Relationship between back muscle endurance and voluntary activation

Journal of Electromyography and Kinesiology 22 (2012) 383–390

Contents lists available at SciVerse ScienceDirect

Journal of Electromyography and Kinesiology journal homepage: www.elsevier.com/locate/jelekin

Relationship between back muscle endurance and voluntary activation Emily Bottle, Paul H. Strutton ⇑ The Nick Davey Laboratory, Human Performance Group, Division of Surgery, Department of Surgery and Cancer, Faculty of Medicine, Imperial College London, Charing Cross Hospital, London W6 8RF, UK

a r t i c l e

i n f o

Article history: Received 21 October 2011 Received in revised form 8 February 2012 Accepted 8 February 2012

Keywords: Back muscle Transcranial magnetic stimulation Back pain Endurance Fatigue

a b s t r a c t There is some evidence that the Biering-Sorensen endurance test can discriminate low back pain sufferers from healthy individuals and can predict future back pain. This test relies on the subject’s ability to voluntarily drive the back muscles. This neural drive, termed voluntary activation (VA) can be measured using the twitch interpolation technique. The aim of the current study was to investigate the relationship between back muscle endurance and VA. Twenty-one healthy volunteers (10 males) participated. Bilateral electromyographic recordings were obtained from erector spinae and rectus abdominis. Back extensor torque was recorded using a dynamometer. The protocol consisted of measurement of VA (using magnetic stimulation of the brain and assessment of the sizes of the evoked twitches) and measurement of endurance. There was a linear correlation (r2 = 1, P < 0.01) between voluntary torque and VA. The mean (SEM) endurance time was 174.9 (12.8) s. There was no correlation between endurance and VA at either 100% MVC (r2 = 0.01, P = 0.72) or at 50% MVC (r2 = 0.11, P = 0.16). These findings indicate that the endurance of the back muscles, as assessed using this widely utilised test does not appear to be related to a subject’s ability to drive their back muscles voluntarily either maximally or submaximally. Ó 2012 Elsevier Ltd. All rights reserved.

1. Introduction Endurance has been defined as ‘‘the time period a constant (non-fatiguing) force output can be maintained’’ (Bigland-Ritchie, 1984). There is evidence showing that low back pain (LBP) is associated with poor back muscle endurance (Biering-Sorensen, 1984; Nicolaisen and Jorgensen, 1985; Hultman et al., 1993; Alaranta et al., 1995) and one widely used and validated test of back muscle endurance is the Biering-Sorensen (BS) test (Latimer et al., 1999; Moreau et al., 2001). This sustained submaximal test has been shown to discriminate between subjects with low back pain LBP and healthy controls and also shown to predict the future occurrence of LBP (Nicolaisen and Jorgensen, 1985; Hultman et al., 1993). Subjects with LBP have also been shown to have alterations in neuromuscular control, including changes in fatigability (Holmstrom et al., 1992; Mannion et al., 1997a,b; Oddsson and De Luca, 2003), activation (Oddsson and De Luca, 2003), postural responses to perturbations (Hodges and Richardson, 1998, 1999; Radebold et al., 2000), implying a change in the central nervous system control. This is supported by recent brain stimulation studies showing brain pathways to back muscles are altered in subjects with LBP (Strutton et al., 2003, 2005; Tsao et al., 2008, 2011). ⇑ Corresponding author. E-mail address: [email protected] (P.H. Strutton). 1050-6411/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved. doi:10.1016/j.jelekin.2012.02.006

During a sustained submaximal elbow flexion there is an appreciable decline in the level of neural drive, termed voluntary activation (VA) (Allen et al., 1995), to the elbow flexors (Sogaard et al., 2006). VA can be estimated using the twitch interpolation method (Merton, 1954), where a stimulus is delivered to the nerve supplying the muscle, or to the motor cortex, during contractions of varying strengths. Transcranial magnetic stimulation (TMS) of the motor cortex has more recently been used to assess VA in a number of limb muscles (Todd et al., 2003; Lee et al., 2008; Goodall et al., 2009; Sidhu et al., 2009), as well as trunk extensors in healthy subjects (Lagan et al., 2008). VA has not been measured using twitch interpolation in LBP subjects to date, since there are practical difficulties in its measurement in these subjects; a requirement of the technique is the undertaking of maximum voluntary contractions (MVC) by the subjects. The presence of back pain is likely to limit maximal efforts in subjects with LBP; it is known that their MVC force is significantly lower (Oddsson and De Luca, 2003); an accurate measure of VA in subjects suffering from LBP is therefore problematic. As a submaximal contraction progresses, VA declines, but recovers rapidly following the cessation of the contraction (Sogaard et al., 2006); suggesting a relationship between the endurance time and the neural drive, or VA. If there is a steady decline in VA during submaximal contractions until task failure, then it seems plausible to suggest that in subjects who have a

