November 1979
TheJournalofPEDIATRICS
793
The consequences of diaphragmatic muscle fatigue in the newborn infant We have previously demonstrated that diaphragmatic muscle fatigue can be diagnosed in infants from spectral frequency analysis of the surface diaphragmatic eleetromyogram. This requires a digital computer, but the analysis takes several days. Spectral frequency changes, however, can be accurately reflected by band pass filtering and expressing the ratio of high-frequency power to low-frequency power. A fall in this ratio of greater than 20% indicates muscle fatigue. Using a simple analog device to obtain this ratio permits the results to be immediately available; we, have used this method to study weaning from mechanical ventilators in ten infants. With a successful weaning step there is no significant change in the ratio, whereas an unsuccessful weaning step invariably leads to a decrease in the ratio of greater than 20%, which precedes C02 retention and clinical deterioration. These data indicate that diaphragmatic muscle fatigue plays an important role in the infant's response to lung disease9 Monitoring of the high~low frequency ratio may be helpful in weaning infants from assisted ventilation.
Nestor Muller, M.D., George Volgyesi, P. Eng., M. Heather Bryan, M.D.,* and A. Charles Bryan, M.D., Toronto, Ont., C a n a d a
THE WORK OF BREATHING is often high in lung disease because the respiratory muscles have to overcome high resistance a n d / o r low compliance to expand the lungs. The work is even higher in infants because the highly compliant rib cage does not provide a stable platform for the muscles. When the diaphragm attempts to create large negative pleural pressures, the rib cage is sucked in (paradoxical respiration) and thus a substantial fraction of the force of the diaphragm is dissipated in distorting the rib cage rather than effecting volume exchange?' 2 Tachypnea is characteristic of lung disease in infancy, and the faster the repetition rate of muscle contraction, the shorter is its endurance time? Histochemically the infant diaphragm is poorly equipped to sustain high work loads, Keens et al 4 have shown that there are relatively few fatigue,resistant high-oxidative fibers in the infant, particularly in the preterm infant's diaphragm. This combination of high work, rapid breathing, and low-oxidative capacity suggests that respiratory muscle From the Research Institute, The Hospital for Sick Children. *Reprint address: The Hospitalfor Sick Children, 555 University Ave., Toronto, Ont., Canada, M5G 1X8.
0022-3476/79/110793+05500.50/0 9 1979 The C. V. Mosby Co.
9fatigue must play a major role in the infant's response to respiratory disease9 Muscle fatigue produces characteristic changes in the electromyogram of skeletal muscles. A fatiguing load produces a fall in the high-frequency power and rise in the low-frequency power of the E M G prior to exhaustion?' ~
See related article, part 2, p. 844. Abbreviations used CPAP: continuous positive airway pressure electromyogram EMG: high/low frequency ratio H/L: intermittent mandatory ventilation IMV: arterial carbon dioxide tension Paeo.~ PETco2 end tidal carbon dioxide tension transcutaneous oxygen tension tCpo2 electrocardiogram ECG: Gross et al 7 have shown that these changes can be detected from surface electrodes recording the diaphragmatic EMG. We have demonstrated that the diaphragmatic EMG spectrum of the infant is the same as that of adult skeletal muscle, and that power spectral shifts character-
Vol. 95, No. 5, part 1, pp. 793-797
Muller et aL
794
The Journal of Pediatrics November 1979
Table
Infant gestational age (wk) 34 26 27 26 36 27 28 27
25 26
Study age (days) 5 31 3 4 12 30 1 18 5 17 17 17 15 3 3
Control (ram Hi)* , Po 2 [ Pco~ 46 73 49 42 44 53 50 53 52 81 82 72 49 50 54
43 42 43 38 44 45 33 40 43 42.8 • 43.5 • 44.5 + 43.5 • 44.8 • 45.0 •
0.8 1.1 0.7 1.1 0.8 1.8
Ventilatory assistance CPAP,-~ CPAP~ CPAP~ --->CPAP~ IMV~o--~ CPAP:. CPAP~ --->CPAP~ !MV~ ---->CPAP~ CPAP~ --->CPAP~ IMV~5~ CPAP5 IMV~ --~ CPAP3 IMV~o~ CPAP~ IMVI0 ~ CPAP, IMV~0--~CPAP~ IMV~0--->CPAP, IMV~--~ CPAP, CPAP,---> CPAP~ CPAP, ~ CPAP~
Time to fatigue (rain) 6 -40 1 31 22 1 13 1 6 5 3
Maximal .fall in H / L (%) 9.6 11.4 12.8 24.0 12.1 70.0 35.7 24.5 42.5 37.5 31.1 65.5 29.2 47.7 63.4
After H/L fall (ram Hi)* I
Po2 66 54 78 54 45 60 51
82 80 65 53 48 52
I Pco~_ 46 40 41 51 38 56 59 Apnea Apnea 54.7 • 52.1 • 53.5 • 49.7 • 56.4 • 59.4 •
Recovery time of H/L (rain) -3
1.3 1.2 1.6 1.4 1.1 2.4
32 5 1 6 1 1 4 1 3 5
CPAP ~ Continuous positive airway pr~ssure in cm H._,O;IMV = intermittent mandatory ventilation in breaths/minute. *Pa%, Pac% in infants 1-7, LP%, PETe%(mean _+SD of breaths) in the others.
