- Email: [email protected]
Muscle temperature alters the EMG power spectrum of the canine diaphragm
Respiration Physiology, 94 (1993) 241-250 © 1993 Elsevier Science Publishers B.V. All rights reserved. 0034-5687/93/$06.00
241
RESP 02079
Muscle temperature alters the EMG power spectrum of the canine diaphragm Jon R. Doud and John M. Walsh* Division of Pulmonary and Critical Care Medicine Loyola UniversiO' Chicago Stritch School c?fMedicine and Edward Hines, Jr. V.A. Hospital Maywood. Illinois 60153, USA (Accepted 26 July 1993) Abstract. Electromyogram (EMG) power spectral analysis has been widely used to detect muscle fatigue. Non-fatigue factors that influence the EMG power spectrum must be well understood to ensure accurate interpretation. We explored the effect of muscle temperature on the EMG power spectrum in an isolated canine diaphragm preparation with intact neuro-vascular supply. In 5 animals, diaphragm temperature was continuously recorded with an intramuscular thermocouple, and the EMG signal was recorded with intramuscular wire electrodes. Data were obtained from spontaneous isometric diaphragm contractions while the diaphragm underwent passive cooling from 40 to 30 °C. The EMG power spectrum was determined using fast Fourier transform and each spectrum was characterized by its center frequency (fc). A direct linear relationship was found between diaphragm temperature and the EMG fc. The mean slope of the temperatureEMG fc relationship was 2.9~'~, fc' °C -~. In conclusion, diaphragm temperature is a major determinant of the diaphragmatic EMG power spectrum, and should be controlled when this index is employed in the investigation of muscle fatigue.
EMG, center frequency; Mammals, dog; Muscle, fatigue; Temperature, EMG spectral analysis
Power spectral analysis of the electromyogram (EMG) is widely employed to identify skeletal muscle fatigue (Gross et al., 1979; Kadefors et al., 1968; LindstrOm et at., 1977; Lindstr~m and Magnusson, 1977). Muscle fatigue produces a shift of the EMG power spectrum towards lower frequencies, which can be identified by a decreased center frequency (fc) (Aldrich et at., 1983). Accurate interpretation of the EMG power spectral shift requires a thorough understanding of the factors that significantly influence the EMG power spectrum of both fatigued and non-fatigued muscle. For example, it has recently been appreciated that muscle length is a significant determinant of the EMG fc (Bazzy et al., 1986; Doud and Walsh, 1992; Doud et al., 1992; Walsh et al., 1990). It is thought that this results from an alteration in muscle fiber conduction velocity related to changes in fiber diameter (Arendt-Nielsen and Mills, 1985; Eberstein and Beattie, 1985; LindstrOm et al., 1970; Mortimer et al., 1970; Stulen and DeLuca, 1981).
*Corresponding author. At: Loyola University Medical Center, 2160 South First Avenue Maywood, IL 60153, USA. Tel.: (708) 216-5404; Fax: (708) 216-8897.
242 Muscle temperature is an additional non-fatigue factor that affects muscle fiber conduction velocity (Buchthal and Engb~ek, 1963; Jarcho et al., 1954; Sfftlberg, 1966), and hence may alter the E M G power spectrum. Accordingly, studies in limb muscle (Merletti et al., 1984; Petrofsky and Lind, 1980) have demonstrated a linear correlation between muscle temperature and the E M G fc. However, the relationship between muscle temperature and the E M G fc has never been investigated in the respiratory muscles. Because of morphologic, functional, and biochemical differences between limb skeletal muscle and the diaphragm, demonstration of this relationship rather than extrapolation is necessary. Significant muscle temperature changes may occur in diaphragm strip preparations (particularly with in situ models) commonly used in fatigue studies (Bark et al., 1987; Hussain et al., 1989; Shindoh et al., 1990). Therefore, a full understanding of the effect of temperature on the diaphragmatic E M G power spectrum is critical for the accurate interpretation of E M G power spectral data in this type of preparation. The purpose of the present study was to examine the effect of muscle temperature on the diaphragmatic E M G fc. We utilized a mechanically isolated strip of costal diaphragm developed in situ in an anesthetized, mechanically ventilated canine model with spontaneous diaphragm contractions.
Methods
The protocol was approved by the Institutional Animal Care and Use Committee of Loyola University Stritch School of Medicine. Five healthy mongrel dogs weighing 26-29 kg were anesthetized with a 30 mg.kg ~ bolus of sodium pentobarbital, followed by a continuous infusion of approximately 2 mg'kg-~'h During the surgical preparation the animals were mechanically hyperventilated to induce apnea. The rate of ventilation was reduced during the experimental protocol to allow spontaneous diaphragm contractions. Two liters per minute of supplemental oxygen was administered continuously, and 5 cm H20 of positive end-expiratory pressure was applied. A catheter was placed in the abdominal aorta via the right femoral artery to monitor arterial blood pressure; a second catheter was placed in the right femoral vein for the administration of fluids and sedation. A heating blanket was used to maintain a core temperature of 38 + 0.5 °C. To prolong inspiratory times, bilateral vagotomy was performed on all animals prior to data acquisition.
Animalpreparation.
A 10 cm wide strip of left costal hemidiaphragm was mechanically isolated using a modification of a previously described technique (Kim et al., 1976). Briefly, the left hemithorax was exposed via a wide thoracotomy incision, and the left internal mammary and the lower 5-6 intercostal arteries were ligated. The left costal diaphragm rib attachments were identified through a left lateral abdominal incision. The involved ribs were severed medially and laterally to isolate a 10 cm wide strip of costal diaphragm; costal margin insertions, phrenic nerve, and phrenic vascu-
Left hemidiaphragm isolation.