384

E. Bottle, P.H. Strutton / Journal of Electromyography and Kinesiology 22 (2012) 383–390

higher level of VA at the start of the task, their time to task failure might be longer. We therefore hypothesise that those subjects with greater levels of VA of the back extensors would have greater endurance times in the BS test. 2. Materials and methods 2.1. Subjects With ethical approval and written informed consent, 21 healthy university students (10 males and 11 females, mean (SEM) age 21.81 (0.75) years, range 19–34, height 170.53 (3.13) cm, body mass 62.64 (2.59) kg) participated in this study. Subjects with a history of back pain or who met the criteria for exclusion from experiments involving TMS (metal implants, cardiac pacemaker, history of epilepsy or fits, brain damage (or neurosurgery), neurological or psychiatric disorders, currently taking antidepressants or other neuromodulatory drugs or pregnancy) were excluded from taking part. The International Physical Activity Questionnaire (iPAQ) was used to determine the level of activity of the subjects (Craig et al., 2003). 2.2. Torque recordings A Cybex Norm Isokinetic Testing System (Henley Healthcare, USA) with an extendable input arm was used to measure back extensor torque. The output from the Cybex was sampled at 500 Hz by a data acquisition interface (1401 plus and Spike2 software: Cambridge Electronic Design, UK) connected to a personal laptop. 2.3. Electromyographic (EMG) recordings Bilateral EMG recordings were obtained from the erector spinae (ES) muscles at vertebral levels T12 and L4 and from rectus abdominis (RA). The recording of the activity of the trunk flexors (RA) was carried out to observe any activation of these muscles by the TMS, which could potentially reduce the size of the evoked extensor twitch thereby introducing error in the calculation of VA. Pairs of Ag/AgCl EMG electrodes (self-adhesive, 2 cm diameter, CareFusion, UK) were positioned vertically, 2 cm apart, 3 cm either side of the spinous processes at the T12 and L4 levels. For the RA, electrodes were placed vertically, 2 cm apart and 3 cm lateral to the linea alba at the level of the umbilicus. The EMG signals were band-pass filtered (10–2 kHz) and amplified (1000; Iso-DAM, WPI, UK) before being sampled at 4 kHz by a data acquisition interface (1401 plus and Spike2 software; Cambridge Electronic Design, UK) connected to a personal computer for subsequent offline analysis. 2.4. Transcranial magnetic stimulation (TMS) TMS was applied to the motor cortex using a Magstim 2002 stimulator (The Magstim Company Ltd., Dyfed, UK) connected to an angled double cone coil (wing outer diameter 12 cm), positioned with its cross-over located over the vertex with the induced current in the brain flowing in a posterior-to-anterior direction. 2.5. Protocol This study comprised of two main experimental components, measurement of VA using the twitch interpolation method and measurement of endurance using the BS test. To avoid any influence on the results by fatigue or practise, the order in which subjects undertook the two tests was randomised and each subject was given appropriate rest in-between.

Subjects were positioned prone on the dynamometer bench, with their legs strapped securely. The iliac crests were aligned with the pivot of the dynamometer lever arm and the lever arm itself was positioned over the lower edge of the scapulae (see Fig. 1A). Subjects performed three brief (2s) maximum voluntary contractions (MVC), consistent verbal encouragement was given throughout and the maximum of the three peak torque values was taken as the control MVC. Submaximal levels (90%, 75%, 50% and 25% MVC) were calculated from the control MVC and target torque indicator lines positioned on the computer screen to provide the subjects with visual feedback. The TMS intensity to be used for the calculation of VA was then determined while the subject maintained a weak contraction (25% MVC). Two stimuli (separated by at least 2 s) were delivered at 50% maximum stimulator output (% MSO) and further pairs of stimuli were delivered at increasing intensities (5% increments) with approximately 5 s rest between pairs at different intensities. A recruitment curve of the stimulus intensity vs. the amplitudes of the motor evoked potentials (MEPs) was constructed to establish the appropriate stimulus intensity (where the size of the MEPs did not increase further when the stimulus intensity was increased); the range of stimulus intensities used was 50–80% MSO. Each subject performed a randomised sequence of 30 contractions, six at each of the submaximal levels (25%, 50%, 75% and 90% MVC) and six at 100% MVC. Each contraction was held stable at the correct level, at which point TMS was delivered. Consistent verbal encouragement was provided throughout to ensure the required contraction strength was maintained. At least 4 s rest was taken between contractions to minimise fatigue (see Fig. 1B). 2.6. Biering-Sorensen test Each subject lay in the prone position on the bench with the upper edge of their iliac crests aligned with the edge of the bench with their arms resting on a chair in front of them. Subjects were secured to the bench with straps placed around their hips and calves. An inclinometer was placed on the subject to allow the experimenter to assess horizontality. To initiate the test, subjects lifted their arms off the chair, crossed them in front of their chest and maintained a straight horizontal position for as long as possible. The test was terminated either by the subject due to excessive fatigue or by the assessor when the subject dropped more than 10° from the horizontal. Verbal encouragement was given throughout. 2.7. Data analysis Statistical analyses were carried out using Sigmaplot 11 (Systat Software, Chicago, USA). To confirm there were no order effects, a paired t-test was used to compare the BS times in those subjects who performed VA before the BS test with those who performed VA after the BS test. Evoked twitches were identified visually. The twitch amplitudes were then calculated as the difference between the pre-stimulus torque level (at 5 ms prior to TMS pulse) and the maximum torque level obtained during the twitch. The time-to peak (TTP) amplitudes (from the TMS pulse) of the twitches were also established. In order to calculate VA, the resting twitch amplitude was first estimated. Twitches were evoked by TMS at contraction strengths of 25%, 50%, 75%, 90% and 100% MVC for each subject and linear regression analysis was used for twitches evoked from 50% to 100% MVC to find the y-axis intercept, which was taken as an appropriate estimated resting twitch (Todd et al., 2003). The twitches evoked at 25% MVC were not included in the regression (Todd et al., 2003; Lagan et al., 2008), since they were disproportionately small (see Fig. 2B), due to reduced corticospinal

