Journal of
Oral Rehabilitation
Journal of Oral Rehabilitation 2014 41; 477--485
Localised task-dependent motor-unit recruitment in the masseter H. J. SCHINDLER*, D. HELLMANN*, N. N. GIANNAKOPOULOS*, U. EIGLSPERG E R † , J . P . V A N D I J K † & B . G . L A P A T K I † *Department of Prosthodontics, University of Heidelberg, Heidelberg and †Department of Orthodontics, University of Ulm, Ulm, Germany
Localised motor-unit (MU) recruitment in the masseter was analysed in this study. We investigated whether differential activation behaviour, which has already been reported for distant masseter regions, can also be detected in small muscle subvolumes at the level of single MUs. Two bipolar fine-wire electrodes and an intra-oral 3D bite-force transmitter were used to record intra-muscular electromyograms (EMG) resulting from controlled bite-forces of 10 healthy human subjects (mean age 241 12 years). Twohundred and seventeen decomposed MUs were organised into localised MU task groups with different (P < 0001) force-direction-specific behaviour. Proportions of MUs involved in one, two, three or four examined tasks were 46%, 31%, SUMMARY
Introduction Motor-unit (MU) recruitment strategies in the masseter muscle are not yet fully understood. In previous studies, either motor behaviour was analysed separately for individual MUs (i.e. not for several MUs simultaneously) or the distribution of activity among distant parts of a muscle was studied. Recruitment of individual MUs in different regions of the human masseter has been examined under specific static biting conditions by means of consecutive recordings and use of needle electrodes. The regions examined were arbitrarily selected on the basis of geometrical or anatomical considerations. Selective task-dependent activation of MUs with force-direction specificity has been detected under such experimental conditions (1–5). Simultaneous intra-muscular recordings © 2014 John Wiley & Sons Ltd
18% and 5%, respectively. This study provides evidence of the ability of the neuromuscular system to modify the mechanical output of small masseter subvolumes by differential control of adjacent MUs belonging to distinct task groups. Localised differential activation behaviour of the masseter may be the crucial factor enabling highly flexible and efficient adjustment of the muscle activity in response to complex local biomechanical needs, for example, continually varying bite-forces during the demanding masticatory process. KEYWORDS: bite force, electromyograms, jaw muscles, masseter, motor units Accepted for publication 13 March 2014
obtained from distant parts of the muscle by use of wire electrodes confirmed differential (i.e. heterogeneous) activation of the masseter under static and dynamic conditions; specifically, during different motor tasks, variable distribution of activity was observed for different parts of the muscle investigated simultaneously (6–9). On the basis of the data available for humans, and supported by animal studies (10, 11), it is supposed that differential activation of the masseter – a phenomenon also described for muscles in other body regions (12) – is reflected mainly by the distinguishable anatomical compartments and the different neuronal supply of the corresponding muscle subdivisions (11–13). A new theory predicting redistribution of activity within an affected muscle at the level of MUs, to protect the muscle from further pain or injury, has doi: 10.1111/joor.12168
478
H . J . S C H I N D L E R et al. recently been proposed (14). This theory would imply, for the masseter with its small MU territories (15), that MU recruitment patterns would have to be reorganised on a local level to relieve discrete injured regions of the muscle. Recent findings in an experimental study of pain (16) supported this concept. The theory of localised differential control of single MUs not only provides a model explaining protective adaptation to pain, but might also appropriately describe the control strategy of the intact jaw muscle, in particular because demanding kinetic tasks during mastication with continually varying bite-force directions require subtle MU control. A prerequisite for this implication should be that taskdependent redistribution of localised MU activity is, in fact, the motor control concept for the masseter. Answering this question, however, requires analysis of populations of MUs in small masseter subvolumes. To date, differential activation behaviour has been proved for distant parts of the masseter, by evaluation of electromyographic (EMG) interference signals; for single MUs, it has been proved by examination of one MU, only, at a time. The purpose of this investigation was to study the recruitment behaviour of several MUs simultaneously in a small subvolume of muscle under conditions of constant bite forces with changing directions. We hypothesised that the differential activation behaviour might also be reflected in small masseter subvolumes by localised differential MU recruitment. Another objective of our study was to obtain an approximate quantitative estimate of which proportion of MUs is task-specific and which proportion of MUs remains active during different tasks (so called ‘polymodal’ MUs).
Material and methods Subjects Ten healthy subjects (five women, five men; mean age 241 12 years) were enrolled in the study. All subjects were naturally dentate, without pain or dysfunction. Clinical assessment was conducted by use of the research diagnostic criteria for temporomandibular disorders (17). The study was approved by the Ethics Committee of the University Medical Centre, Heidelberg (S-213/2008). All participating subjects provided written consent to the experiments. The experimental procedures were conducted in accordance with the declaration of Helsinki. Bite-force measurement Bite-force was transmitted by means of an intra-oral device (Fig. 1a) consisting of an upper metal splint with a “bearing pin” equipped with strain gauges and a lower metal splint with a contact plate mounted midway between the lower first molars and parallel to the occlusal plane of the mandible (18). A perforation in the contact plate enabled connection with the bearing pin for force transmission in centric jaw relation. Jaw separation at the incisors was adjusted to 5 mm. The transducer measured forces in three orthogonal directions with reference to the occlusal and mid-sagittal planes. The bite-force vectors ! ! F res; maxilla ¼ F res; mand were described by the horizontal–circumferential angle φ, the vertical–sagittal angle h (its tangent is determined by the relationship between the vertical and horizontal force components), and the resultant force magnitude (i.e. the
(b)
(a)
Transducer
z Metal splints
x
Insertion needle
y
Fig. 1. (a) Intra-oral force-measurement device. (b) The stereotactic instrumentation used for placement of the fine-wire electrodes in pre-determined masseter subvolumes. © 2014 John Wiley & Sons Ltd
MOTOR-UNIT RECRUITMENT IN THE MASSETER
Vert
Tasks 3D force vector
Pro
80 s
Lat
Med
Analysis period (5 s)
20 N Raw EMG 0·5 mV
Raw EMG
Fig. 2. Force and EMG signals for one subject and masseter region. The upper three traces show the complete force and EMG recordings (total 80 s) obtained during performance of the four motor tasks, i.e. vertical (Vert), protrusive (Pro), lateral (Lat) and medial (Med) bite-force directions. In the lower trace, the extended 5 s EMG data window selected for decomposition into single MUs is shown for the protrusive task. Window selection was based on the maximum stability of the 3D force vector. Its magnitude was kept constant while its direction was altered for the different motor tasks. Data windows selected for the other three tasks are indicated by horizontal bars above the force signal.