243 lature were all maintained intact. Electro-cautery was used to create the diaphragm strip by dividing the costal diaphragm from the costal margin to the central tendon at the midline, and 10 cm lateral to the midline (Fig. 1). The position of the central tendon was fixed to a metal rack with a large clamp, taking care not to distort the phrenic vessel blood flow. The free costal margin of the din-
DIAPHRAGM FREE MARGIN
~
I
III 1''' I '1--'-/' i
If
l, //
! J!!,J.JlJ
PHRENIC VESSELS Fig. l. Schematic representation of the in situ diaphragm preparation, with an abdominal view of the mechanically isolated left costal diaphragm strip. The phrenic nerve is intact, and phrenic vasculature is similarly undisturbed. The mechanical fixation of the costal margin and central tendon to the rigid metal rack is not shown.
244 phragm strip was suspended from an isometric strain gauge force transducer fixed to the rack. Plastic wrap and surgical towels were used to limit desiccation and heat loss.
Measurement of diaphragm length. A pair of piezoelectric crystals (2.5 mm diameter) were implanted on the thoracic side of the diaphragm with an inter-crystal distance of approximately 15 ram. The crystals were placed in line with the muscle fibers and secured with 4 - 0 prolene suture followed by cyanoacrylate adhesive for maximal stabilization. The crystals were connected to a 4-channel ultrasonic device (Sonomicrometer model 120, Triton Technology, San Diego, CA). The crystals provided accurate recordings of diaphragm length (Newman et al., 1984) throughout the data acquisition period. The potential temperature-induced error in length measurement is acceptably low (approximately 0.1 °C ~). Muscle length was fixed at the resting in situ length, defined as the maximal length at which the relaxed muscle fails to exert a force. Contractions were isometric in order to exclude the influence of muscle length on the E M G power spectrum.
EMG signal. Two teflon-coated intramuscular wire electrodes (0.005" diameter,0.5" coiled length) were positioned at the distal margin of the diaphragm strip to record the E M G signal. Wires were placed in line with the piezoelectric crystals, with a distance of approximately 10 mm separating the 2 electrodes. The electrodes were secured with 4 - 0 prolene suture and cyanoacrylate adhesive to eliminate movement. The E M G signal was amplified (Grass, model P511K) and band pass filtered (CWE, model BPF-935) at 20-500 Hz.
Diaphragm temperature.
A one-inch 33 gauge stainless steel thermocouple (Model HYP-0, Omega Engineering Inc), calibrated from 20-40 ° C, was implanted within the muscle on the thoracic side of the diaphragm preparation to continuously record diaphragm temperature. The thermocouple was positioned within 2 cm of the E M G electrodes (Fig. 1) and secured with fine suture. An external radiant heat source (placed no closer than 40 cm) was used as necessary to maintain desired diaphragm temperature. Laboratory room temperature was kept at 25 ° C.
Experimentalprotocol.
Data from ten sequential spontaneous breaths were acquired at approximately one degree intervals while diaphragm temperature fell passively towards room temperature. The diaphragm temperature ranged from 40 to 30 ° C. To address the possibility of irreversible muscle injury induced by the experimental protocol, the diaphragm was rewarmed in one animal. Muscle rewarming with a heat lamp was performed from 30 to 40 °C, at a similar rate as cooling. E M G signal and temperature data were again acquired at 1 °C intervals. To investigate the mechanism responsible for the temperature-induced power spectral shift, compound diaphragm action potentials (CDAP) were evoked in one animal via phrenic nerve stimulation. A self-retaining bi-polar phrenic nerve stimulator was placed circumferentially around the phrenic nerve at the level of the heart. The phrenic
245 nerve was supra-maximally stimulated (duration of 0.25 ms) at the two temperature extremes (5 stimulations at each temperature), and the resultant CDAPs were acquired for power spectral analysis.
Data acquisition and analysis.
The E M G signal and muscle temperature were recorded on a Gould chart recorder (Series 2800, 8-channel) and simultaneously digitized at a sampling frequency of 2000 Hz via the analog/digital board in an IBM Model 70 computer and stored on hard disk (DataSponge, BioScience Analysis Software Ltd). EMG frequency component analysis was modeled after that reported by Richardson and Mitchell (1982). Each contraction was divided into 1024 ms segments of data, with an overlap of 768 ms (75 ~o common data) between successive segments. The number of segments was determined by the duration of diaphragm contraction. After EMG signal pretreatment by trend removal and application of a Hanning window, the E M G fc was determined for each data segment by applying a fast Fourier transform (VGA; Inspirational Software, 1990). The parameter used to characterize the power spectra and follow the change in signal frequency content was the center (or centroid) frequency, defined as the weighted mathematical mean frequency of the power spectrum: fc = .Y-,AiFi/ZAi, where Ai = amplitude associated with frequency F i. Temperature and E M G fc values fi'om each 1 °C interval were averaged and plotted. To facilitate comparison of data between animals, E M G fc values were normalized with respect to values taken at 38 °C. The evoked CDAPs elicited via phrenic nerve stimulation were analyzed in a similar manner to the spontaneous E M G signal. Fast Fourier transform was applied to the CDAP data, and each resulting power spectrum was represented by its fc.
Statistical analysis. Linear regression analysis of the relationship between diaphragm temperature and EMG fc was performed on the data from each animal. Slopes of the resulting regression lines (in ~ofc.°C 1) were calculated for each animal, and significance determined at a P < 0 . 0 5 level.