385

E. Bottle, P.H. Strutton / Journal of Electromyography and Kinesiology 22 (2012) 383–390

Fig. 1. (A) Experimental setup and (B) protocols for measurement of voluntary activation. Subjects performed a randomised sequence of 30 contractions, six MVCs and six at each sub-maximal level (25%, 50%, 75%, 90% MVC). Transcranial magnetic stimulation (TMS) was delivered to the motor cortex (arrows) once the appropriate torque level had been reached (reproduced with permission from Lagan et al., 2008 and modified).

MEPs were also identified visually from the averaged (six trials per level of voluntary torque) traces and the amplitudes and areas were obtained using a custom made script. Fixed cursors were positioned around the MEP identified from the averaged trace. These fixed cursors were then applied to the individual frames of EMG and the amplitudes and areas of individual MEPs were obtained. These were then averaged per contraction strength to give a single MEP amplitude and area per subject, for each level of voluntary torque. For each subject MEP amplitude and area were normalised to the maximum obtained from

excitability which occurs with low levels of voluntary contraction (Hess et al., 1987; Ugawa et al., 1995). VA was then calculated as a percentage: (1-superimposed twitch amplitude/estimated resting twitch amplitude)  100. Superimposed twitch amplitudes and levels of voluntary torque for each subject were normalised to the maximum MVC force attained throughout the whole experiment. Linear regression analysis was used to assess the relationship between VA and the BS test time. It was also used to assess the relationship between voluntary torque vs. VA and voluntary torque vs. time to peak amplitude.

A

B

Estimated resting twitch amplitude

70 60

50 ms

25%

Twitch size (Nm)

50% MVC 15Nm

TMS

50 75% MVC

40

90% MVC

30

MVC

20 25% MVC

10 75%

MVC 90%

50%

r2=0.883, P<0.0001

0 0

50

100

150

200

250

300

Voluntary force (Nm) Fig. 2. Data from a single subject showing (A) raw torque traces of evoked twitches at contraction strengths MVC; 90% MVC; 75% MVC; 50% MVC; 25% MVC. The traces have been offset with respect to the level of voluntary torque to allow comparisons of twitch sizes. (B) Superimposed twitch amplitude following TMS of the motor cortex at varying contraction strengths. The solid line is the regression line between 50% MVC to MVC. The dashed line represents an extrapolation of this line back to the y-axis intercept, which is taken as an estimate of the resting twitch amplitude. Closed symbols of differing types indicate different levels of voluntary torque.

386

E. Bottle, P.H. Strutton / Journal of Electromyography and Kinesiology 22 (2012) 383–390

A

B

100

14

200

90

180

80

160

r² = 1, P<0.001

70 60

8

50 6

40 30

4

Time to peak amplitude (ms)

Superimposed Twitch (%MVC)

r² = 0.998, P<0.001 10

Voluntary Activation (%)

12

140 120 100 80 60

20

40

10

20

2

0

40

50

60

70

80

90

0 100

R² = 0.9978

0 0

20

Voluntary Torque (%MVC)

40

60

80

100

Voluntary Torque (%MVC)

Fig. 3. Group mean (±SEM) data for (A) voluntary activation (open symbols) and superimposed twitch size (closed symbols) and (B) time-to-peak amplitude of the superimposed twitches evoked by TMS of the motor cortex at varying levels of voluntary torque. Regression lines (and their statistics) are for group data. In (B), individual data are also plotted as smaller black symbols to show range of data across the levels of voluntary torque.

