- Email: [email protected]
Bite force and masseter muscle electromyographic activity during onset of an isometric clench in man
0003-9969/X5 $3.00 + 0.00 Copyright (‘J 1985 Pergamon Press Ltd
Archs ornl Btol Vol. 30. No 3. pp. 213-215. 1985 Prmted in Great Bntam All rlghrs reserved
MUSCLE BITE FORCE AND MASSETER ELECTROMYOGRAPHIC ACTIVITY DURING ONSET AN ISOMETRIC CLENCH IN MAN Departments
OF
H. DEVLIN* and D. G. WAsTELL? of *Prosthetic Dentistry and tMedical Computation, University of Manchester, Higher Cambridge Street, Manchester Ml5 6FH, England, U.K.
Summary-The surface electromyogram (EMG) of the masseter muscle and interocclusal force were recorded in six human subjects during fast and slow clenches. The simple linear force/EMG relation that applies for steady bite-force levels was inadequate to explain their relationship under conditions of changing force: a second component of force, its rate of change, was necessary. EMG/force equations for fast and slow clenches were broadly similar. The rate of change factor may account for the apparent EMG/force lag observed in earlier studies.
INTRODUCTION
A linear relationship exists between integrated surface electromyographic (EMG) activity and the steady level of force generated by a muscle during an isometric contraction (e.g. Lippold, 1952; Bigland and Lippold. 1964; Manns, Miralles and Palazzi, 1979). However, when force is changing, we should not necessarily expect such a simple linear relationship to hold. When the output of physical systems is changed, resistance (friction) is often encountered in proportion to the rate at which the change is effected. Our hypothesis is that a term involving the rate of change of force as well as its absolute level would be important in describing the EMG-force relation during increase or decrease of force. We attempted to quantify the relative importance of these two force variables in accounting for the human masseter EMG recorded during onset of a clench. MATERIALS
AND METHODS
The subjects Six people were used, four males and two females, aged 3&45 years. All were healthy volunteers who gave their informed consent to participate in these experiments.
A brass transducer, 1.8 cm in diameter and 0.4 mm thick, was used to measure bite force. The device fulfilled most of the criteria described by Duxbury et al. (1973) for bite-force measurement. Load applied to the device was detected by a semiconductor strain gauge arranged in a Wheatstone bridge circuit. The amplified voltage from the transducer was linearly proportional to loads up to 490 N. The device was compensated for changes of temperature within the range 2&4O”C , and had a response time of 70 ps.
The following tracking task procedure was used to elicit clenches with standard onset profiles. Bite force was fed back to subjects using a moving spot on an oscilloscope screen which they were asked to maintain between a pair of parallel lines that demarcated the desired profile. The spot moved at 10 cm/s; screen
Either right or left masseter was chosen randomly. Silver-silver chloride electrodes were placed 2 cm apart over the main body of the muscle. An imaginary line connecting the electrodes would bisect the occlusal plane at a right angle. Inter-electrode resistances were maintained below 5 kohm throughout. The earth electrode was positioned on the skin of the neck at the level of the 5th cervical vertebra. The electrodes were connected to a Medelec EMG machine via Medelec PA62 pre-amplifiers. The filter settings on the EMG machine employed a band width of 16-800 Hz.
Cobalt n
The measurement of bite force force was measured between two cast
The bite-force transducer
The experimental procedure
The EMG recordings
The bite compressed
(0.8 mm thick) made of cobalt-chromium alloy (see Fig. 1). The transducer was positioned over the palate midline, opposite the premolar teeth. The purpose of the metal platforms was to distribute the biting load over the posterior teeth of both sides, while keeping the teeth out of contact. The cobalt-chromium platforms were parallel to the occlusal plane and covered the occlusal surface of the posterior teeth. The platforms extended over the occlusal surfaces of the molar and premolar teeth present in each arch, as well as the palatal and lingual surfaces of the canines.
chromium platforms
Fig. 1. The bite-force transducer attached to the upper cobalt-chromium platform, separating the posterior teeth slightly.
by a transducer metal platforms 213
H. DEVLIN
214
and
D.G.
width was IO cm. The parallel lines were drawn 1 cm apart on a Perspex overlay. Two templates were used to produce an S-shaped fast clench (200 ms duration) and a ramp-shaped slow clench (400 ms duration). Both clenches had the same overall amplitude and were held for approx. 0.5 s before release. Fast clenches achieved a maximum rate of rise of force roughly twice that for slow clenches.