Results
A reduction in diaphragm temperature from 40 to 30 °C resulted in a significant reduction in E M G fc in each animal. The mean decrease in E M G fc across this temperature range exceeded 25 ~o, and was similar in all animals (Fig. 2). Linear regression analysis determined significant slopes in each individual animal, ranging from 2.4 to 3.4~o fc' °C 1. Correlation coefficients calculated from all data points from each animal ranged from 0.79 to 0.95 (Table 1). Group data displayed in Fig. 2 had a significant slope of 2.9~/o fc' °C -t and an r value of 0.95. In one animal, rewarming of the cooled diaphragm produced an increase in the E M G
246
115 ~9
105 ~> 0
O3
L-v3
95
~2 85
75 S o
•
/ o
65 28
o
I
I
I
I
I
I
I
30
32
34
36
38
40
42
temperature (°C) Fig. 2. The relationship between the EMG center frequency (fc) and diaphragm temperature in the five animals studied. The regression line is superimposed on the data points (r = 0.97). The EMG fc is expressed as a percentage of the value obtained at 38 °C. Data points represent mean data for a one degree temperature interval; data derived from each animal are represented by a unique symbol.
fc. After r e w a r m i n g , the E M G fc values were e q u i v a l e n t to those o b t a i n e d at the start o f the cooling p r o t o c o l (Fig. 3). T h e C D A P d u r a t i o n of the w a r m e d m u s c l e (40 ° C ) was c o m p a r e d to that o f the cooled m u s c l e (30 ° C ) in o n e a n i m a l . T h e 5 C D A P s at each t e m p e r a t u r e were superi m p o s a b l e . R e p r e s e n t a t i v e w a v e f o r m s from the two t e m p e r a t u r e extremes are s h o w n in Fig. 4. T h e C D A P d u r a t i o n was a p p r o x i m a t e l y 25 ~o longer u n d e r the colder c o n ditions. This p r o l o n g a t i o n o f the action p o t e n t i a l in cooled m u s c l e was a s s o c i a t e d with an a p p r o x i m a t e l y 3 0 % r e d u c t i o n in E M G fc o f the C D A P (103 H z to 74 Hz). TABLE 1 Slopes and regression coefficients for temperature-EMG fc relationship. Animal
Slope (°ofc.°C i)
r value
1 2 3 4 5
3.2 2.6 2.4 3.4 2.8
0.87 0.79 0.90 0.95 0.95
fc, center frequency.
247
120
...,.-'''"
ii0
[] ,...'''' ..~" i00 0
cO c~
90
o
80
...../ .,..." .~ ~
"'''''''''
70
60 28
30
I
I
I
[
32
34
36
38
40
42
temperature (°C) Fig. 3. Data obtained during diaphragm cooling and rewarming in one animal. The EMG center frequency (fc) is expressed as a percentage of the value obtained at 38 °C. Data points represent mean data for a one degree temperature interval. • = Data from cooling protocol; [] = data from rewarming protocol.
Systemic arterial pH and Pco, were similar at the initiation and conclusion of each experimental protocol. In all animals, change in arterial pH was < 0.05 units, and Pco, < 4 Torr.
Discussion
The canine diaphragmatic EMG power spectrum is strongly influenced by changes in intramuscular temperature. We found a significant linear relationship between the temperature and the EMG fc (mean slope 2.9~o fc. °C 1). The effect of temperature on the EMG power spectrum is probably due to a temperature-induced alteration in muscle fiber conduction velocity. Jarcho et al. (1954) described a significant positive relationship between temperature and conduction velocity between 27 and 40 °C in the rat anterior gracilis muscle. This strong association between temperature and muscle fiber conduction velocity was confirmed in the human biceps by St51berg (1966). In addition, Buchthal and Engb~ek (1963) demonstrated approximately 3.5 ~o change in conduction velocity per °C in the frog sartorius muscle, consistent with the temperature-induced E M G fc changes observed in the present study. Slowing of the muscle fiber conduction velocity prolongs the motor unit action
248
,,- - A = 4 0 o, EMG f c = 1 0 3 *
0
* = 3 0 °, EMG f c = 7 4
Hz Hz
I
[
I
5
i0
15
time (msec) Fig. 4. Compound diaphragm action potentials evoked via supra-maximal phrenic nerve stimulation at two different temperatures in one animal. The action potential duration was markedly prolonged at 30 °C compared with 40 °C; this was associated with an approximately 30% reduction in EMG fc (74 Hz and 103 Hz, respectively).
potential duration (Iaizzo and Poppele, 1990), resulting in a decreased E M G fc. In the present investigation C D A P duration was markedly prolonged at 30 ° compared with 40 ~C (Fig. 4). This observation is consistent with a temperature-related slowing of the muscle fiber conduction velocity ultimately manifested as a decreased E M G fc. This mechanism is supported by the findings of Bigland-Ritchie et al. (1981), where cooling of the adductor pollicis muscle resulted in slowing of the muscle fiber conduction velocity, with an associated reduction in the E M G high/low ratio (a parameter proportional to the E M G fc). The magnitude of E M G fc alteration caused by changes in temperature in our study is similar to that observed in non-respiratory muscles. Petrofsky and Lind (1980) studied the effect of cooling on the brachioradialis E M G ; the resulting E M G fc was shown to decrease by approximately 4 ?J~,. ° C a over a similar temperature range to that used in the present study. Additionally, Merletti et al. (1984) investigated cooling of the first dorsal interosseous muscle, and found the slope of the t e m p e r a t u r e - E M G fc regression line to be approximately 3g; f c ' ° C When the muscle was rewarmed to its initial temperature, the E M G fc values approximated the values acquired prior to cooling. In addition, the E M G fc followed temperature in a linear fashion during rewarming (Fig. 3). The difference between
249 cooling and rewarming fc values in the 30 to 35 °C range is likely due to a lag in rewarming at the E M G electrode site. This may have occurred as a result from relatively rapid rewarming over this temperature range. These observations imply that the electrode position remained unchanged throughout the experiment, and that muscle cooling did not induce irreversible tissue injury. Muscle length changes throughout the study were minimized in order to diminish the influence of muscle length on the E M G fc. Diaphragm length was fixed, and shortening was less than 5°,o in all experiments, and of similar magnitude at both high and low temperatures. The implications of this study may be most important in animal models employing isolated and exposed muscles where large temperature changes may occur. This model has been used extensively to study the mechanical aspects of muscle fatigue (Bark et al., 1987; Hussain et al., 1989; Shindoh et al., 1990), however E M G power spectral analysis in this model has been overlooked. We feel that this model may represent a unique opportunity to resolve many of the issues regarding E M G power spectral analysis for fatigue detection of the diaphragm. Our results indicate that the application of E M G power spectral analysis for fatigue detection in isolated and exposed preparations will require control of muscle temperature for accurate interpretation. Unrecognized muscle cooling will result in a fall in the E M G fc, and therefore may falsely imply muscle fatigue. Conversely, an elevation of muscle temperature will increase the E M G fc and could potentially obscure the detection of fatigue development. In other experimental models with temperature changes of a smaller magnitude, the effect of muscle temperature on the E M G power spectrum will likely be reduced. h i s u m m a r y , we have shown that changes in the muscle temperature of an isolated diaphragm strip preparation results in significant alterations in the EMG power spectrum; this represents the first demonstration of this effect in the diaphragm. The effect of diaphragm temperature on the E M G signal should be considered when utilizing E M G power spectral analysis in diaphragm investigations, particularly those employing isolated or in situ diaphragm strip preparations.