vector length). These ‘target variables’ were displayed on a PC monitor as feedback for the subject and experimenter (19). Desired values for the target variables could be marked on the desktop of the force-acquisition software. The force signals were sampled at 1000 Hz and recorded synchronously with the EMG signal. Masseter EMG recordings Two bipolar fine-wire electrode pairs (9) with 15 mm of wire exposed at the ends and an interelectrode distance of 15 mm were used to record EMG signals from deep and superficial regions of the right masseter. To enable recording from approximately the same regions of the muscles of different subjects, we initially marked the intersection of two lines on the right cheek. These diagonally arranged lines were constructed by connecting opposing corners of the tetragon formed by the anterior and posterior borders of the masseter, the lower edge of the zygomatic arch and the lower mandibular border. The point of intersection was the target point for insertion of the wires. The target recording depths below the surface of the muscle were 18 2 mm and 9 1 mm, respectively. To achieve this, the depth of penetration of the two fine-wire electrode pairs below the skin surface had to be individually inferred for each subject from ultrasound images of the slightly contracted masseter taken before the study. Precise insertion of the two intra-muscular fine-wire electrode pairs into the muscle, i.e. parallel guidance of the carrier needles with a © 2014 John Wiley & Sons Ltd
defined cranio–caudal distance of 5 mm and a predefined difference of 9 mm between their depths of penetration of the muscle, was ensured by use of a stereotactic device specially developed for this study (Fig. 1b). The signals from the wire electrode pairs were differentially amplified (Dantec Keypoint*), filtered (bandwidth 20 Hz–10 kHz) and digitised at a sampling rate of 20 kHz. Procedure After electrode placement, the intra-oral force transducer was mounted on both jaws. Four motor tasks involving different bite force vectors in different directions exerted in centric jaw relation (Fig. 2) were then performed. The first vector was purely vertical (h = 0°); starting from an initial value of 20N. Force magnitude was then varied (with the pure vertical direction maintained) until the maximum number of distinguishable MUs could be observed in the signal trains on the EMG monitor. Distinguishable means that large overlap of waveforms of subsequent action potentials of different MUs occurred only exceptionally. Using the optimum 3D force vector magnitude for the vertical task as target criterion, simultaneous bite force and EMG recordings were also performed during application of oblique bite forces with different angles φ
*Natus Medical, San Carlos, CA, USA.
479
480
H . J . S C H I N D L E R et al. (anteriorly = 0°, anteriorly/left 60°, anteriorly/right 300°), but a constant value of the angle h (60°). Data were recorded for 20 s for each of the four bite-force vectors. To avoid muscle fatigue, individual recordings were separated by 1 min. Data processing and analysis As a first step, for each subject and for each of the 20 s task-specific recordings, a 5 s period with the least force variability was selected (Fig. 2). The root mean square (RMS) of the two bipolar interference signals was calculated separately for the 5 s periods, to estimate muscle-activity differences for distinct motor tasks at the level of localised muscle areas. Next, the two-channel interference EMG from the four selected 5 s periods were decomposed into single MU action potential trains by use of the validated semiautomatic open-source software EMGLAB (20). This software enables full decomposition of single-channel or multichannel interference EMG signals recorded by intra-muscular electrodes into the waveforms of the individual MUs. Because it contains advanced algorithms for template matching and for resolving superimposition of interfering activity from several simultaneously active MUs, more complex signals recorded during moderate levels of muscle activity can also be decomposed. A user interface enables inspection and editing of the signals manually to complete the decomposition and verify the results. Because we concatenated the selected 5 s periods for each subject into one 20 s file and decomposed the merged files in one run, correct assignment of decomposed action potentials to specific MUs over different tasks was ensured. Three experienced experimenters performed the decomposition procedure independently, and only MUs found by all three investigators were used for further analysis. The recruitment behaviour of masseter MUs was characterised in three ways: 1 to analyse which task combinations are managed by the same MUs, we introduced a so-called ‘coincidence index’ (CI), which enables quantitative comparison of the different recruitment patterns for two different motor tasks i and j, as follows: CI ¼ sqrtððnActMUtask i and j =nActMUtask i Þ ðnActMUtask i and j =nActMUtask j ÞÞ 100%
for example, when five MUs were active in a task i (e.g. vertical bite force) and four in task j (e.g. protrusive bite force), and two MUs were active in both tasks, CI = sqrt (2/5*2/4)*100 = 45%. Thus, the CI for a complete match of MUs between two tasks would be 100%, whereas that for no match would be 0%. Because we observed no active MUs for several tasks, inadmissible division by zero would have occurred. In these cases, the CI was defined as zero; 2 we determined the number of tasks in which each MU was recruited. Thus, we obtained the number and proportion of MUs involved in one, two, three or four tasks (unimodal, bimodal, trimodal and quadrimodal); 3 the overall number of MUs activated during each task was evaluated. Deviation of the actually measured bite-force vectors from the corresponding target values was characterised by use of the coefficient of variation (cv) for the variability during the 5 s tasks averaged over all tasks (cv1) and for the variability among the four tasks (cv2). Differences between the RMS values and the CI values were evaluated separately for the superficial and deep recordings by one-way repeated-measures ANOVA. Differences between the number of active MUs during the four tasks and the task-specificity of the MUs were analysed between and within the superficial and deep masseter by two-way repeatedmeasures ANOVA.
Results The bite-forces (mean, SD and cv) exerted by the individual subjects during the different tasks are shown in Table 1. Obviously, the 3D force vector necessary to activate the maximum number of distinguishable MUs was relatively constant during all motor tasks for individual subjects. However, force magnitudes varied substantially among individuals. The distinct motor tasks were associated with significantly different (F = 397, df = 3; P < 005) local muscle activity. The corresponding RMS values for the deep and superficial interference EMG signals are illustrated in Fig. 3. In total, 217 single MUs were decomposed from the interference signals from the 10 subjects (Table 1). A typical MU recruitment scheme during the different motor tasks is shown in Fig. 4 for one representative
© 2014 John Wiley & Sons Ltd
MOTOR-UNIT RECRUITMENT IN THE MASSETER Table 1. Bite-force magnitudes (intra-individual, averaged over the four tasks) exerted by the individual subjects to achieve the maximum number of decomposable MUs; cv1 = mean coefficient of variation (SD/mean 9 100) describing variability during the 5 s recording periods, cv2 = mean coefficient of variation describing variability among the four tasks. The two columns on the right indicate the number of MUs identified in the individual subjects Motor units per subject
Force vector Variability (%)
Masseter
Subject
Newton
cv1
cv2
Deep
Superficial
1 2 3 4 5 6 7 8 9 10 Mean SD Σ
22 41 16 17 37 26 16 9 43 17 244 119
59 49 69 35 149 115 63 33 09 112 69 43
63 104 61 62 144 93 66 57 143 165 96 41
12 9 10 10 15 15 18 5 12 10 116 39 116
8 5 8 8 10 13 17 9 16 7 101 40 101
subject. Figure 5 shows the numbers of active MUs during the different tasks for all the individual subjects. The 116 MUs detected in the deep part of the masseter were recruited 229 times during the four motor tasks. The corresponding involvement of the 101 MUs detected in the superficial masseter, 164 recruitments, was significantly smaller (F = 774, df = 1; P < 002). Within both parts, significant differ-
ences in the number of recruited MUs were found between protrusive (86 deep; 67 superficial) and lateral (31 deep; 17 superficial) motor tasks (F = 718, df = 3; P < 0001). The highest mean CI values (i.e. 69% and 52% for the deep and superficial masseter, respectively) were found for paired vertical and protrusive motor tasks. Mean CI values for other task combinations were much lower, with a minimum for the combination of vertical and lateral force directions (Table 2A, B). Mean CI values for the six different pairs of tasks were found to be significantly different (F = 3012, df = 3; P < 0001) within both parts of the masseter. In addition, the CI values were significantly (F = 2058, df = 1; P < 0001) higher for the deep masseter. Quantitative estimates of the task-specificity of masseter MUs are given in Table 2C; these indicate that approximately 46% of the detected MUs were unimodal, 31% bimodal, 18% trimodal and 5% quadrimodal. Statistical analysis revealed that these proportions were significantly different, both within (F = 739, df = 3; P < 0001) and between (F = 827, df = 1; P < 002) the deep and superficial parts of the masseter.