istic of fatigue can be detected in normal infants during R E M sleep and in infants with cardiopulmonary disease? These studies were done using a digital computer. Since the results were not immediately available, they were not clinically useful. W e have therefore developed a simple analog device to indicate immediately and continuously the spectral frequency changes in the diaphragm of infants during weaning from mechanical ventilation.
METHODS We monitored ten intubated premature newborn infants during 15 studies w h e n they were being switched from IMV to CPAP or w h e n CPAP was reduced. Their gestational age and age at the time of the study are shown in Table I. The study included infants with respiratory distress syndrome, patent ductus arteriosus and congestive heart failure, and pneumonia. Respiratory movements were monitored using anteroposterior magnetometer pairs placed in the midline, one at the level of the nipple, and the other 1 cm above the umbilicus.' In seven studies arterial Pac% was measured; in the others end-tidal Pco~ was monitored continuously at the endotracheal tube using a Beckman LB~ CO~ analyzer adapted for the low respiratory flow rates present in newborn babies. 8 The diaphragmatic E M G was recorded using Beckman surface electrodes in the right sixth and seventh interspaces or subcostally, between the mid-clavicular and midaxillary lines. 2 These electrodes were secured to t h e skin with double adhesive tape rings. The accepted
interelectrode impedance was below 20 kohm. The E M G signal was led to a hfgh input impedance, high c o m m o n mode rejection E M G preamplifier* placed in the incubator close to the infant. F r o m there it was taken to a Honeywell Accudata 134A biomedical amplifier* with band pass filters of 15-1000 Hz for further conditioning. To analyze the diaphragmatic frequency spectrum it is necessary to gate the superimposed E C G signal in a method similar to the one described by Prechtl et aP and ourselves? ~ Briefly, the E M G signal was introduced into two channels. In channel one it was led to a so-called bucket brigade circuit which delayed this signal for a period of up to 50 milliseconds, In channel 2, the E C G artifact in the E M G signal acted as a trigger for a pulse generator, the output ~ which reached the delayed E M G signal in channel one just prior to the cardiac artifact and blocked out any signal for the duration o f the pulse. The blocking pulse had a variable time setting so that it was possible to suppress the entire E C G signal. The gated E M G signal was then fed into two-band pass filters: a high-frequency one (150-350 Hz) and a low-frequency one (20-,46.7.Hz). These band pass filters have a similar frequency, range to those used by Kadefors et al 5 and by Gross et al. 7 The rectified output of each filter passed through a similar electronic switch into an averager circuit. W h e n the switch was closed the averager developed a voltage proportional to the average value of the band o f frequencies. W h e n the switch opened (during the *Honeywell Test Instruments Division, P.O. Box 5227, Denver, CO 80217.
Volume 95 Number 5, part 1
Diaphragmatic muscle fatigue
795
HIGH/LOW FREQUENCY ,(percent normal)
::::::
: ?
:::
,
::
;:
--:,
:
::::;;:;:
Figure. The tracing shows the H/L frequency ratio monitored from the diaphragmatic EMG, the PETeo2,rib cage and abdominal movement. The infant has been weaned from IMV to CPAP alone and the results show the progressive fall in the H/L ratio followed by rising PETe%.Recovery of the H/L ratio is rapid when IMV is restarted (arrow), while PExc% shows a slower return to normal.