Acknowledgements.Supported by the Chicago Lung Association and the Parker B. Francis Foundation. We are grateful to Mr. Scott Dornseif for his technical assistance.
References Aldrich, T.K., J.M. Adams, N.S. Arora and D.F. Rochester (1983). Power spectral analysis of the diaphragm electromyogram. J. Appl. Physiol. 54: 1579-1584. Arendt-Nielsen, L. and K.R. Mills (1985). The relationship between mean power frequency of the E M G power spectrum and muscle fiber conduction velocity. Electroencephalogr. Clin. Neurophysiol. 6 0 : 1 3 0 134. Bark, H., G.S. Supinski, J.C. L a m a n n a and S.G. Kelsen (1987). Relationship of changes in diaphragmatic muscle blood flow to muscle contractile activity. J~ Appl. Physiol. 62: 291-299.
250 Bazzy, A.R., J.B. Korten and G.G. Haddad (1986). Increase in electromyogram low-frequency power in nonfatigued contracting skeletal muscle. J. Appl. Physiol. 61: 1012-1017. Bigland-Ritchie, B., E.F. Donovan and C.S. Roussos (1981). Conduction velocity and EMG power spectrum changes in fatigue of sustained maximal efforts. J. Appl. Physiol. 51: 1300-1305. Buchthal, F. and I. Engb~ek (1963). Refractory period and conduction velocity of the striated muscle fiber. Acta Physiol. Stand. 59: 199-220. Doud, J.R. and J.M. Walsb (1992). Effect of fatigue on the relationship between muscle length and center frequency of the EMG power spectrum. Am. Rev. Respir. Dis. 145: A149. Doud, J.R., P.J. Fahey and J.M. Walsh (1992). The effect of diaphragm length on the EMG frequency components. Chest 102: 58S. Eberstein, A. and B. Beattie (1985). Simultaneous measurement of muscle conduction velocity and EMG power spectrum changes during fatigue. Muscle Nerve 8: 768-773. Gross, D., A. Grassino, W.R.D. Ross and P.T. Macklem (1979). Electromyogram pattern of diaphragmatic fatigue. J. Appl. Physiol. 46: 1-7. Hussain, S.N.A., C. Roussos and S. Magder (1989). In situ perfused and innervated left hemidiaphragm preparation. J. Appl. Physiol. 67: 2141-2146. Iaizzo, P.A. and R.E. Poppele (1990). Twitch relaxation of the cat soleus muscle at different lengths and temperatures. Muscle Nerve l 3 :1105-1112. Jarcho, L.W., B. Berman, R.M. Dowben and J . L Lilienthal (1954). Site of origin and velocity of contraction of fibrillary potentials in denervated skeletal muscle. Am. J. Physiol. 178: 129-134. Kadefors, R., E. Kaiser and I. Peters6n (1968). Dynamic spectrum analysis of myo-potentials and with special reference to muscle fatigue. Electro~o'ography 8: 39-74. Kim, M.J., W.S. Druz, J. Danon, W. Machnach and J.T. Sharp (1976). Mechanics of the canine diaphragm. J. Appl. Physiol. 41: 369-382. LindstrOm, L., R. Magnusson and I. Peters6n (1970). Muscular fatigue and action potential conduction velocity changes studied with frequency analysis of EMG signals. Electromyogr. Clin. Neurophysiol. 10: 341-353. Lindstr6m, L., R. Kadefors and I. Petersdn (1977). An electromyographic index for localized muscle fatigue. J. Appl. Physiol. 43: 750-754. LindstrOm, L. and R. Magnusson (1977). Interpretation of myoelectric power spectra: A model and its applications. Proc. IEEE 65: 653-662. Merletti, R., M.A. Sabbahi and C.J. De Luca (1984). Median frequency of the myoelectric signal: effects of muscle ischemia and cooling. Eur. J. Appl. Physiol. 52: 258-265. Mortimer, J.T., R. Magnusson and I. Peters6n (1970). Conduction velocity in ischemic muscle: effect on EMG frequcncy spectrum. Am. J. Ptowiol. 219: 1324-1329. Newman, S., J. Road, F. Bellemare, J.P. Clozel, C.M. Lavigne and A. Grassino (1984). Respiratory muscle length measured by sonomicrometry. J. Appl. Ph~wiol. 56: 753-764. Petrofsky, J.S. and A. Lind (1980). The influence of temperature on the amplitude and frequency components of the EMG during brief and sustained isometric contractions. Eur. J. Appl. Physiol. 44:189-200. Richardson, C.A. and R.A. Mitchell (1982). Power spectral analysis if inspiratory nerve activity in the decerebrate cat. Bra#7 Res. 233: 317-336. Shindoh, C., A. Dimarco, A. Thomas, P. Manubay and G. Supinski (1990). Effect of N-acetylcysteine on diaphragm fatigue. J. Appl. Physiol. 68:2107-2113. Stgtlberg, E. (1966). Propagation velocity in human muscle fibers in situ. Acta Physiol. Scand. 70 (Suppl): 89-95. Stulen, F.B. and C.J. De Luca (1981). Frequency parameters of the myoelectric signal as a measure of conduction velocity. 1EEE Trans. Biomed. Eng. 28:515-523. Walsh, J.M., S. Romano and A. Grassino (1990). Intrabreath change in center frequency of the diaphragmatic EMG power spectrum. Am. Rev. Respir. Dis. 141: A369.