Discussion In this study, task-dependent MU recruitment in the masseter was examined. Stationary bipolar fine-wire electrodes with a short interelectrode distance were used and, therefore, myoelectric activity from relatively small muscle subvolumes only was recorded. Application of advanced decomposition software enabled isola-
140 120
RMS [ V]
100 Superficial wire
80
Deep wire
60 40 20 0
Vert
Pro
Lat
Med
Motor task Fig. 3. Localised integrated EMG activity (RMS) of the superficial and deep masseter for the different motor tasks. Asterisks indicate significant differences between the tasks (P < 005). © 2014 John Wiley & Sons Ltd
481
482
H . J . S C H I N D L E R et al. (a) Arbitrary units 200
48 6
37 1
2
4
6
3
8 9
1
2 4
100 0 –100 –200 60·82
60·84
60·86
60·88
60·90
60·92
Motor task (b)
Pro
Vert
Lat
60·94 s
Med
MU number 1 2 3 4 5 6 7 8 9 Selected 5 s time frames
(c) MU number
Pro
Vert
Lat
Med
1 2 3 4 5 6 7 8 9 0
10
20
30
40
50
60
70
80
s
Fig. 4. EMG decomposition and analysis of recruitment behaviour for a representative subject. (a) The signal peaks and corresponding numbers in the interference signal shown in the upper trace (01 s recording in the deep masseter during lateral bite-force) indicate firing events of eight (MU 1, 2, 3, 4, 6, 7, 8, 9) of the nine decomposed MUs active in the four tasks shown below. The lower trace is the remaining interference signal after subtraction of the corresponding templates. The fact that this signal is, almost, baseline noise only is indicative of successful decomposition. (b) For demonstration purposes, the nine decomposed MU action potentials were averaged in this example over the separate 5 s time periods (instead of averaging them over the combined 20 s signal window). The finding of very small variation of action potential shapes of individual MUs in the different motor tasks shows that recruitment and/or derecruitment of identical MUs in the different motor tasks could be unambiguously identified. (c) Analysis of the recruitment behaviour. The black bars indicate which MU is active in which task, that is, in the vertical (Vert), protrusive (Pro), lateral (Lat) and medial (Med) bite-force directions. For this subject, three of the MUs are unimodal (MU5 for the vertical force direction, MU7 and MU8 for the lateral force direction), two MUs are bimodal (MU1 for vertical plus lateral tasks, MU9 for protrusive plus lateral tasks), three MUs are trimodal (MU2 and MU3 for vertical plus protrusive plus lateral tasks, and MU6 for protrusive plus lateral plus medial tasks), and MU4 is quadrimodal.
© 2014 John Wiley & Sons Ltd
MOTOR-UNIT RECRUITMENT IN THE MASSETER Deep recording
16
Superficial recording
Number of active MUs
14 12 10 8 6 4 2 0 Vert
Pro
Lat
Med
Vert
Motor task
Pro
Lat
Med
Motor task
Fig. 5. Numbers of MUs active in the superficial and deep masseter in the four different tasks for all individual subjects (n = 10).
tion of up to 10 MUs from the recording for a single motor task and up to 18 MUs from one concatenated EMG file containing the signals for all four motor tasks. By this means it was revealed (as far as we are aware, for the first time) that the heterogeneous activation behaviour of the masseter manifests itself even in small muscle subvolumes as a very localised phenomenon at the level of its smallest functional units. These findings confirm the initially stated hypothesis and add substan-
tially to those of previous reports (4, 5) in which taskspecificity was separately proved for individual MUs with locations in distinct muscle regions only. These results are, moreover, also supported by the local muscle activity differences identified at the interference EMG level. Quantification of the coincidence of MU recruitment for pairs of different motor tasks (i.e. defined bite-force vectors of the same length in four different
Table 2. Mean values of coincidence indexes (CI, %) with standard deviations (in parentheses). The CI correlates positively with the proportions of MUs recruited in both tasks of a specific task combination. CI values were averaged over all subjects (n = 10); Vert, vertical; Pro, protrusive; Lat, lateral; Med, medial bite-force directions. (A) CI values for the deep masseter. (B) CI values for the superficial masseter; N indicates the number of MUs identified as being involved in the task of the corresponding row. (C) Summary of the number of MUs active in one, two, three, or four tasks (A) CI values for the deep masseter Task
Vert
Pro
Lat
Med
N
Vert Pro Lat Med
100% – – –
69% (17%) 100% – –
23% (31%) 28% (25%) 100% –
36% (19%) 40% (24%) 30% (35%) 100%
59 86 31 55
(B) CI values for the superficial masseter Task
Vert
Pro
Lat
Med
N
Vert Pro Lat Med
100% – – –
52% (33%) 100% – –
7% (11%) 17% (28%) 100% –
11%(21%) 31% (18%) 10% (26%) 100%
44 67 17 32
(C) MU numbers and proportions for the different grades of task specificity Task per MU
Unimodal
Bimodal
Trimodal
Quadrimodal
N
Deep MUs Superficial MUs Total Proportion
47 (41%) 54 (53%) 101 46%
33 (28%) 34 (34%) 67 31%
28 (24%) 10 (10%) 38 18%
8 (7%) 3 (3%) 11 5%
116 101 217
© 2014 John Wiley & Sons Ltd
483
484
H . J . S C H I N D L E R et al. directions) revealed force-direction-specific recruitment for a specific proportion of the MUs only. The occurrence of task-specific and polymodal MUs in the masseter confirms results from a previous study (2). The exclusive finding of polymodal MUs in another study (4) might be because of the biomechanical similarity of some of the tasks examined. Because of the simultaneous observation of multiple MUs in this study we could subdivide the MUs into task-specific and polymodal types. In this manner, it was possible to approximately estimate which model of functional MU behaviour is dominant in the masseter. The variability of the CI between the task combinations and between the deep and superficial parts of the muscle reflects the distinct biomechanical contributions of highly direction-specific MUs in the different bite-force directions and in both parts of the masseter. Systematic characterisation of MUs into specific and polymodal types revealed that most of the MUs were single-task-specific (46%), followed by bimodal (31%), trimodal (18°) and quadrimodal (5%) MU recruitment behaviour. The polymodal behaviour might be related to biomechanical requirements. Indeed, recent anatomical studies using magnetic resonance imaging revealed highly variable inner muscle structure consisting of multiple small aponeuroses with different spatial orientations and dense interconnections (21). Such a complex structure requires stabilizing elements between the aponeuroses, especially during differential activation of the muscle. In principle, force transmission by MUs must be integrated into task-specific lines of action in the active muscle. Biomechanical interactions between single muscle fibres, whole fibre bundles, and aponeuroses in the complex muscle are incompletely understood. On the basis of the knowledge available (22, 23), however, it might be speculated that the polymodal MUs of the masseter provide, at least, supplementary force for internal muscle stabilisation, maintaining the basic muscle shape during function, and stiffening the complex tendon system to ensure task-related directional force application by other MUs. In this context, the different overall MU activation between and within both parts of the muscle might be indicative of different stabilisation needs of superficial and deep parts of the masseter and for distinct tasks. Polymodal MUs have been also found in other muscles, for example the trapezius muscle (24). Interestingly, these motor units, which are not task-specific, are most likely to
be type I motor units, sometime also described as ‘Cinderella’ motor units (25). When interpreting the quantitative results of this study, it must be taken into consideration that co-contraction, a well-known phenomenon during execution of untrained motor tasks (18), might partially explain the activity of task-unspecific MUs. Consequently, the extent of polymodality might have been somewhat overestimated in our study with untrained subjects. Moreover, because the results were obtained from two specific masseter subvolumes only, the suggested recruitment behaviour cannot be automatically assumed for other regions or for the whole muscle. This limitation also applies to the individual force values and their effect on recruitment modalities. It might, however, be assumed that the characteristics of neuromuscular behaviour in other muscle regions, and possibly also at distinct force levels, are, in principle, similar.