that muscle during the entire experiment (MEPmax), regardless of the level of voluntary torque during which this was obtained. MEP amplitudes and areas in ES and RA were compared using repeated measures two-way ANOVA with factors; muscle [left T12, left L4, right T12, right L4] and ES contraction status [25%, 50%, 75%, 90% and 100% MVC]. The raw EMG traces from the BS test were evaluated for characteristics of fatigue. They were divided into 10% epochs over the total duration of the endurance test. Power spectra were constructed over 10% intervals and the median frequency was calculated for each interval. The RMS of EMG was also calculated over the same epochs and normalised to the peak RMS EMG ob-

A

Left Erector Spinae (T12)

Left Rectus Abdominis

tained during a 100 ms window centred around the peak torque value during the three MVCs carried out at the start of the protocol. Median frequency and RMS at each 10% interval from the ES muscles at vertebral level T12 was compared using repeated measures two-way ANOVA with factors side [left, right] and epoch [10%, 20%, etc. to 100%]. This was repeated for vertebral level L4. Linear regression analysis was performed between the iPAQ scores and either the VA (at 100% or 50% MVC) or the BS test time. Statistical significance was taken as P < 0.05 (or as appropriate for the Holm–Sidak post hoc tests) and all data are presented as mean (SEM).

B

25% MVC

80

MEP amplitude (% MEP max)

70

50% MVC

75% MVC

90% MVC

60

*

50 40 30 20 10 0

0

20

40

60

80

100

Voluntary torque (%MVC) MVC

1mV TMS

50ms

TMS

Fig. 4. (A) Representative data from the left RA and left EST12 from a single subject showing the EMG responses evoked by TMS of the motor cortex (arrows) at varying levels of voluntary torque. (B) Grouped data showing mean (±SEM) MEP amplitudes in left EST12 (squares), right EST12 (diamonds), left ESL4 (triangles) and right ESL4 (circles). ⁄ P < 0.001 with respect to MEP amplitudes obtained at all other levels of voluntary torque.

387

E. Bottle, P.H. Strutton / Journal of Electromyography and Kinesiology 22 (2012) 383–390

MVC (target torque of 50% MVC) to 84.19 (1.84) % at a voluntary torque of 92.60 (0.75) % MVC (target torque of 100% MVC). The linear regression between target torques of 50% and 100% MVC was statistically significant (r2 = 1, P < 0.001).

3.1. Twitch interpolation Mean MVC torque was 220.37 (12.73) Nm. Twitches were evoked by TMS in 95.8% of maximal contractions and in 100% of sub maximal contractions (90%, 75%, 50% and 25% MVC). Fig. 2 shows data from a single subject; Fig. 2A shows an example of the raw torque traces obtained in response to TMS. Fig. 2B shows the graph of TMS-evoked superimposed twitch amplitude against voluntary torque and the linear regression line obtained. Extrapolation of the line of best fit for voluntary torque against superimposed twitches for contractions of 50–100% MVC revealed variability between subjects. One subject’s data were excluded, since the linear regression was not statistically significant. The estimated resting twitch (and VA derived using this) would therefore be inaccurate and inappropriate to use in the group analysis. The mean regression coefficient of the remaining 20 subjects was 0.70 (0.04). The mean estimated resting twitch amplitude was 20.75 (1.78) % MVC. Fig. 3A shows grouped data for level of voluntary torque against TMS-evoked superimposed twitch amplitudes, it shows that twitch amplitudes decrease linearly as the voluntary torque increases from 50% to 100% MVC (r2 = 0.998, P < 0.001). 3.2. Voluntary activation Fig. 3A shows group data of voluntary torque against VA. VA increased from 42.76 (4.25) % at a voluntary torque of 47.27 (0.56) %

A

110 100

Median Frequency (Hz)

90 80

*

70 60 50 40 30 20 10

3.3. Time-to-peak amplitude of the superimposed twitch There was a significant correlation between voluntary torque and the time-to-peak amplitudes of the superimposed twitches (r2 = 0.9981, P < 0.001, see Fig. 3B). 3.4. Motor evoked potentials (MEPs) Fig. 4 shows (A) single subject EMG data (average traces of six stimuli for left RA and left EST12) and (B) grouped data of voluntary torque against MEP amplitudes (in back muscles only) expressed as %MEPmax. For the ES muscles, there were no overall significant differences in either the MEP amplitudes (F = 0.392, P = 0.759) or MEP areas (F = 0.560, P = 0.643) between muscles but significant differences in MEP amplitudes (F = 26.76, P < 0.001) and MEP areas (F = 9.33, P < 0.001) between different levels of voluntary torque. Holm–Sidak post hoc comparisons revealed MEP amplitudes and areas at 25% MVC were significantly smaller than those at 50%, 75%, 90% and 100% MVC (P < 0.001 in all cases). For the RA muscles there were no significant differences in MEP amplitudes or areas between the left and right sides or between differing levels of voluntary torque. RMS EMG amplitude (%EMG during MVC)