241
RESP 02079
Muscle temperature alters the EMG power spectrum of the canine diaphragm Jon R. Doud and John M. Walsh* Division of Pulmonary and Critical Care Medicine Loyola UniversiO' Chicago Stritch School c?fMedicine and Edward Hines, Jr. V.A. Hospital Maywood. Illinois 60153, USA (Accepted 26 July 1993) Abstract. Electromyogram (EMG) power spectral analysis has been widely used to detect muscle fatigue. Non-fatigue factors that influence the EMG power spectrum must be well understood to ensure accurate interpretation. We explored the effect of muscle temperature on the EMG power spectrum in an isolated canine diaphragm preparation with intact neuro-vascular supply. In 5 animals, diaphragm temperature was continuously recorded with an intramuscular thermocouple, and the EMG signal was recorded with intramuscular wire electrodes. Data were obtained from spontaneous isometric diaphragm contractions while the diaphragm underwent passive cooling from 40 to 30 °C. The EMG power spectrum was determined using fast Fourier transform and each spectrum was characterized by its center frequency (fc). A direct linear relationship was found between diaphragm temperature and the EMG fc. The mean slope of the temperatureEMG fc relationship was 2.9~'~, fc' °C -~. In conclusion, diaphragm temperature is a major determinant of the diaphragmatic EMG power spectrum, and should be controlled when this index is employed in the investigation of muscle fatigue.
EMG, center frequency; Mammals, dog; Muscle, fatigue; Temperature, EMG spectral analysis
Power spectral analysis of the electromyogram (EMG) is widely employed to identify skeletal muscle fatigue (Gross et al., 1979; Kadefors et al., 1968; LindstrOm et at., 1977; Lindstr~m and Magnusson, 1977). Muscle fatigue produces a shift of the EMG power spectrum towards lower frequencies, which can be identified by a decreased center frequency (fc) (Aldrich et at., 1983). Accurate interpretation of the EMG power spectral shift requires a thorough understanding of the factors that significantly influence the EMG power spectrum of both fatigued and non-fatigued muscle. For example, it has recently been appreciated that muscle length is a significant determinant of the EMG fc (Bazzy et al., 1986; Doud and Walsh, 1992; Doud et al., 1992; Walsh et al., 1990). It is thought that this results from an alteration in muscle fiber conduction velocity related to changes in fiber diameter (Arendt-Nielsen and Mills, 1985; Eberstein and Beattie, 1985; LindstrOm et al., 1970; Mortimer et al., 1970; Stulen and DeLuca, 1981).
*Corresponding author. At: Loyola University Medical Center, 2160 South First Avenue Maywood, IL 60153, USA. Tel.: (708) 216-5404; Fax: (708) 216-8897.
242 Muscle temperature is an additional non-fatigue factor that affects muscle fiber conduction velocity (Buchthal and Engb~ek, 1963; Jarcho et al., 1954; Sfftlberg, 1966), and hence may alter the E M G power spectrum. Accordingly, studies in limb muscle (Merletti et al., 1984; Petrofsky and Lind, 1980) have demonstrated a linear correlation between muscle temperature and the E M G fc. However, the relationship between muscle temperature and the E M G fc has never been investigated in the respiratory muscles. Because of morphologic, functional, and biochemical differences between limb skeletal muscle and the diaphragm, demonstration of this relationship rather than extrapolation is necessary. Significant muscle temperature changes may occur in diaphragm strip preparations (particularly with in situ models) commonly used in fatigue studies (Bark et al., 1987; Hussain et al., 1989; Shindoh et al., 1990). Therefore, a full understanding of the effect of temperature on the diaphragmatic E M G power spectrum is critical for the accurate interpretation of E M G power spectral data in this type of preparation. The purpose of the present study was to examine the effect of muscle temperature on the diaphragmatic E M G fc. We utilized a mechanically isolated strip of costal diaphragm developed in situ in an anesthetized, mechanically ventilated canine model with spontaneous diaphragm contractions.
Methods
The protocol was approved by the Institutional Animal Care and Use Committee of Loyola University Stritch School of Medicine. Five healthy mongrel dogs weighing 26-29 kg were anesthetized with a 30 mg.kg ~ bolus of sodium pentobarbital, followed by a continuous infusion of approximately 2 mg'kg-~'h During the surgical preparation the animals were mechanically hyperventilated to induce apnea. The rate of ventilation was reduced during the experimental protocol to allow spontaneous diaphragm contractions. Two liters per minute of supplemental oxygen was administered continuously, and 5 cm H20 of positive end-expiratory pressure was applied. A catheter was placed in the abdominal aorta via the right femoral artery to monitor arterial blood pressure; a second catheter was placed in the right femoral vein for the administration of fluids and sedation. A heating blanket was used to maintain a core temperature of 38 + 0.5 °C. To prolong inspiratory times, bilateral vagotomy was performed on all animals prior to data acquisition.
Animalpreparation.
A 10 cm wide strip of left costal hemidiaphragm was mechanically isolated using a modification of a previously described technique (Kim et al., 1976). Briefly, the left hemithorax was exposed via a wide thoracotomy incision, and the left internal mammary and the lower 5-6 intercostal arteries were ligated. The left costal diaphragm rib attachments were identified through a left lateral abdominal incision. The involved ribs were severed medially and laterally to isolate a 10 cm wide strip of costal diaphragm; costal margin insertions, phrenic nerve, and phrenic vascu-
Left hemidiaphragm isolation.
243 lature were all maintained intact. Electro-cautery was used to create the diaphragm strip by dividing the costal diaphragm from the costal margin to the central tendon at the midline, and 10 cm lateral to the midline (Fig. 1). The position of the central tendon was fixed to a metal rack with a large clamp, taking care not to distort the phrenic vessel blood flow. The free costal margin of the din-
DIAPHRAGM FREE MARGIN
~
I
III 1''' I '1--'-/' i
If
l, //
! J!!,J.JlJ
PHRENIC VESSELS Fig. l. Schematic representation of the in situ diaphragm preparation, with an abdominal view of the mechanically isolated left costal diaphragm strip. The phrenic nerve is intact, and phrenic vasculature is similarly undisturbed. The mechanical fixation of the costal margin and central tendon to the rigid metal rack is not shown.