Conclusions Our findings for healthy subjects suggest the capability of the neuromuscular system to modify the mechanical output of small masseter subvolumes by differential control of adjacent MUs belonging to distinct task groups. Such a subtle control mechanism, supported by discrete MU territories and different lines of action of MUs, might enable the muscle to adjust its activity in a highly flexible and efficient manner in response to complex local biomechanical needs, for example, continually varying bite-forces during a demanding masticatory process. It might also be the neuromuscular basis of motor adaptations to pain, as has already been reported for the masseter (16).
Acknowledgments The authors wish to express their gratitude to Mr Willi Wendler (University of Karlsruhe) for his invaluable contribution to the design and fabrication of diverse mechanical components for the experimental setup and Dr. med. dent. Oliver Thiele (University of Heidelberg) for his assistance with the ultrasound imaging of the masseter.
Conflict of interests The authors received no financial support and declare no potential conflict of interests in respect of patent © 2014 John Wiley & Sons Ltd
MOTOR-UNIT RECRUITMENT IN THE MASSETER rights, corporate affiliations or consultancies for any product or process mentioned in the article. 14.
References 15. 1. Yemm R. The orderly recruitment of motor units of the masseter and temporal muscles during voluntary isometric contraction in man. J Physiol. 1977;265:163–174. 2. Eriksson PO, Stalberg E, Antoni L. Flexibility in motor-unit firing pattern in the human temporal and masseter muscles related to type of activation and location. Arch Oral Biol. 1984;29:707–712. 3. Tonndorf ML, Sasaki K, Hannam AG. Single-wire recording of regional activity in the human masseter muscle. Brain Res Bull. 1989;23:155–159. 4. McMillan AS, Hannam AG. Task-related behavior of motor units in different regions of the human masseter muscle. Arch Oral Biol. 1992;37:849–857. 5. Ogawa T, Kawata T, Tsuboi A, Hattori Y, Watanabe M, Sasaki K. Functional properties and regional differences of human masseter motor units related to three-dimensional bite force. J Oral Rehabil. 2006;33:729–740. 6. Blanksma NG, Van Eijden TM, Weijs WA. Electromyographic heterogeneity in the human masseter muscle. J Dent Res. 1992;71:47–52. 7. Blanksma NG, van Eijden TM, van Ruijven LJ, Weijs WA. Electromyographic heterogeneity in the human temporalis and masseter muscles during dynamic tasks guided by visual feedback. J Dent Res. 1997;76:542–551. 8. Schindler HJ, T€ urp JC, Blaser R, Lenz J. Differential activity patterns in the masseter muscle under simulated clenching and grinding forces. J Oral Rehabil. 2005;32:552–563. 9. T€ urp JC, Schindler HJ, Pritsch M, Rong Q. Antero-posterior activity changes in the superficial masseter muscle after exposure to experimental pain. Eur J Oral Sci. 2002;110:83– 91. 10. Widmer CG, Klugman D, English AW. Anatomical partitioning and nerve branching patterns in the adult rabbit masseter. Acta Anat (Basel). 1997;159:222–232. 11. Herring SW, Wineski LE, Anapol FC. Neural organization of the masseter muscle in the pig. J Comp Neurol. 1989;280:563–576. 12. Windhorst U, Hamm TM, Stuart DG. On the function of muscle and reflex partitioning. Behav Brain Sci. 1989;12:629–681. 13. Widmer CG, Carrasco DI, English AW. Differential activation of neuromuscular compartments in the rabbit masseter
© 2014 John Wiley & Sons Ltd
16.
17.
18.
19.
20.
21.
22. 23.
24.
25.
muscle during different oral behaviors. Exp Brain Res. 2003;150:297–307. Hodges PW, Tucker K. Moving differently in pain: a new theory to explain the adaptation to pain. Pain. 2011;152(3 Suppl):S90–S98. Tonndorf ML, Hannam AG. Motor unit territory in relation to tendons in the human masseter muscle. Muscle Nerve. 1994;17:436–443. Minami I, Akhter R, Albersen I, Burger C, Whittle T, Lobbezoo F et al. Masseter motor unit recruitment is altered in experimental jaw muscle pain. J Dent Res. 2013;92: 143–148. Dworkin S, LeResche L. Research diagnostic criteria for temporomandibular disorders: review, criteria, examinations and specifications, critique. J Craniomandib Disord Facial Oral Pain. 1992;6:301–355. Schindler HJ, Rues S, T€ urp JC, Schweizerhof K, Lenz J. Jaw clenching: muscle and joint forces, optimization strategies. J Dent Res. 2007;86:843–847. van Eijden TM, Koolstra JH, Brugman P, Weijs WA. A feedback method to determine the three-dimensional bite-force capabilities of the human masticatory system. J Dent Res. 1988;67:450–454. McGill KC, Lateva ZC, Marateb HR. EMGLAB: an interactive EMG decomposition program. J Neurosci Methods. 2005;149:121–133. Cioffi I, Gallo LM, Palla S, Erni S, Farella M. Macroscopic analysis of human masseter compartments assessed by magnetic resonance imaging. Cells Tissues Organs. 2012;195:465–472. Monti RJ, Roy RR, Edgerton VR. Role of motor unit structure in defining function. Muscle Nerve. 2001;24:848–866. Huijing PA. Epimuscular myofascial force transmission: a historical review and implications for new research. International Society of Biomechanics Muybridge Award Lecture, Taipei, 2007. J Biomech. 2009;42:9–21. Kadefors R, Forsman M, Zoega B, Herberts P. Recruitment of low threshold motor-units in the trapezius muscle in different static arm positions. Ergonomics. 1999;42:359–375. H€agg GM. Static work and myalgia - a new explanation model. In: Andersson PA, Hobart DJ, Danoff JV, eds. Electromyographical kinesiology. Amsterdam: Elsevier Science; 1991:141–144.