WASTELL 00
50
0
50
100
150
200
100
200
300
400
The computer analysis Force and EMG were fed into the analogue-todigital converters of a PDP I l/34 computer (digitization rate 3 kHz). At the end of each series of recordings, the program recalled the data for vetting: clenches not conforming to the template were excluded. Typically, 10-I 5 satisfactory clenches of each type were obtained. A second computer program was used to measure EMG amplitude (root mean square value for a sliding 50 ms window), force level (F) and increment in force (AF) at 10ms intervals across the period of clench onset. In order to compare data from different subjects, EMG and force were normalized for each subject by dividing through by their respective maximum values. An average fast and an average slow bite were computed for each subject. These averages were then transferred to a mainframe computer for fitting regression lines using the Genstat statistical package.
L Time
( ms)
Fig. 3. EMG and force (percentage maximum) are shown as a functron of time for fast (A) and slow (B) clenches for a typical subject.
EMG-force relationships one typical subject):
(Fig. 3 shows averages
for
RESULTS
Figure 2 illustrates raw EMG and force recordings from one subject, typical of the others, during a fast clench. There were clear separations between the points at which EMG and force attained their maximum values: integrated EMG reached its peak about half way through the period of clench onset, i.e. at the time that force is increasing most rapidly. The following equations were fitted to each subject’s fast and slow averages in order to quantify the
of
maximum
clenching
EMG=a+bxF+cxAF
C-4
where F = the absolute level of the bite force; AF = the rate of change of the bite force per ms; a, b and c are coefficients to be estimated. The percentage of EMG variance explained by each equation was determined by analysis of variance. Using the first equation, an average of 36 per cent of
5.
force
(b)
Raw EMG frace
(c)
of maximum
100
Percentage integrated
(I)
100
Percentage (a)
EMG=a+bxF
I
50 EMG
,
I, O_
0
50
*--*
I
,
,
,
,
,
,
,
,
?OO
150
200
250
300
350
400
450
Time
(ms)
Fig. 2. Raw data for a typical trial for one subject is shown: Traces a, b, and c respectively illustrate force (percentage maximum), raw EMG and integrated EMG (percentage maximum) as a function of time during a fast clench.
EMG
and force
the EMG was accounted for in a fast bite; during a slow bite, this figure rose to 70 per cent. By including rate of change of force (AF), the percentage of EMG variance explained for a fast bite increased to 90 per cent; for a slow bite the figure rose from 70 to 93 per cent. Averaging coefficients across subjects we have equation 2 for fast and slow clenches as follows: EMG (fast) = -0.02 + 0.54 F + 114 AF EMG (slow) = 0.02 + 0.66 F + 82 AF
215
during clench onset
(3) (4)
The form of the equations for fast and slow bites are broadly similar. Statistical analysis showed that the constant term was not significantly different from zero for either case (t = 0.5,~ > 0.3; t = 0.4,~ > 0.3). Both F and AF were reliable across subjects for both fast (t = 5.0, p > 0.01; t = 6.3, p ~0.01) and slow bites (t =6.1, p
Our results confirm the hypothesis that the relationship between EMG and force is not a simple linear one. During conditions of changing force, such as clench onset, EMG is composed of two components, one related to absolute level of force (already known from steady-state studies) and a second reflecting rate of change of force. During clench onset, force accelerates, reaches maximum and then decelerates; the rate-related EMG component follows a similar up-and-down course; that superimposed on the linear component gives rise to a peak in EMG occurring before force reaches its maximum final value. For fast clenches, this peak in EMG will tend to coincide with the point of maximum rate of change of force, which will generally occur around the middle of clench onset. For slower clenches where the rate factor is relatively less important, EMG would be less peaked and the peak would occur later in the onset period. Other studies have noted a lag between peak EMG and peak force: Inman et al. (1952) reported a variable lag of approx. 80 ms ( f 20 ms) between peak integrated EMG and peak tension of human limb muscles during voluntary isometric contractions. Hannam, Inkster and Scott (1975), in their study relating peak EMG of the right anterior temporalis muscle to jaw-closing force in man, found a mean delay between peak-electrical activity and peak force of 73 ms (SD + 12 ms). Ahlgren and Owall (1970) report a lag of 43 ms during chewing. Our study is important in showing these delays not to be simple lags but to reflect the presence of an EMG factor