244 phragm strip was suspended from an isometric strain gauge force transducer fixed to the rack. Plastic wrap and surgical towels were used to limit desiccation and heat loss.
Measurement of diaphragm length. A pair of piezoelectric crystals (2.5 mm diameter) were implanted on the thoracic side of the diaphragm with an inter-crystal distance of approximately 15 ram. The crystals were placed in line with the muscle fibers and secured with 4 - 0 prolene suture followed by cyanoacrylate adhesive for maximal stabilization. The crystals were connected to a 4-channel ultrasonic device (Sonomicrometer model 120, Triton Technology, San Diego, CA). The crystals provided accurate recordings of diaphragm length (Newman et al., 1984) throughout the data acquisition period. The potential temperature-induced error in length measurement is acceptably low (approximately 0.1 °C ~). Muscle length was fixed at the resting in situ length, defined as the maximal length at which the relaxed muscle fails to exert a force. Contractions were isometric in order to exclude the influence of muscle length on the E M G power spectrum.
EMG signal. Two teflon-coated intramuscular wire electrodes (0.005" diameter,0.5" coiled length) were positioned at the distal margin of the diaphragm strip to record the E M G signal. Wires were placed in line with the piezoelectric crystals, with a distance of approximately 10 mm separating the 2 electrodes. The electrodes were secured with 4 - 0 prolene suture and cyanoacrylate adhesive to eliminate movement. The E M G signal was amplified (Grass, model P511K) and band pass filtered (CWE, model BPF-935) at 20-500 Hz.
Diaphragm temperature.
A one-inch 33 gauge stainless steel thermocouple (Model HYP-0, Omega Engineering Inc), calibrated from 20-40 ° C, was implanted within the muscle on the thoracic side of the diaphragm preparation to continuously record diaphragm temperature. The thermocouple was positioned within 2 cm of the E M G electrodes (Fig. 1) and secured with fine suture. An external radiant heat source (placed no closer than 40 cm) was used as necessary to maintain desired diaphragm temperature. Laboratory room temperature was kept at 25 ° C.
Experimentalprotocol.
Data from ten sequential spontaneous breaths were acquired at approximately one degree intervals while diaphragm temperature fell passively towards room temperature. The diaphragm temperature ranged from 40 to 30 ° C. To address the possibility of irreversible muscle injury induced by the experimental protocol, the diaphragm was rewarmed in one animal. Muscle rewarming with a heat lamp was performed from 30 to 40 °C, at a similar rate as cooling. E M G signal and temperature data were again acquired at 1 °C intervals. To investigate the mechanism responsible for the temperature-induced power spectral shift, compound diaphragm action potentials (CDAP) were evoked in one animal via phrenic nerve stimulation. A self-retaining bi-polar phrenic nerve stimulator was placed circumferentially around the phrenic nerve at the level of the heart. The phrenic
245 nerve was supra-maximally stimulated (duration of 0.25 ms) at the two temperature extremes (5 stimulations at each temperature), and the resultant CDAPs were acquired for power spectral analysis.
Data acquisition and analysis.
The E M G signal and muscle temperature were recorded on a Gould chart recorder (Series 2800, 8-channel) and simultaneously digitized at a sampling frequency of 2000 Hz via the analog/digital board in an IBM Model 70 computer and stored on hard disk (DataSponge, BioScience Analysis Software Ltd). EMG frequency component analysis was modeled after that reported by Richardson and Mitchell (1982). Each contraction was divided into 1024 ms segments of data, with an overlap of 768 ms (75 ~o common data) between successive segments. The number of segments was determined by the duration of diaphragm contraction. After EMG signal pretreatment by trend removal and application of a Hanning window, the E M G fc was determined for each data segment by applying a fast Fourier transform (VGA; Inspirational Software, 1990). The parameter used to characterize the power spectra and follow the change in signal frequency content was the center (or centroid) frequency, defined as the weighted mathematical mean frequency of the power spectrum: fc = .Y-,AiFi/ZAi, where Ai = amplitude associated with frequency F i. Temperature and E M G fc values fi'om each 1 °C interval were averaged and plotted. To facilitate comparison of data between animals, E M G fc values were normalized with respect to values taken at 38 °C. The evoked CDAPs elicited via phrenic nerve stimulation were analyzed in a similar manner to the spontaneous E M G signal. Fast Fourier transform was applied to the CDAP data, and each resulting power spectrum was represented by its fc.
Statistical analysis. Linear regression analysis of the relationship between diaphragm temperature and EMG fc was performed on the data from each animal. Slopes of the resulting regression lines (in ~ofc.°C 1) were calculated for each animal, and significance determined at a P < 0 . 0 5 level.
Results
A reduction in diaphragm temperature from 40 to 30 °C resulted in a significant reduction in E M G fc in each animal. The mean decrease in E M G fc across this temperature range exceeded 25 ~o, and was similar in all animals (Fig. 2). Linear regression analysis determined significant slopes in each individual animal, ranging from 2.4 to 3.4~o fc' °C 1. Correlation coefficients calculated from all data points from each animal ranged from 0.79 to 0.95 (Table 1). Group data displayed in Fig. 2 had a significant slope of 2.9~/o fc' °C -t and an r value of 0.95. In one animal, rewarming of the cooled diaphragm produced an increase in the E M G
246
115 ~9
105 ~> 0
O3
L-v3
95
~2 85
75 S o
•
/ o
65 28
o
I
I
I
I
I
I
I
30
32
34
36
38
40
42
temperature (°C) Fig. 2. The relationship between the EMG center frequency (fc) and diaphragm temperature in the five animals studied. The regression line is superimposed on the data points (r = 0.97). The EMG fc is expressed as a percentage of the value obtained at 38 °C. Data points represent mean data for a one degree temperature interval; data derived from each animal are represented by a unique symbol.