3. Results

B

60

50

40

30

20

0 0

10

20

30

40

50

60

70

80

90

100

0

10

60

D

20

BS test time (% total time)

C

100

VA (%) at 50% MVC

VA (%) at MVC

90

80

70

r2=0.008, P=0.717

60 60

80

100

120 140

160

180 200 220

BS test time (s)

240

260

280

30 40 50 60 70 BS test time (% total time)

80

90

100

50

40

30

r2=0.111, P=0.163

20 60

80

100

120

140

160

180

200

220

240

260

280

BS test time (s)

Fig. 5. Upper panels: (A) grouped data for median frequency of EMG, (B) RMS EMG amplitude during the Biering-Sorensen test for left EST12 (squares), right EST12 (diamonds), left ESL4 (triangles) and right ESL4 (circles). ⁄P < 0.001 with respect to values during first 10% of BS test. The closed symbols in (B) indicate where RMS EMG was significantly higher in the respective muscle from the first 10% epoch. Lower panels: scatter plot showing BS test times against voluntary activation measured at (C) MVC and (D) 50% MVC. Regression lines (and their statistics) are shown.

388

E. Bottle, P.H. Strutton / Journal of Electromyography and Kinesiology 22 (2012) 383–390

3.5. Biering-Sorensen test The mean BS time was 174.9 (16.85) s. There was no significant difference in BS time if the test was performed before or after measurement of VA (P = 0.50). One subject with a score of more than two standard deviations from the mean was excluded from group analysis. Fig. 5 (A) shows grouped data for median frequency of EMG and (B) RMS EMG amplitude during the BS test. At both vertebral levels T12 and L4, ANOVA revealed a significant difference between the median frequency values among the different 10% epochs, T12 (F = 68.91, P < 0.001) and L4 (F = 102.024, P < 0.001). Multiple comparisons showed a significant difference between the first 10% epoch and last 10% epoch, P < 0.001 for both ES at T12 and L4. For the RMS EMG amplitude, there was a significant difference between muscles (F = 11.323, P < 0.001) and between 10% epochs (F = 5.040, P < 0.001). Multiple comparisons showed EMG was significantly higher from 60–100% vs. 10% for RES T12 (P = <0.005), between 70–100% vs. 10% for RES L4 (P = <0.008) and between 80% vs. 10% for LES T12 (P = 0.002). There was no significant correlation between BS test times and VA at either 100% MVC ((r2 = 0.01, P = 0.717) or at 50% MVC (r2 = 0.11, P = 0.163); see Fig. 5C and D). Furthermore there was no correlation between the slope of the decline in median frequency of the EMG activity of the extensors over the course of the BS test and VA at either 100% MVC (r2 = 0.01–0.09) or at 50% MVC (r2 = 0.001–0.09). There was also no correlation between the iPAQ scores and either the VA (at 100% or 50% MVC) or the BS test time.

4. Discussion The results of this study reveal that there is no significant correlation between back muscle endurance and neural drive, as assessed using the twitch interpolation method, in our cohort of healthy controls. The ability of a subject to maintain a low level contraction during a back extension test is dependent on the central nervous system to voluntarily activate these muscles. This level of neural drive (termed voluntary activation; see (Gandevia et al., 1995) can be determined using the twitch interpolation technique where, during a contraction, a stimulus is delivered to the nerve supplying that muscle, or the motor cortex using TMS. This technique has been used in a number of other muscles (Todd et al., 2003; Lee et al., 2008; Goodall et al., 2009; Sidhu et al., 2009), including the back extensors (Lagan et al., 2008). The calculation of VA is derived from any increase in force (a superimposed twitch) during a maximal contraction in response to the artificial stimulus divided by the size of the twitch evoked when the muscle is at rest. When using TMS over the motor cortex as the stimulus, the calculation involves using an estimated resting twitch (see (Todd et al., 2003), due to the considerable changes in corticospinal excitability which occur with voluntary contraction (Hess et al., 1987; Ugawa et al., 1995). Using the twitch interpolation technique, high levels of VA have been calculated for the biceps (Todd et al., 2003) and this occurs in both young and old adults (De Serres & Enoka, 1998), but there is considerable variability in the levels of VA obtained between different subjects (VA of 80–100% during maximal efforts (see (Allen et al., 1995). The level of VA measured during MVCs of the back muscles in the current study (84%) was lower than that reported previously using this technique in other muscles (Todd et al., 2003; Sogaard et al., 2006; Taylor and Gandevia, 2008), but higher than that reported previously (68%) in the same muscles (Lagan et al., 2008); that study, however, had subjects with a lower mean back extensor torque and estimated rest twitch amplitudes. It is possible that the reduced level of VA found in the current study