reflecting rate of clenching. From results and illustrations in Inman et al. and Ahlgren and Owall, peak EMG occurred, as we would predict, during the mid-phase of clenching when rate of force change is greatest. A system in which output is a simple linear function of input is known as a zero-order system (Grodins, 1963). With the appearance of a second term involving the first derivative of output (velocity) the system becomes first-order. This second component is often involved in describing frictional forces in physical systems. Using this terminology, we may describe the EMG-force relationship under steadystate conditions as a zero-order system. However, during clench onset we have shown that a first-order model is required involving two components of force, its absolute level and its rate of change. We might speculate that the first of these components reflects muscle activity required to sustain the current-force level; and that the second expresses the additional increment of activity required in order to bring about a given increase in force, which may reflect frictionlike resistances within joints and muscles. Acknowledgements-We wish to express our gratitude to Professor A. A. Grant and Dr A. J. Duxbury for their advice, encouragement and helpful criticisms throughout the period during which these experiments were undertaken. REFERENCES
Ahlgren J. and Owall B. (1970) Muscular activity and chewing force: a polygraphic study of human mandibular movements. Archs oral Biol. 15, 271-280. Bigland B. and Lippold 0. C. J. (1954) Motor unit activity in voluntary contraction of human muscle. J. Phyriol. 125, 322-335. Duxbury A. J., Frame J. W., Rothwell P. S. and Jackson P. D. (1973) An instrument for the measurement of incisive bite force. J. Dent. 1, 246-250. Hannam A. G.. Inkster W. C. and Scott J. D. (1975) Peak electromyographic activity and jaw closing force in man. J. dent. Res. 54, 694. Grodins F. S. (1963) Control Theory and Biological Svstems. Colombia University Press, New York. Inman V. T., Ralston H. J., Saunders J. C., Fernstein C. M. de, Bertram C. and Wright E. W. Jr (1952) Relation of human electromyogram to muscular tension. Elertroenceph. clin. Neurophysiol. 4, 187-194. Lippold 0. C. J. (1952) The relation between integrated action potentials in a human muscle and its isometric tension. J. Physiol. 117, 492499. Manns A., Miralles R. and Palazzi C. (1979) EMG, bite force and elongation of the masseter muscle under isometric contractions and variations of vertical dimension. J. prosthet. Dent. 42, 674-682.
Archs ornl Btol Vol. 30. No 3. pp. 213-215. 1985 Prmted in Great Bntam All rlghrs reserved
MUSCLE BITE FORCE AND MASSETER ELECTROMYOGRAPHIC ACTIVITY DURING ONSET AN ISOMETRIC CLENCH IN MAN Departments
OF
H. DEVLIN* and D. G. WAsTELL? of *Prosthetic Dentistry and tMedical Computation, University of Manchester, Higher Cambridge Street, Manchester Ml5 6FH, England, U.K.
Summary-The surface electromyogram (EMG) of the masseter muscle and interocclusal force were recorded in six human subjects during fast and slow clenches. The simple linear force/EMG relation that applies for steady bite-force levels was inadequate to explain their relationship under conditions of changing force: a second component of force, its rate of change, was necessary. EMG/force equations for fast and slow clenches were broadly similar. The rate of change factor may account for the apparent EMG/force lag observed in earlier studies.
INTRODUCTION
A linear relationship exists between integrated surface electromyographic (EMG) activity and the steady level of force generated by a muscle during an isometric contraction (e.g. Lippold, 1952; Bigland and Lippold. 1964; Manns, Miralles and Palazzi, 1979). However, when force is changing, we should not necessarily expect such a simple linear relationship to hold. When the output of physical systems is changed, resistance (friction) is often encountered in proportion to the rate at which the change is effected. Our hypothesis is that a term involving the rate of change of force as well as its absolute level would be important in describing the EMG-force relation during increase or decrease of force. We attempted to quantify the relative importance of these two force variables in accounting for the human masseter EMG recorded during onset of a clench. MATERIALS
AND METHODS
The subjects Six people were used, four males and two females, aged 3&45 years. All were healthy volunteers who gave their informed consent to participate in these experiments.
A brass transducer, 1.8 cm in diameter and 0.4 mm thick, was used to measure bite force. The device fulfilled most of the criteria described by Duxbury et al. (1973) for bite-force measurement. Load applied to the device was detected by a semiconductor strain gauge arranged in a Wheatstone bridge circuit. The amplified voltage from the transducer was linearly proportional to loads up to 490 N. The device was compensated for changes of temperature within the range 2&4O”C , and had a response time of 70 ps.