Correspondence: Daniel Hellmann, Department of Prosthodontics, University of Heidelberg, Im Neuenheimer Feld 400, 69120 Heidelberg, Germany. E-mail: [email protected]
485
Oral Rehabilitation
Journal of Oral Rehabilitation 2014 41; 477--485
Localised task-dependent motor-unit recruitment in the masseter H. J. SCHINDLER*, D. HELLMANN*, N. N. GIANNAKOPOULOS*, U. EIGLSPERG E R † , J . P . V A N D I J K † & B . G . L A P A T K I † *Department of Prosthodontics, University of Heidelberg, Heidelberg and †Department of Orthodontics, University of Ulm, Ulm, Germany
Localised motor-unit (MU) recruitment in the masseter was analysed in this study. We investigated whether differential activation behaviour, which has already been reported for distant masseter regions, can also be detected in small muscle subvolumes at the level of single MUs. Two bipolar fine-wire electrodes and an intra-oral 3D bite-force transmitter were used to record intra-muscular electromyograms (EMG) resulting from controlled bite-forces of 10 healthy human subjects (mean age 241 12 years). Twohundred and seventeen decomposed MUs were organised into localised MU task groups with different (P < 0001) force-direction-specific behaviour. Proportions of MUs involved in one, two, three or four examined tasks were 46%, 31%, SUMMARY
Introduction Motor-unit (MU) recruitment strategies in the masseter muscle are not yet fully understood. In previous studies, either motor behaviour was analysed separately for individual MUs (i.e. not for several MUs simultaneously) or the distribution of activity among distant parts of a muscle was studied. Recruitment of individual MUs in different regions of the human masseter has been examined under specific static biting conditions by means of consecutive recordings and use of needle electrodes. The regions examined were arbitrarily selected on the basis of geometrical or anatomical considerations. Selective task-dependent activation of MUs with force-direction specificity has been detected under such experimental conditions (1–5). Simultaneous intra-muscular recordings © 2014 John Wiley & Sons Ltd
18% and 5%, respectively. This study provides evidence of the ability of the neuromuscular system to modify the mechanical output of small masseter subvolumes by differential control of adjacent MUs belonging to distinct task groups. Localised differential activation behaviour of the masseter may be the crucial factor enabling highly flexible and efficient adjustment of the muscle activity in response to complex local biomechanical needs, for example, continually varying bite-forces during the demanding masticatory process. KEYWORDS: bite force, electromyograms, jaw muscles, masseter, motor units Accepted for publication 13 March 2014
obtained from distant parts of the muscle by use of wire electrodes confirmed differential (i.e. heterogeneous) activation of the masseter under static and dynamic conditions; specifically, during different motor tasks, variable distribution of activity was observed for different parts of the muscle investigated simultaneously (6–9). On the basis of the data available for humans, and supported by animal studies (10, 11), it is supposed that differential activation of the masseter – a phenomenon also described for muscles in other body regions (12) – is reflected mainly by the distinguishable anatomical compartments and the different neuronal supply of the corresponding muscle subdivisions (11–13). A new theory predicting redistribution of activity within an affected muscle at the level of MUs, to protect the muscle from further pain or injury, has doi: 10.1111/joor.12168
478
H . J . S C H I N D L E R et al. recently been proposed (14). This theory would imply, for the masseter with its small MU territories (15), that MU recruitment patterns would have to be reorganised on a local level to relieve discrete injured regions of the muscle. Recent findings in an experimental study of pain (16) supported this concept. The theory of localised differential control of single MUs not only provides a model explaining protective adaptation to pain, but might also appropriately describe the control strategy of the intact jaw muscle, in particular because demanding kinetic tasks during mastication with continually varying bite-force directions require subtle MU control. A prerequisite for this implication should be that taskdependent redistribution of localised MU activity is, in fact, the motor control concept for the masseter. Answering this question, however, requires analysis of populations of MUs in small masseter subvolumes. To date, differential activation behaviour has been proved for distant parts of the masseter, by evaluation of electromyographic (EMG) interference signals; for single MUs, it has been proved by examination of one MU, only, at a time. The purpose of this investigation was to study the recruitment behaviour of several MUs simultaneously in a small subvolume of muscle under conditions of constant bite forces with changing directions. We hypothesised that the differential activation behaviour might also be reflected in small masseter subvolumes by localised differential MU recruitment. Another objective of our study was to obtain an approximate quantitative estimate of which proportion of MUs is task-specific and which proportion of MUs remains active during different tasks (so called ‘polymodal’ MUs).
Material and methods Subjects Ten healthy subjects (five women, five men; mean age 241 12 years) were enrolled in the study. All subjects were naturally dentate, without pain or dysfunction. Clinical assessment was conducted by use of the research diagnostic criteria for temporomandibular disorders (17). The study was approved by the Ethics Committee of the University Medical Centre, Heidelberg (S-213/2008). All participating subjects provided written consent to the experiments. The experimental procedures were conducted in accordance with the declaration of Helsinki. Bite-force measurement Bite-force was transmitted by means of an intra-oral device (Fig. 1a) consisting of an upper metal splint with a “bearing pin” equipped with strain gauges and a lower metal splint with a contact plate mounted midway between the lower first molars and parallel to the occlusal plane of the mandible (18). A perforation in the contact plate enabled connection with the bearing pin for force transmission in centric jaw relation. Jaw separation at the incisors was adjusted to 5 mm. The transducer measured forces in three orthogonal directions with reference to the occlusal and mid-sagittal planes. The bite-force vectors ! ! F res; maxilla ¼ F res; mand were described by the horizontal–circumferential angle φ, the vertical–sagittal angle h (its tangent is determined by the relationship between the vertical and horizontal force components), and the resultant force magnitude (i.e. the
(b)
(a)
Transducer
z Metal splints
x
Insertion needle
y
Fig. 1. (a) Intra-oral force-measurement device. (b) The stereotactic instrumentation used for placement of the fine-wire electrodes in pre-determined masseter subvolumes. © 2014 John Wiley & Sons Ltd
MOTOR-UNIT RECRUITMENT IN THE MASSETER
Vert
Tasks 3D force vector
Pro
80 s
Lat
Med
Analysis period (5 s)
20 N Raw EMG 0·5 mV
Raw EMG
Fig. 2. Force and EMG signals for one subject and masseter region. The upper three traces show the complete force and EMG recordings (total 80 s) obtained during performance of the four motor tasks, i.e. vertical (Vert), protrusive (Pro), lateral (Lat) and medial (Med) bite-force directions. In the lower trace, the extended 5 s EMG data window selected for decomposition into single MUs is shown for the protrusive task. Window selection was based on the maximum stability of the 3D force vector. Its magnitude was kept constant while its direction was altered for the different motor tasks. Data windows selected for the other three tasks are indicated by horizontal bars above the force signal.