in comparison to that reported in other muscles relates to the nature of the task used to measure VA. A maximal back extension in the prone position is a task very unlikely to be frequently carried out by the subjects and they may be unaccustomed to this. Indeed, it has been reported that average levels of EMG activity in the back muscles of moderately active subjects, recorded over the working day (from the start of work until bedtime) are of the order of 2– 4% (Mork and Westgaard, 2005). It is possible to calculate VA for a submaximal contraction, provided it falls on the linear part of the torque/twitch curve (i.e. from 50% to 100% MVC; see Fig. 2A and (Todd et al., 2004). With this in mind we used a calculation of VA made at 50% MVC to investigate if this correlated with the BS test results. Although the calculated level of VA at 50% MVC was close to 50% MVC (43%), implying validity of the technique, there was no correlation between the level of VA and the BS test times, indicating that high VA of the back muscles is not associated with a long BS test time, as was our original hypothesis. It has been shown that the level and type of physical activity can both increase corticospinal drive to a muscle, reduce a subject’s susceptibility to central fatigue and alter motor unit behaviour and proportions (Thayer et al., 2000; Fulton et al., 2002; Perez et al., 2004; Triscott et al., 2008; Vila-Cha et al., 2010). When subjects completed 6 weeks endurance training on a bicycle ergometer, quadriceps motor unit discharge rates decreased and, conversely, following strength training (high load bilateral leg exercises), they increased (Vila-Cha et al., 2010), suggesting neural adaptations to the exercise regimes. Furthermore, long-term aerobic training, such as long distance running or other endurance sports, has been shown to be associated with an increased proportion of slow twitch non-fatiguable muscle fibres and a reduced proportion of fast twitch fatiguable muscle fibres in comparison to non-trained subjects (Thayer et al., 2000), suggesting the conversion of one fibre type to another by the training. To determine levels of activity in our subjects, each completed the International Physical Activity Questionnaire (iPAQ); (Craig et al., 2003); the scores derived from these questionnaires did not correlate with the BS test time and they had a wide range which might be indicative of the differences in type of motor units in the back muscles of our subjects. This is supported by the observation of a wide range of time-to-peak (TTP) amplitudes for the evoked twitches (see Fig. 3B). At a low contraction strength (25% MVC), where more slow twitch fibres would be recruited, the range of TTP amplitudes was 90 ms (82–172 ms) and at high contraction strength (MVC) the range was 100 ms (23–127 ms). So although the mean TTP amplitudes appear to be shorter at 100% MVC than 25% MVC, there was still wide variability within a particular contraction strength (see Fig. 3B). It is likely that this wide range of TTP amplitudes reflects the differing proportions of slow/fast twitch fibres in the back muscles amongst our subjects. It has been shown that the back extensor muscles in healthy subjects contain a higher proportion of slow-twitch fibres than fast twitch fibres (Mannion et al., 1997a) which is in keeping with their significant postural role. Furthermore, the variability within a particular fibre type in that study was not inconsiderable; it is therefore reasonable to assume that the variability in the proportions of fibre types in our cohort of healthy subjects would be comparable, but this is not possible to determine definitively without histochemical analysis of biopsies. Mean BS times have been previously shown to vary greatly between studies (Demoulin et al., 2006). The mean time of 174.9 (16.85) s found in this study is similar to Biering-Sorensen’s original study (Biering-Sorensen, 1984) and Holmstron (Holmstrom et al., 1992), but higher than other later studies (Alaranta et al., 1995; Mannion et al., 1997a,b; Latimer et al., 1999). Muscle fatigue plays an important role in endurance tests and during sub maximal contractions, such as the BS test, measurable fatigue can occur without a decrement in task performance as other motor units or

