited by strong and weak single- and paired-pulse transcranial

IP

T

magnetic stimulation in the vastus lateralis, rectus femoris, and vastus medialis muscles

0.504

0.893

0.538

0.600

0.871

0.919

0.630

0.489

0.942

0.878

0.801

0.546

0.900

0.820

ED

Overall

0.818

0.968

0.904

Inter-session

0.839

0.956

0.933

Intra-session

0.875

0.985

Overall

0.840

Inter-session

0.799

0.799

0.763

0.357

0.898

0.864

Intra-session

0.948

0.956

0.897

0.827

0.973

0.826

0.823

0.901

0.791

0.596

0.381

0.882

Inter-session

0.840

0.840

0.848

0.565

0.529

0.879

0.827

0.966

0.843

0.570

0.200

0.933

PT CE

Vastus medialis Overall

Intra-session a b

AN

M

Rectus femoris

US

0.591

AC

Vastus lateralis

CR

MEPweak MEPstrong SICIweak SICIstrong LICIweak a LICIstrongb

analysis of non-zero values only (n = 10) analysis of non-zero values only (n = 19 for vastus lateralis and vastus medialis; n = 20 for

rectus femoris)

28

ACCEPTED MANUSCRIPT Table 3

Mean coefficient of variation (CV) for normalized motor-evoked potential amplitudes

elicited by strong and weak single- and paired-pulse transcranial magnetic stimulation in the vastus lateralis, rectus femoris, and vastus medialis muscles across all testing sessions

MEPstrong

SICIweak

SICIstrong

LICIweak a

LICIstrongb

Vastus lateralis

21.3 ± 7.8

11.2 ± 5.2

24.7 ± 8.9

13.4 ± 6.5

31.5 ± 11.9

17.2 ± 10.3

Rectus femoris

21.2 ± 14.0

6.8 ± 3.8

25.6 ± 12.0

9.5 ± 8.6

26.8 ± 12.1

12.3 ± 11.6

Vastus medialis

24.2 ± 10.6

11.6 ± 4.6

28.5 ± 11.4

13.6 ± 5.9

34.5 ± 12.7

19.9 ± 13.2

b

IP

CR

analysis of non-zero values only (n = 10 for all subjects)

AN

a

US

Values are presented as means ± standard deviation.

T

MEPweak

analysis of non-zero values only (n = 19 for vastus lateralis and vastus medialis; n = 20 for

AC

CE

PT

ED

M

rectus femoris)

29

ACCEPTED MANUSCRIPT Highlights Both strong and weak single-pulse TMS have excellent reliability.



Strong long-interval paired-pulse TMS has excellent reliability.



Weak short-interval paired-pulse TMS has excellent reliability.



Weak and strong test pulses are recommended to assess SICI and LICI, respectively.

AC

CE

PT

ED

M

AN

US

CR

IP

T



30