vector length). These ‘target variables’ were displayed on a PC monitor as feedback for the subject and experimenter (19). Desired values for the target variables could be marked on the desktop of the force-acquisition software. The force signals were sampled at 1000 Hz and recorded synchronously with the EMG signal. Masseter EMG recordings Two bipolar fine-wire electrode pairs (9) with 15 mm of wire exposed at the ends and an interelectrode distance of 15 mm were used to record EMG signals from deep and superficial regions of the right masseter. To enable recording from approximately the same regions of the muscles of different subjects, we initially marked the intersection of two lines on the right cheek. These diagonally arranged lines were constructed by connecting opposing corners of the tetragon formed by the anterior and posterior borders of the masseter, the lower edge of the zygomatic arch and the lower mandibular border. The point of intersection was the target point for insertion of the wires. The target recording depths below the surface of the muscle were 18 2 mm and 9 1 mm, respectively. To achieve this, the depth of penetration of the two fine-wire electrode pairs below the skin surface had to be individually inferred for each subject from ultrasound images of the slightly contracted masseter taken before the study. Precise insertion of the two intra-muscular fine-wire electrode pairs into the muscle, i.e. parallel guidance of the carrier needles with a © 2014 John Wiley & Sons Ltd
defined cranio–caudal distance of 5 mm and a predefined difference of 9 mm between their depths of penetration of the muscle, was ensured by use of a stereotactic device specially developed for this study (Fig. 1b). The signals from the wire electrode pairs were differentially amplified (Dantec Keypoint*), filtered (bandwidth 20 Hz–10 kHz) and digitised at a sampling rate of 20 kHz. Procedure After electrode placement, the intra-oral force transducer was mounted on both jaws. Four motor tasks involving different bite force vectors in different directions exerted in centric jaw relation (Fig. 2) were then performed. The first vector was purely vertical (h = 0°); starting from an initial value of 20N. Force magnitude was then varied (with the pure vertical direction maintained) until the maximum number of distinguishable MUs could be observed in the signal trains on the EMG monitor. Distinguishable means that large overlap of waveforms of subsequent action potentials of different MUs occurred only exceptionally. Using the optimum 3D force vector magnitude for the vertical task as target criterion, simultaneous bite force and EMG recordings were also performed during application of oblique bite forces with different angles φ
*Natus Medical, San Carlos, CA, USA.
479
480
H . J . S C H I N D L E R et al. (anteriorly = 0°, anteriorly/left 60°, anteriorly/right 300°), but a constant value of the angle h (60°). Data were recorded for 20 s for each of the four bite-force vectors. To avoid muscle fatigue, individual recordings were separated by 1 min. Data processing and analysis As a first step, for each subject and for each of the 20 s task-specific recordings, a 5 s period with the least force variability was selected (Fig. 2). The root mean square (RMS) of the two bipolar interference signals was calculated separately for the 5 s periods, to estimate muscle-activity differences for distinct motor tasks at the level of localised muscle areas. Next, the two-channel interference EMG from the four selected 5 s periods were decomposed into single MU action potential trains by use of the validated semiautomatic open-source software EMGLAB (20). This software enables full decomposition of single-channel or multichannel interference EMG signals recorded by intra-muscular electrodes into the waveforms of the individual MUs. Because it contains advanced algorithms for template matching and for resolving superimposition of interfering activity from several simultaneously active MUs, more complex signals recorded during moderate levels of muscle activity can also be decomposed. A user interface enables inspection and editing of the signals manually to complete the decomposition and verify the results. Because we concatenated the selected 5 s periods for each subject into one 20 s file and decomposed the merged files in one run, correct assignment of decomposed action potentials to specific MUs over different tasks was ensured. Three experienced experimenters performed the decomposition procedure independently, and only MUs found by all three investigators were used for further analysis. The recruitment behaviour of masseter MUs was characterised in three ways: 1 to analyse which task combinations are managed by the same MUs, we introduced a so-called ‘coincidence index’ (CI), which enables quantitative comparison of the different recruitment patterns for two different motor tasks i and j, as follows: CI ¼ sqrtððnActMUtask i and j =nActMUtask i Þ ðnActMUtask i and j =nActMUtask j ÞÞ 100%
for example, when five MUs were active in a task i (e.g. vertical bite force) and four in task j (e.g. protrusive bite force), and two MUs were active in both tasks, CI = sqrt (2/5*2/4)*100 = 45%. Thus, the CI for a complete match of MUs between two tasks would be 100%, whereas that for no match would be 0%. Because we observed no active MUs for several tasks, inadmissible division by zero would have occurred. In these cases, the CI was defined as zero; 2 we determined the number of tasks in which each MU was recruited. Thus, we obtained the number and proportion of MUs involved in one, two, three or four tasks (unimodal, bimodal, trimodal and quadrimodal); 3 the overall number of MUs activated during each task was evaluated. Deviation of the actually measured bite-force vectors from the corresponding target values was characterised by use of the coefficient of variation (cv) for the variability during the 5 s tasks averaged over all tasks (cv1) and for the variability among the four tasks (cv2). Differences between the RMS values and the CI values were evaluated separately for the superficial and deep recordings by one-way repeated-measures ANOVA. Differences between the number of active MUs during the four tasks and the task-specificity of the MUs were analysed between and within the superficial and deep masseter by two-way repeatedmeasures ANOVA.
Results The bite-forces (mean, SD and cv) exerted by the individual subjects during the different tasks are shown in Table 1. Obviously, the 3D force vector necessary to activate the maximum number of distinguishable MUs was relatively constant during all motor tasks for individual subjects. However, force magnitudes varied substantially among individuals. The distinct motor tasks were associated with significantly different (F = 397, df = 3; P < 005) local muscle activity. The corresponding RMS values for the deep and superficial interference EMG signals are illustrated in Fig. 3. In total, 217 single MUs were decomposed from the interference signals from the 10 subjects (Table 1). A typical MU recruitment scheme during the different motor tasks is shown in Fig. 4 for one representative
© 2014 John Wiley & Sons Ltd
MOTOR-UNIT RECRUITMENT IN THE MASSETER Table 1. Bite-force magnitudes (intra-individual, averaged over the four tasks) exerted by the individual subjects to achieve the maximum number of decomposable MUs; cv1 = mean coefficient of variation (SD/mean 9 100) describing variability during the 5 s recording periods, cv2 = mean coefficient of variation describing variability among the four tasks. The two columns on the right indicate the number of MUs identified in the individual subjects Motor units per subject
Force vector Variability (%)
Masseter
Subject
Newton
cv1
cv2
Deep
Superficial
1 2 3 4 5 6 7 8 9 10 Mean SD Σ
22 41 16 17 37 26 16 9 43 17 244 119
59 49 69 35 149 115 63 33 09 112 69 43
63 104 61 62 144 93 66 57 143 165 96 41
12 9 10 10 15 15 18 5 12 10 116 39 116
8 5 8 8 10 13 17 9 16 7 101 40 101
subject. Figure 5 shows the numbers of active MUs during the different tasks for all the individual subjects. The 116 MUs detected in the deep part of the masseter were recruited 229 times during the four motor tasks. The corresponding involvement of the 101 MUs detected in the superficial masseter, 164 recruitments, was significantly smaller (F = 774, df = 1; P < 002). Within both parts, significant differ-
ences in the number of recruited MUs were found between protrusive (86 deep; 67 superficial) and lateral (31 deep; 17 superficial) motor tasks (F = 718, df = 3; P < 0001). The highest mean CI values (i.e. 69% and 52% for the deep and superficial masseter, respectively) were found for paired vertical and protrusive motor tasks. Mean CI values for other task combinations were much lower, with a minimum for the combination of vertical and lateral force directions (Table 2A, B). Mean CI values for the six different pairs of tasks were found to be significantly different (F = 3012, df = 3; P < 0001) within both parts of the masseter. In addition, the CI values were significantly (F = 2058, df = 1; P < 0001) higher for the deep masseter. Quantitative estimates of the task-specificity of masseter MUs are given in Table 2C; these indicate that approximately 46% of the detected MUs were unimodal, 31% bimodal, 18% trimodal and 5% quadrimodal. Statistical analysis revealed that these proportions were significantly different, both within (F = 739, df = 3; P < 0001) and between (F = 827, df = 1; P < 002) the deep and superficial parts of the masseter.