The following tracking task procedure was used to elicit clenches with standard onset profiles. Bite force was fed back to subjects using a moving spot on an oscilloscope screen which they were asked to maintain between a pair of parallel lines that demarcated the desired profile. The spot moved at 10 cm/s; screen
Either right or left masseter was chosen randomly. Silver-silver chloride electrodes were placed 2 cm apart over the main body of the muscle. An imaginary line connecting the electrodes would bisect the occlusal plane at a right angle. Inter-electrode resistances were maintained below 5 kohm throughout. The earth electrode was positioned on the skin of the neck at the level of the 5th cervical vertebra. The electrodes were connected to a Medelec EMG machine via Medelec PA62 pre-amplifiers. The filter settings on the EMG machine employed a band width of 16-800 Hz.
Cobalt n
The measurement of bite force force was measured between two cast
The bite-force transducer
The experimental procedure
The EMG recordings
The bite compressed
(0.8 mm thick) made of cobalt-chromium alloy (see Fig. 1). The transducer was positioned over the palate midline, opposite the premolar teeth. The purpose of the metal platforms was to distribute the biting load over the posterior teeth of both sides, while keeping the teeth out of contact. The cobalt-chromium platforms were parallel to the occlusal plane and covered the occlusal surface of the posterior teeth. The platforms extended over the occlusal surfaces of the molar and premolar teeth present in each arch, as well as the palatal and lingual surfaces of the canines.
chromium platforms
Fig. 1. The bite-force transducer attached to the upper cobalt-chromium platform, separating the posterior teeth slightly.
by a transducer metal platforms 213
H. DEVLIN
214
and
D.G.
width was IO cm. The parallel lines were drawn 1 cm apart on a Perspex overlay. Two templates were used to produce an S-shaped fast clench (200 ms duration) and a ramp-shaped slow clench (400 ms duration). Both clenches had the same overall amplitude and were held for approx. 0.5 s before release. Fast clenches achieved a maximum rate of rise of force roughly twice that for slow clenches.
WASTELL 00
50
0
50
100
150
200
100
200
300
400
The computer analysis Force and EMG were fed into the analogue-todigital converters of a PDP I l/34 computer (digitization rate 3 kHz). At the end of each series of recordings, the program recalled the data for vetting: clenches not conforming to the template were excluded. Typically, 10-I 5 satisfactory clenches of each type were obtained. A second computer program was used to measure EMG amplitude (root mean square value for a sliding 50 ms window), force level (F) and increment in force (AF) at 10ms intervals across the period of clench onset. In order to compare data from different subjects, EMG and force were normalized for each subject by dividing through by their respective maximum values. An average fast and an average slow bite were computed for each subject. These averages were then transferred to a mainframe computer for fitting regression lines using the Genstat statistical package.
L Time
( ms)
Fig. 3. EMG and force (percentage maximum) are shown as a functron of time for fast (A) and slow (B) clenches for a typical subject.
EMG-force relationships one typical subject):
(Fig. 3 shows averages
for
RESULTS
Figure 2 illustrates raw EMG and force recordings from one subject, typical of the others, during a fast clench. There were clear separations between the points at which EMG and force attained their maximum values: integrated EMG reached its peak about half way through the period of clench onset, i.e. at the time that force is increasing most rapidly. The following equations were fitted to each subject’s fast and slow averages in order to quantify the
of
maximum
clenching
EMG=a+bxF+cxAF
C-4
where F = the absolute level of the bite force; AF = the rate of change of the bite force per ms; a, b and c are coefficients to be estimated. The percentage of EMG variance explained by each equation was determined by analysis of variance. Using the first equation, an average of 36 per cent of
5.
force
(b)
Raw EMG frace
(c)
of maximum
100
Percentage integrated
(I)
100
Percentage (a)
EMG=a+bxF
I
50 EMG
,
I, O_
0
50
*--*
I
,
,
,
,
,
,
,
,
?OO
150
200
250
300
350
400
450
Time
(ms)
Fig. 2. Raw data for a typical trial for one subject is shown: Traces a, b, and c respectively illustrate force (percentage maximum), raw EMG and integrated EMG (percentage maximum) as a function of time during a fast clench.