fc. After r e w a r m i n g , the E M G fc values were e q u i v a l e n t to those o b t a i n e d at the start o f the cooling p r o t o c o l (Fig. 3). T h e C D A P d u r a t i o n of the w a r m e d m u s c l e (40 ° C ) was c o m p a r e d to that o f the cooled m u s c l e (30 ° C ) in o n e a n i m a l . T h e 5 C D A P s at each t e m p e r a t u r e were superi m p o s a b l e . R e p r e s e n t a t i v e w a v e f o r m s from the two t e m p e r a t u r e extremes are s h o w n in Fig. 4. T h e C D A P d u r a t i o n was a p p r o x i m a t e l y 25 ~o longer u n d e r the colder c o n ditions. This p r o l o n g a t i o n o f the action p o t e n t i a l in cooled m u s c l e was a s s o c i a t e d with an a p p r o x i m a t e l y 3 0 % r e d u c t i o n in E M G fc o f the C D A P (103 H z to 74 Hz). TABLE 1 Slopes and regression coefficients for temperature-EMG fc relationship. Animal
Slope (°ofc.°C i)
r value
1 2 3 4 5
3.2 2.6 2.4 3.4 2.8
0.87 0.79 0.90 0.95 0.95
fc, center frequency.
247
120
...,.-'''"
ii0
[] ,...'''' ..~" i00 0
cO c~
90
o
80
...../ .,..." .~ ~
"'''''''''
70
60 28
30
I
I
I
[
32
34
36
38
40
42
temperature (°C) Fig. 3. Data obtained during diaphragm cooling and rewarming in one animal. The EMG center frequency (fc) is expressed as a percentage of the value obtained at 38 °C. Data points represent mean data for a one degree temperature interval. • = Data from cooling protocol; [] = data from rewarming protocol.
Systemic arterial pH and Pco, were similar at the initiation and conclusion of each experimental protocol. In all animals, change in arterial pH was < 0.05 units, and Pco, < 4 Torr.
Discussion
The canine diaphragmatic EMG power spectrum is strongly influenced by changes in intramuscular temperature. We found a significant linear relationship between the temperature and the EMG fc (mean slope 2.9~o fc. °C 1). The effect of temperature on the EMG power spectrum is probably due to a temperature-induced alteration in muscle fiber conduction velocity. Jarcho et al. (1954) described a significant positive relationship between temperature and conduction velocity between 27 and 40 °C in the rat anterior gracilis muscle. This strong association between temperature and muscle fiber conduction velocity was confirmed in the human biceps by St51berg (1966). In addition, Buchthal and Engb~ek (1963) demonstrated approximately 3.5 ~o change in conduction velocity per °C in the frog sartorius muscle, consistent with the temperature-induced E M G fc changes observed in the present study. Slowing of the muscle fiber conduction velocity prolongs the motor unit action
248
,,- - A = 4 0 o, EMG f c = 1 0 3 *
0
* = 3 0 °, EMG f c = 7 4
Hz Hz
I
[
I
5
i0
15
time (msec) Fig. 4. Compound diaphragm action potentials evoked via supra-maximal phrenic nerve stimulation at two different temperatures in one animal. The action potential duration was markedly prolonged at 30 °C compared with 40 °C; this was associated with an approximately 30% reduction in EMG fc (74 Hz and 103 Hz, respectively).
potential duration (Iaizzo and Poppele, 1990), resulting in a decreased E M G fc. In the present investigation C D A P duration was markedly prolonged at 30 ° compared with 40 ~C (Fig. 4). This observation is consistent with a temperature-related slowing of the muscle fiber conduction velocity ultimately manifested as a decreased E M G fc. This mechanism is supported by the findings of Bigland-Ritchie et al. (1981), where cooling of the adductor pollicis muscle resulted in slowing of the muscle fiber conduction velocity, with an associated reduction in the E M G high/low ratio (a parameter proportional to the E M G fc). The magnitude of E M G fc alteration caused by changes in temperature in our study is similar to that observed in non-respiratory muscles. Petrofsky and Lind (1980) studied the effect of cooling on the brachioradialis E M G ; the resulting E M G fc was shown to decrease by approximately 4 ?J~,. ° C a over a similar temperature range to that used in the present study. Additionally, Merletti et al. (1984) investigated cooling of the first dorsal interosseous muscle, and found the slope of the t e m p e r a t u r e - E M G fc regression line to be approximately 3g; f c ' ° C When the muscle was rewarmed to its initial temperature, the E M G fc values approximated the values acquired prior to cooling. In addition, the E M G fc followed temperature in a linear fashion during rewarming (Fig. 3). The difference between
249 cooling and rewarming fc values in the 30 to 35 °C range is likely due to a lag in rewarming at the E M G electrode site. This may have occurred as a result from relatively rapid rewarming over this temperature range. These observations imply that the electrode position remained unchanged throughout the experiment, and that muscle cooling did not induce irreversible tissue injury. Muscle length changes throughout the study were minimized in order to diminish the influence of muscle length on the E M G fc. Diaphragm length was fixed, and shortening was less than 5°,o in all experiments, and of similar magnitude at both high and low temperatures. The implications of this study may be most important in animal models employing isolated and exposed muscles where large temperature changes may occur. This model has been used extensively to study the mechanical aspects of muscle fatigue (Bark et al., 1987; Hussain et al., 1989; Shindoh et al., 1990), however E M G power spectral analysis in this model has been overlooked. We feel that this model may represent a unique opportunity to resolve many of the issues regarding E M G power spectral analysis for fatigue detection of the diaphragm. Our results indicate that the application of E M G power spectral analysis for fatigue detection in isolated and exposed preparations will require control of muscle temperature for accurate interpretation. Unrecognized muscle cooling will result in a fall in the E M G fc, and therefore may falsely imply muscle fatigue. Conversely, an elevation of muscle temperature will increase the E M G fc and could potentially obscure the detection of fatigue development. In other experimental models with temperature changes of a smaller magnitude, the effect of muscle temperature on the E M G power spectrum will likely be reduced. h i s u m m a r y , we have shown that changes in the muscle temperature of an isolated diaphragm strip preparation results in significant alterations in the EMG power spectrum; this represents the first demonstration of this effect in the diaphragm. The effect of diaphragm temperature on the E M G signal should be considered when utilizing E M G power spectral analysis in diaphragm investigations, particularly those employing isolated or in situ diaphragm strip preparations.