E. Bottle, P.H. Strutton / Journal of Electromyography and Kinesiology 22 (2012) 383–390

muscles are recruited to compensate for those that are fatiguing (Gandevia et al., 1996). Our study confirms fatigue occurring during the BS test, as demonstrated by a significant rise in EMG RMS and decline in median frequency (Viitasalo and Komi, 1977; Luttmann et al., 1996). Although it was clear from our results that fatigue occurs throughout the BS test, the task itself may not be representative of the type of fatigue experienced in normal daily activities. It has been shown that there are differences in the EMG variables between different types of fatiguing task, known as task dependency (Elfving and Dedering, 2007). The type of task associated with performing a BS test is not commonly found in daily living tasks. It is therefore possible that fatigue associated with normal daily activities may correlate with neural drive where BS test scores may not. It should be noted that the calculation of VA in this study was carried out once and it remains to be seen if the change in VA following the BS test correlates with these test times. Although we have found no correlation between neural drive and back muscle endurance, our group consisted only of healthy controls. There has been much research showing that people with low back pain have poor endurance (Biering-Sorensen, 1984; Nicolaisen and Jorgensen, 1985; Hultman et al., 1993; Alaranta et al., 1995), differences in the level of neuromuscular fatigability during submaximal contractions (Holmstrom et al., 1992; Mannion et al., 1997a,b; Oddsson and De Luca, 2003; Heydari et al., 2010) and lower corticospinal excitability compared to healthy controls (Strutton et al., 2005). It remains to be determined if those subjects who are at risk from future low back pain, as predicted by changes in their EMG characteristics (see (Heydari et al., 2010)) or those who have a history of low back pain (but free from pain at the time of experimentation) have reduced voluntary activation. To conclude, this study on healthy subjects has revealed that voluntary activation of the back extensor muscles, as assessed by the twitch interpolation method does not correlate with endurance time. Conflict of interest statement None. Acknowledgements The authors would like to thank all participants of this study, Arthritis Research UK for funding some of the equipment and also Imperial College London for providing the funds for this undergraduate research project. References Alaranta H, Luoto S, Heliovaara M, Hurri H. Static back endurance and the risk of low-back pain. Clin biomech 1995;10:323–4. Allen GM, Gandevia SC, McKenzie DK. Reliability of measurements of muscle strength and voluntary activation using twitch interpolation. Muscle Nerve 1995;18:593–600. Biering-Sorensen F. Physical measurements as risk indicators for low-back trouble over a one-year period. Spine 1984;9:106–19. Bigland-Ritchie B. Muscle fatigue and the influence of changing neural drive. Clin Chest Med 1984;5:21–34. Craig CL, Marshall AL, Sjostrom M, Bauman AE, Booth ML, Ainsworth BE, et al. International physical activity questionnaire: 12-country reliability and validity. Med Sci Sports Exerc 2003;35:1381–95. De Serres SJ, Enoka RM. Older adults can maximally activate the biceps brachii muscle by voluntary command. J Appl Physiol 1998;84:284–91. Demoulin C, Vanderthommen M, Duysens C, Crielaard JM. Spinal muscle evaluation using the Sorensen test: a critical appraisal of the literature. Joint, bone, spine : revue du rhumatisme 2006;73:43–50. Elfving B, Dedering A. Task dependency in back muscle fatigue–correlations between two test methods. Clin biomech 2007;22:28–33. Fulton RC, Strutton PH, McGregor AH, Davey NJ. Fatigue-induced change in corticospinal drive to back muscles in elite rowers. Exp Physiol 2002;87:593–600.