EMG
and force
the EMG was accounted for in a fast bite; during a slow bite, this figure rose to 70 per cent. By including rate of change of force (AF), the percentage of EMG variance explained for a fast bite increased to 90 per cent; for a slow bite the figure rose from 70 to 93 per cent. Averaging coefficients across subjects we have equation 2 for fast and slow clenches as follows: EMG (fast) = -0.02 + 0.54 F + 114 AF EMG (slow) = 0.02 + 0.66 F + 82 AF
215
during clench onset
(3) (4)
The form of the equations for fast and slow bites are broadly similar. Statistical analysis showed that the constant term was not significantly different from zero for either case (t = 0.5,~ > 0.3; t = 0.4,~ > 0.3). Both F and AF were reliable across subjects for both fast (t = 5.0, p > 0.01; t = 6.3, p ~0.01) and slow bites (t =6.1, p
Our results confirm the hypothesis that the relationship between EMG and force is not a simple linear one. During conditions of changing force, such as clench onset, EMG is composed of two components, one related to absolute level of force (already known from steady-state studies) and a second reflecting rate of change of force. During clench onset, force accelerates, reaches maximum and then decelerates; the rate-related EMG component follows a similar up-and-down course; that superimposed on the linear component gives rise to a peak in EMG occurring before force reaches its maximum final value. For fast clenches, this peak in EMG will tend to coincide with the point of maximum rate of change of force, which will generally occur around the middle of clench onset. For slower clenches where the rate factor is relatively less important, EMG would be less peaked and the peak would occur later in the onset period. Other studies have noted a lag between peak EMG and peak force: Inman et al. (1952) reported a variable lag of approx. 80 ms ( f 20 ms) between peak integrated EMG and peak tension of human limb muscles during voluntary isometric contractions. Hannam, Inkster and Scott (1975), in their study relating peak EMG of the right anterior temporalis muscle to jaw-closing force in man, found a mean delay between peak-electrical activity and peak force of 73 ms (SD + 12 ms). Ahlgren and Owall (1970) report a lag of 43 ms during chewing. Our study is important in showing these delays not to be simple lags but to reflect the presence of an EMG factor
reflecting rate of clenching. From results and illustrations in Inman et al. and Ahlgren and Owall, peak EMG occurred, as we would predict, during the mid-phase of clenching when rate of force change is greatest. A system in which output is a simple linear function of input is known as a zero-order system (Grodins, 1963). With the appearance of a second term involving the first derivative of output (velocity) the system becomes first-order. This second component is often involved in describing frictional forces in physical systems. Using this terminology, we may describe the EMG-force relationship under steadystate conditions as a zero-order system. However, during clench onset we have shown that a first-order model is required involving two components of force, its absolute level and its rate of change. We might speculate that the first of these components reflects muscle activity required to sustain the current-force level; and that the second expresses the additional increment of activity required in order to bring about a given increase in force, which may reflect frictionlike resistances within joints and muscles. Acknowledgements-We wish to express our gratitude to Professor A. A. Grant and Dr A. J. Duxbury for their advice, encouragement and helpful criticisms throughout the period during which these experiments were undertaken. REFERENCES
Ahlgren J. and Owall B. (1970) Muscular activity and chewing force: a polygraphic study of human mandibular movements. Archs oral Biol. 15, 271-280. Bigland B. and Lippold 0. C. J. (1954) Motor unit activity in voluntary contraction of human muscle. J. Phyriol. 125, 322-335. Duxbury A. J., Frame J. W., Rothwell P. S. and Jackson P. D. (1973) An instrument for the measurement of incisive bite force. J. Dent. 1, 246-250. Hannam A. G.. Inkster W. C. and Scott J. D. (1975) Peak electromyographic activity and jaw closing force in man. J. dent. Res. 54, 694. Grodins F. S. (1963) Control Theory and Biological Svstems. Colombia University Press, New York. Inman V. T., Ralston H. J., Saunders J. C., Fernstein C. M. de, Bertram C. and Wright E. W. Jr (1952) Relation of human electromyogram to muscular tension. Elertroenceph. clin. Neurophysiol. 4, 187-194. Lippold 0. C. J. (1952) The relation between integrated action potentials in a human muscle and its isometric tension. J. Physiol. 117, 492499. Manns A., Miralles R. and Palazzi C. (1979) EMG, bite force and elongation of the masseter muscle under isometric contractions and variations of vertical dimension. J. prosthet. Dent. 42, 674-682.