Acknowledgements.Supported by the Chicago Lung Association and the Parker B. Francis Foundation. We are grateful to Mr. Scott Dornseif for his technical assistance.
References Aldrich, T.K., J.M. Adams, N.S. Arora and D.F. Rochester (1983). Power spectral analysis of the diaphragm electromyogram. J. Appl. Physiol. 54: 1579-1584. Arendt-Nielsen, L. and K.R. Mills (1985). The relationship between mean power frequency of the E M G power spectrum and muscle fiber conduction velocity. Electroencephalogr. Clin. Neurophysiol. 6 0 : 1 3 0 134. Bark, H., G.S. Supinski, J.C. L a m a n n a and S.G. Kelsen (1987). Relationship of changes in diaphragmatic muscle blood flow to muscle contractile activity. J~ Appl. Physiol. 62: 291-299.
250 Bazzy, A.R., J.B. Korten and G.G. Haddad (1986). Increase in electromyogram low-frequency power in nonfatigued contracting skeletal muscle. J. Appl. Physiol. 61: 1012-1017. Bigland-Ritchie, B., E.F. Donovan and C.S. Roussos (1981). Conduction velocity and EMG power spectrum changes in fatigue of sustained maximal efforts. J. Appl. Physiol. 51: 1300-1305. Buchthal, F. and I. Engb~ek (1963). Refractory period and conduction velocity of the striated muscle fiber. Acta Physiol. Stand. 59: 199-220. Doud, J.R. and J.M. Walsb (1992). Effect of fatigue on the relationship between muscle length and center frequency of the EMG power spectrum. Am. Rev. Respir. Dis. 145: A149. Doud, J.R., P.J. Fahey and J.M. Walsh (1992). The effect of diaphragm length on the EMG frequency components. Chest 102: 58S. Eberstein, A. and B. Beattie (1985). Simultaneous measurement of muscle conduction velocity and EMG power spectrum changes during fatigue. Muscle Nerve 8: 768-773. Gross, D., A. Grassino, W.R.D. Ross and P.T. Macklem (1979). Electromyogram pattern of diaphragmatic fatigue. J. Appl. Physiol. 46: 1-7. Hussain, S.N.A., C. Roussos and S. Magder (1989). In situ perfused and innervated left hemidiaphragm preparation. J. Appl. Physiol. 67: 2141-2146. Iaizzo, P.A. and R.E. Poppele (1990). Twitch relaxation of the cat soleus muscle at different lengths and temperatures. Muscle Nerve l 3 :1105-1112. Jarcho, L.W., B. Berman, R.M. Dowben and J . L Lilienthal (1954). Site of origin and velocity of contraction of fibrillary potentials in denervated skeletal muscle. Am. J. Physiol. 178: 129-134. Kadefors, R., E. Kaiser and I. Peters6n (1968). Dynamic spectrum analysis of myo-potentials and with special reference to muscle fatigue. Electro~o'ography 8: 39-74. Kim, M.J., W.S. Druz, J. Danon, W. Machnach and J.T. Sharp (1976). Mechanics of the canine diaphragm. J. Appl. Physiol. 41: 369-382. LindstrOm, L., R. Magnusson and I. Peters6n (1970). Muscular fatigue and action potential conduction velocity changes studied with frequency analysis of EMG signals. Electromyogr. Clin. Neurophysiol. 10: 341-353. Lindstr6m, L., R. Kadefors and I. Petersdn (1977). An electromyographic index for localized muscle fatigue. J. Appl. Physiol. 43: 750-754. LindstrOm, L. and R. Magnusson (1977). Interpretation of myoelectric power spectra: A model and its applications. Proc. IEEE 65: 653-662. Merletti, R., M.A. Sabbahi and C.J. De Luca (1984). Median frequency of the myoelectric signal: effects of muscle ischemia and cooling. Eur. J. Appl. Physiol. 52: 258-265. Mortimer, J.T., R. Magnusson and I. Peters6n (1970). Conduction velocity in ischemic muscle: effect on EMG frequcncy spectrum. Am. J. Ptowiol. 219: 1324-1329. Newman, S., J. Road, F. Bellemare, J.P. Clozel, C.M. Lavigne and A. Grassino (1984). Respiratory muscle length measured by sonomicrometry. J. Appl. Ph~wiol. 56: 753-764. Petrofsky, J.S. and A. Lind (1980). The influence of temperature on the amplitude and frequency components of the EMG during brief and sustained isometric contractions. Eur. J. Appl. Physiol. 44:189-200. Richardson, C.A. and R.A. Mitchell (1982). Power spectral analysis if inspiratory nerve activity in the decerebrate cat. Bra#7 Res. 233: 317-336. Shindoh, C., A. Dimarco, A. Thomas, P. Manubay and G. Supinski (1990). Effect of N-acetylcysteine on diaphragm fatigue. J. Appl. Physiol. 68:2107-2113. Stgtlberg, E. (1966). Propagation velocity in human muscle fibers in situ. Acta Physiol. Scand. 70 (Suppl): 89-95. Stulen, F.B. and C.J. De Luca (1981). Frequency parameters of the myoelectric signal as a measure of conduction velocity. 1EEE Trans. Biomed. Eng. 28:515-523. Walsh, J.M., S. Romano and A. Grassino (1990). Intrabreath change in center frequency of the diaphragmatic EMG power spectrum. Am. Rev. Respir. Dis. 141: A369.