Discussion In this study, task-dependent MU recruitment in the masseter was examined. Stationary bipolar fine-wire electrodes with a short interelectrode distance were used and, therefore, myoelectric activity from relatively small muscle subvolumes only was recorded. Application of advanced decomposition software enabled isola-
140 120
RMS [ V]
100 Superficial wire
80
Deep wire
60 40 20 0
Vert
Pro
Lat
Med
Motor task Fig. 3. Localised integrated EMG activity (RMS) of the superficial and deep masseter for the different motor tasks. Asterisks indicate significant differences between the tasks (P < 005). © 2014 John Wiley & Sons Ltd
481
482
H . J . S C H I N D L E R et al. (a) Arbitrary units 200
48 6
37 1
2
4
6
3
8 9
1
2 4
100 0 –100 –200 60·82
60·84
60·86
60·88
60·90
60·92
Motor task (b)
Pro
Vert
Lat
60·94 s
Med
MU number 1 2 3 4 5 6 7 8 9 Selected 5 s time frames
(c) MU number
Pro
Vert
Lat
Med
1 2 3 4 5 6 7 8 9 0
10
20
30
40
50
60
70
80
s
Fig. 4. EMG decomposition and analysis of recruitment behaviour for a representative subject. (a) The signal peaks and corresponding numbers in the interference signal shown in the upper trace (01 s recording in the deep masseter during lateral bite-force) indicate firing events of eight (MU 1, 2, 3, 4, 6, 7, 8, 9) of the nine decomposed MUs active in the four tasks shown below. The lower trace is the remaining interference signal after subtraction of the corresponding templates. The fact that this signal is, almost, baseline noise only is indicative of successful decomposition. (b) For demonstration purposes, the nine decomposed MU action potentials were averaged in this example over the separate 5 s time periods (instead of averaging them over the combined 20 s signal window). The finding of very small variation of action potential shapes of individual MUs in the different motor tasks shows that recruitment and/or derecruitment of identical MUs in the different motor tasks could be unambiguously identified. (c) Analysis of the recruitment behaviour. The black bars indicate which MU is active in which task, that is, in the vertical (Vert), protrusive (Pro), lateral (Lat) and medial (Med) bite-force directions. For this subject, three of the MUs are unimodal (MU5 for the vertical force direction, MU7 and MU8 for the lateral force direction), two MUs are bimodal (MU1 for vertical plus lateral tasks, MU9 for protrusive plus lateral tasks), three MUs are trimodal (MU2 and MU3 for vertical plus protrusive plus lateral tasks, and MU6 for protrusive plus lateral plus medial tasks), and MU4 is quadrimodal.
© 2014 John Wiley & Sons Ltd
MOTOR-UNIT RECRUITMENT IN THE MASSETER Deep recording
16
Superficial recording
Number of active MUs
14 12 10 8 6 4 2 0 Vert
Pro
Lat
Med
Vert
Motor task
Pro
Lat
Med
Motor task
Fig. 5. Numbers of MUs active in the superficial and deep masseter in the four different tasks for all individual subjects (n = 10).
tion of up to 10 MUs from the recording for a single motor task and up to 18 MUs from one concatenated EMG file containing the signals for all four motor tasks. By this means it was revealed (as far as we are aware, for the first time) that the heterogeneous activation behaviour of the masseter manifests itself even in small muscle subvolumes as a very localised phenomenon at the level of its smallest functional units. These findings confirm the initially stated hypothesis and add substan-
tially to those of previous reports (4, 5) in which taskspecificity was separately proved for individual MUs with locations in distinct muscle regions only. These results are, moreover, also supported by the local muscle activity differences identified at the interference EMG level. Quantification of the coincidence of MU recruitment for pairs of different motor tasks (i.e. defined bite-force vectors of the same length in four different
Table 2. Mean values of coincidence indexes (CI, %) with standard deviations (in parentheses). The CI correlates positively with the proportions of MUs recruited in both tasks of a specific task combination. CI values were averaged over all subjects (n = 10); Vert, vertical; Pro, protrusive; Lat, lateral; Med, medial bite-force directions. (A) CI values for the deep masseter. (B) CI values for the superficial masseter; N indicates the number of MUs identified as being involved in the task of the corresponding row. (C) Summary of the number of MUs active in one, two, three, or four tasks (A) CI values for the deep masseter Task
Vert
Pro
Lat
Med
N
Vert Pro Lat Med
100% – – –
69% (17%) 100% – –
23% (31%) 28% (25%) 100% –
36% (19%) 40% (24%) 30% (35%) 100%
59 86 31 55
(B) CI values for the superficial masseter Task
Vert
Pro
Lat
Med
N
Vert Pro Lat Med
100% – – –
52% (33%) 100% – –
7% (11%) 17% (28%) 100% –
11%(21%) 31% (18%) 10% (26%) 100%
44 67 17 32
(C) MU numbers and proportions for the different grades of task specificity Task per MU
Unimodal
Bimodal
Trimodal
Quadrimodal
N
Deep MUs Superficial MUs Total Proportion
47 (41%) 54 (53%) 101 46%
33 (28%) 34 (34%) 67 31%
28 (24%) 10 (10%) 38 18%
8 (7%) 3 (3%) 11 5%
116 101 217
© 2014 John Wiley & Sons Ltd
483
484
H . J . S C H I N D L E R et al. directions) revealed force-direction-specific recruitment for a specific proportion of the MUs only. The occurrence of task-specific and polymodal MUs in the masseter confirms results from a previous study (2). The exclusive finding of polymodal MUs in another study (4) might be because of the biomechanical similarity of some of the tasks examined. Because of the simultaneous observation of multiple MUs in this study we could subdivide the MUs into task-specific and polymodal types. In this manner, it was possible to approximately estimate which model of functional MU behaviour is dominant in the masseter. The variability of the CI between the task combinations and between the deep and superficial parts of the muscle reflects the distinct biomechanical contributions of highly direction-specific MUs in the different bite-force directions and in both parts of the masseter. Systematic characterisation of MUs into specific and polymodal types revealed that most of the MUs were single-task-specific (46%), followed by bimodal (31%), trimodal (18°) and quadrimodal (5%) MU recruitment behaviour. The polymodal behaviour might be related to biomechanical requirements. Indeed, recent anatomical studies using magnetic resonance imaging revealed highly variable inner muscle structure consisting of multiple small aponeuroses with different spatial orientations and dense interconnections (21). Such a complex structure requires stabilizing elements between the aponeuroses, especially during differential activation of the muscle. In principle, force transmission by MUs must be integrated into task-specific lines of action in the active muscle. Biomechanical interactions between single muscle fibres, whole fibre bundles, and aponeuroses in the complex muscle are incompletely understood. On the basis of the knowledge available (22, 23), however, it might be speculated that the polymodal MUs of the masseter provide, at least, supplementary force for internal muscle stabilisation, maintaining the basic muscle shape during function, and stiffening the complex tendon system to ensure task-related directional force application by other MUs. In this context, the different overall MU activation between and within both parts of the muscle might be indicative of different stabilisation needs of superficial and deep parts of the masseter and for distinct tasks. Polymodal MUs have been also found in other muscles, for example the trapezius muscle (24). Interestingly, these motor units, which are not task-specific, are most likely to
be type I motor units, sometime also described as ‘Cinderella’ motor units (25). When interpreting the quantitative results of this study, it must be taken into consideration that co-contraction, a well-known phenomenon during execution of untrained motor tasks (18), might partially explain the activity of task-unspecific MUs. Consequently, the extent of polymodality might have been somewhat overestimated in our study with untrained subjects. Moreover, because the results were obtained from two specific masseter subvolumes only, the suggested recruitment behaviour cannot be automatically assumed for other regions or for the whole muscle. This limitation also applies to the individual force values and their effect on recruitment modalities. It might, however, be assumed that the characteristics of neuromuscular behaviour in other muscle regions, and possibly also at distinct force levels, are, in principle, similar.