389

Gandevia SC, Allen GM, Butler JE, Taylor JL. Supraspinal factors in human muscle fatigue: evidence for suboptimal output from the motor cortex. J Physiol 1996;490(Pt 2):529–36. Gandevia SC, Allen GM, McKenzie DK. Central fatigue. Critical issues, quantification and practical implications. Adv Exp Med Biol 1995;384:281–94. Goodall S, Romer LM, Ross EZ. Voluntary activation of human knee extensors measured using transcranial magnetic stimulation. Exp Physiol 2009;94:995–1004. Hess CW, Mills KR, Murray NM. Responses in small hand muscles from magnetic stimulation of the human brain. J Physiol 1987;388:397–419. Heydari A, Nargol AV, Jones AP, Humphrey AR, Greenough CG. EMG analysis of lumbar paraspinal muscles as a predictor of the risk of low-back pain. Eur Spine J 2010;19:1145–52. Hodges PW, Richardson CA. Delayed postural contraction of transversus abdominis in low back pain associated with movement of the lower limb. J Spinal Disord 1998;11:46–56. Hodges PW, Richardson CA. Altered trunk muscle recruitment in people with low back pain with upper limb movement at different speeds. Arch Phys Med Rehabil 1999;80:1005–12. Holmstrom E, Moritz U, Andersson M. Trunk muscle strength and back muscle endurance in construction workers with and without low back disorders. Scand J Rehabil Med 1992;24:3–10. Hultman G, Nordin M, Saraste H, Ohlsen H. Body composition, endurance, strength, cross-sectional area, and density of MM erector spinae in men with and without low back pain. J Spinal Disord 1993;6:114–23. Lagan J, Lang P, Strutton PH. Measurement of voluntary activation of the back muscles using transcranial magnetic stimulation. Clin Neurophysiol 2008;119:2839–45. Latimer J, Maher CG, Refshauge K, Colaco I. The reliability and validity of the Biering-Sorensen test in asymptomatic subjects and subjects reporting current or previous nonspecific low back pain. Spine 1999;24:2085–9. Lee M, Gandevia SC, Carroll TJ. Cortical voluntary activation can be reliably measured in human wrist extensors using transcranial magnetic stimulation. Clin Neurophysiol 2008;119:1130–8. Luttmann A, Jager M, Sokeland J, Laurig W. Electromyographical study on surgeons in urology. II. Determination of muscular fatigue. Ergonomics 1996;39:298–313. Mannion AF, Dumas GA, Cooper RG, Espinosa FJ, Faris MW, Stevenson JM. Muscle fibre size and type distribution in thoracic and lumbar regions of erector spinae in healthy subjects without low back pain: normal values and sex differences. J Anat 1997a;190(Pt 4):505–13. Mannion AF, Weber BR, Dvorak J, Grob D, Muntener M. Fibre type characteristics of the lumbar paraspinal muscles in normal healthy subjects and in patients with low back pain. J orthop res 1997b;15:881–7. Merton PA. Voluntary strength and fatigue. J Physiol 1954;123:553–64. Moreau CE, Green BN, Johnson CD, Moreau SR. Isometric back extension endurance tests: a review of the literature. J Manipulative Physiol Ther 2001;24:110–22. Mork PJ, Westgaard RH. Long-term electromyographic activity in upper trapezius and low back muscles of women with moderate physical activity. J Appl Physiol 2005;99:570–8. Nicolaisen T, Jorgensen K. Trunk strength, back muscle endurance and low-back trouble. Scand J Rehabil Med 1985;17:121–7. Oddsson LI, De Luca CJ. Activation imbalances in lumbar spine muscles in the presence of chronic low back pain. J Appl Physiol 2003;94:1410–20. Perez MA, Lungholt BK, Nyborg K, Nielsen JB. Motor skill training induces changes in the excitability of the leg cortical area in healthy humans. Exp Brain Res 2004;159:197–205. Radebold A, Cholewicki J, Panjabi MM, Patel TC. Muscle response pattern to sudden trunk loading in healthy individuals and in patients with chronic low back pain. Spine 2000;25:947–54. Sidhu SK, Bentley DJ, Carroll TJ. Cortical voluntary activation of the human knee extensors can be reliably estimated using transcranial magnetic stimulation. Muscle Nerve 2009;39:186–96. Sogaard K, Gandevia SC, Todd G, Petersen NT, Taylor JL. The effect of sustained lowintensity contractions on supraspinal fatigue in human elbow flexor muscles. J Physiol 2006;573:511–23. Strutton PH, Catley M, McGregor AH, Davey NJ. Corticospinal excitability in patients with unilateral sciatica. Neurosci Lett 2003;353:33–6. Strutton PH, Theodorou S, Catley M, McGregor AH, Davey NJ. Corticospinal excitability in patients with chronic low back pain. J Spinal Disord Tech 2005;18:420–4. Taylor JL, Gandevia SC. A comparison of central aspects of fatigue in submaximal and maximal voluntary contractions. J Appl Physiol 2008;104:542–50. Thayer R, Collins J, Noble EG, Taylor AW. A decade of aerobic endurance training: histological evidence for fibre type transformation. J Sports Med Phys Fitness 2000;40:284–9. Todd G, Taylor JL, Gandevia SC. Measurement of voluntary activation of fresh and fatigued human muscles using transcranial magnetic stimulation. J Physiol 2003;551:661–71. Todd G, Taylor JL, Gandevia SC. Reproducible measurement of voluntary activation of human elbow flexors with motor cortical stimulation. J Appl Physiol 2004;97:236–42. Triscott S, Gordon J, Kuppuswamy A, King N, Davey N, Ellaway P. Differential effects of endurance and resistance training on central fatigue. J Sports Sci 2008;26:941–51.

390

E. Bottle, P.H. Strutton / Journal of Electromyography and Kinesiology 22 (2012) 383–390

Tsao H, Danneels LA, Hodges PW. Smudging the motor brain in young adults with recurrent low back pain. Spine 2011;36:1721–7. Tsao H, Galea MP, Hodges PW. Reorganization of the motor cortex is associated with postural control deficits in recurrent low back pain. Brain 2008;131:2161–71. Ugawa Y, Terao Y, Hanajima R, Sakai K, Kanazawa I. Facilitatory effect of tonic voluntary contraction on responses to motor cortex stimulation. Electroen Clin Neuro 1995;97:451–4. Viitasalo JH, Komi PV. Signal characteristics of EMG during fatigue. Eur J Appl Physiol O 1977;37:111–21. Vila-Cha C, Falla D, Farina D. Motor unit behavior during submaximal contractions following six weeks of either endurance or strength training. J Appl Physiol 2010;109:1455–66.

Emily Bottle (BSc) is a medical student at Imperial College London. As part of her BSc in Neuroscience and Mental health she undertook the current study for her dissertation.

Paul Strutton (BSc, PhD) is a senior lecturer in Neurophysiology at Imperial College London. His research interests include motor control in low back pain; exercise and central fatigue and postural control.