Conclusions Our findings for healthy subjects suggest the capability of the neuromuscular system to modify the mechanical output of small masseter subvolumes by differential control of adjacent MUs belonging to distinct task groups. Such a subtle control mechanism, supported by discrete MU territories and different lines of action of MUs, might enable the muscle to adjust its activity in a highly flexible and efficient manner in response to complex local biomechanical needs, for example, continually varying bite-forces during a demanding masticatory process. It might also be the neuromuscular basis of motor adaptations to pain, as has already been reported for the masseter (16).
Acknowledgments The authors wish to express their gratitude to Mr Willi Wendler (University of Karlsruhe) for his invaluable contribution to the design and fabrication of diverse mechanical components for the experimental setup and Dr. med. dent. Oliver Thiele (University of Heidelberg) for his assistance with the ultrasound imaging of the masseter.
Conflict of interests The authors received no financial support and declare no potential conflict of interests in respect of patent © 2014 John Wiley & Sons Ltd
MOTOR-UNIT RECRUITMENT IN THE MASSETER rights, corporate affiliations or consultancies for any product or process mentioned in the article. 14.
References 15. 1. Yemm R. The orderly recruitment of motor units of the masseter and temporal muscles during voluntary isometric contraction in man. J Physiol. 1977;265:163–174. 2. Eriksson PO, Stalberg E, Antoni L. Flexibility in motor-unit firing pattern in the human temporal and masseter muscles related to type of activation and location. Arch Oral Biol. 1984;29:707–712. 3. Tonndorf ML, Sasaki K, Hannam AG. Single-wire recording of regional activity in the human masseter muscle. Brain Res Bull. 1989;23:155–159. 4. McMillan AS, Hannam AG. Task-related behavior of motor units in different regions of the human masseter muscle. Arch Oral Biol. 1992;37:849–857. 5. Ogawa T, Kawata T, Tsuboi A, Hattori Y, Watanabe M, Sasaki K. Functional properties and regional differences of human masseter motor units related to three-dimensional bite force. J Oral Rehabil. 2006;33:729–740. 6. Blanksma NG, Van Eijden TM, Weijs WA. Electromyographic heterogeneity in the human masseter muscle. J Dent Res. 1992;71:47–52. 7. Blanksma NG, van Eijden TM, van Ruijven LJ, Weijs WA. Electromyographic heterogeneity in the human temporalis and masseter muscles during dynamic tasks guided by visual feedback. J Dent Res. 1997;76:542–551. 8. Schindler HJ, T€ urp JC, Blaser R, Lenz J. Differential activity patterns in the masseter muscle under simulated clenching and grinding forces. J Oral Rehabil. 2005;32:552–563. 9. T€ urp JC, Schindler HJ, Pritsch M, Rong Q. Antero-posterior activity changes in the superficial masseter muscle after exposure to experimental pain. Eur J Oral Sci. 2002;110:83– 91. 10. Widmer CG, Klugman D, English AW. Anatomical partitioning and nerve branching patterns in the adult rabbit masseter. Acta Anat (Basel). 1997;159:222–232. 11. Herring SW, Wineski LE, Anapol FC. Neural organization of the masseter muscle in the pig. J Comp Neurol. 1989;280:563–576. 12. Windhorst U, Hamm TM, Stuart DG. On the function of muscle and reflex partitioning. Behav Brain Sci. 1989;12:629–681. 13. Widmer CG, Carrasco DI, English AW. Differential activation of neuromuscular compartments in the rabbit masseter
© 2014 John Wiley & Sons Ltd
16.
17.
18.
19.
20.
21.
22. 23.
24.
25.
muscle during different oral behaviors. Exp Brain Res. 2003;150:297–307. Hodges PW, Tucker K. Moving differently in pain: a new theory to explain the adaptation to pain. Pain. 2011;152(3 Suppl):S90–S98. Tonndorf ML, Hannam AG. Motor unit territory in relation to tendons in the human masseter muscle. Muscle Nerve. 1994;17:436–443. Minami I, Akhter R, Albersen I, Burger C, Whittle T, Lobbezoo F et al. Masseter motor unit recruitment is altered in experimental jaw muscle pain. J Dent Res. 2013;92: 143–148. Dworkin S, LeResche L. Research diagnostic criteria for temporomandibular disorders: review, criteria, examinations and specifications, critique. J Craniomandib Disord Facial Oral Pain. 1992;6:301–355. Schindler HJ, Rues S, T€ urp JC, Schweizerhof K, Lenz J. Jaw clenching: muscle and joint forces, optimization strategies. J Dent Res. 2007;86:843–847. van Eijden TM, Koolstra JH, Brugman P, Weijs WA. A feedback method to determine the three-dimensional bite-force capabilities of the human masticatory system. J Dent Res. 1988;67:450–454. McGill KC, Lateva ZC, Marateb HR. EMGLAB: an interactive EMG decomposition program. J Neurosci Methods. 2005;149:121–133. Cioffi I, Gallo LM, Palla S, Erni S, Farella M. Macroscopic analysis of human masseter compartments assessed by magnetic resonance imaging. Cells Tissues Organs. 2012;195:465–472. Monti RJ, Roy RR, Edgerton VR. Role of motor unit structure in defining function. Muscle Nerve. 2001;24:848–866. Huijing PA. Epimuscular myofascial force transmission: a historical review and implications for new research. International Society of Biomechanics Muybridge Award Lecture, Taipei, 2007. J Biomech. 2009;42:9–21. Kadefors R, Forsman M, Zoega B, Herberts P. Recruitment of low threshold motor-units in the trapezius muscle in different static arm positions. Ergonomics. 1999;42:359–375. H€agg GM. Static work and myalgia - a new explanation model. In: Andersson PA, Hobart DJ, Danoff JV, eds. Electromyographical kinesiology. Amsterdam: Elsevier Science; 1991:141–144.
Correspondence: Daniel Hellmann, Department of Prosthodontics, University of Heidelberg, Im Neuenheimer Feld 400, 69120 Heidelberg, Germany. E-mail: [email protected]
485