QRS complex) the averager maintained its output at the level that existed just prior to the opening of the switch. The gated and averaged output of the high-frequency filter was then electronically divided by the similarly processed output of the low-frequency filter to provide a continous high-to-low frequency ratio, which was shown on a visual display. Since there are small changes in the frequency spectrum at the beginning of inspiration," the analysis of the H / L ratio was done ,at midinspiration as judged from the magnetometer tracing when the baby was breathing quietly without gross body movements. Control values for the H / L ratio were obtained from the mean of 10 breaths when the infants were stable just before any changes were made in the ventilatory mode. In order to assess the stability of this control value, 10 to 20 sections of 10 breath intervals were analyzed while the infant was stable. Following ventilatory mode changes, similar 10 breath intervals were analyzed. All mean values were expressed as a percentage of the initial control H / L ratio. The H / L ratio measurements did not influence clinical management. The usual practices of nursing care and r e p e a t e d blood gas measurements were continued. Because diaphragmatic muscle fatigue occurs rapidly in the presence of hypoxia, 12 oxygenation was monitored
using a transcutaneous Po~ electrode (Radiometer). Initially the infants were stable m inspired oxygen concentrations ranging from 25 to 40%. During the studies the inspired oxygen was altered in order to maintain a stable transcutaneous Po2 reading. None of the infants required increases of more than 20% inspired oxygen. All data were recorded on a Hewlett-Packard strip chart recorder or a Brush pen recorder. Statistical analysis was done using the unpaired t test. RESULTS During the 15 studies while the infants were stable and breathing on assisted ventilation, the mean H/L ratio did not vary more than 10% (the coefficient of variation ranged from 1.1 to 4.5%). When assisted ventilation was altered, we observed two different patterns. In four infants, two going from IMV to CPAP and two in whom CPAP was reduced, the H / L ratio did not fall more than 12.8% from the control value at any time throughout the studies, which continued for two to five hours (Table). These infants had normal blood gas values throughout the study and remained stable for at least 12 hours afterward. In 11 studies, after assisted ventilation had been altered, the H / L ratio fell progressively to levels of 20% or more
796
Muller et al.
below the control value (P < 0.001). The time observed for the fall to occur varied from infant to infant (Table). There was a significant rise in the arterial Pco~ in three infants when blood gas values were determined after the development of fatigue. Two babies developed severe apneic spells and were put on assisted ventilation before arterial blood gases were obtained (Table)~ In the six studies in which end-tidal CO2 was monitored, there was no significant change in PETCO~ when the H / L ratio remained within 10% of the control. When the H / L ratio fell 20% or more below the control value, there was a progressive and significant rise in PETCO2 (P < 0.001). The maximum levels are listed in the Table. One study is shown in the Figure. There was a steady fall in the H / L ratio with a slower rise in PETco~ after weaning from IMV. The time lapse between the H / L ratio fall of over 20% and the rise in PETe% varied in the infants from 0.5 to 10 minutes. The time for recovery from fatigue, after assisted ventilation had been reinstituted, varied in the infants studied (Table). The longest recovery time of 32 minutes occurred in an infant with the largest fall (70%) in the H / L ratio (Table). DISCUSSION The changes in the E M G spectrum with muscle fatigue have been known since the beginning of this century. 1~ These changes are best described by spectral frequency analysis, using either multiple band pass filters or fast Fourier transform analysis. The shift in the E M G power to lower frequencies with a fatiguing load is well established?' ~ The reasons for these changes are still conjectural. Mortimer et al TM showed that the changes were related to a decrease in the conduction velocity within the muscle fiber and suggested that this was due to the accumulation of metabolic end products within the muscle. However, when a muscle is presented with a fatiguing load there is an almost immediate increase in love-frequency power and a gradual decrease in the high-frequency power? Thus the changes predict eventual exhaustion, and the application of this technique to the diaphragm was a major step. 7 We established that the spectrum of the infant diaphragm was the same as for any other skeletal muscle, 2 and that the essential information could be obtained from a high and a low band pass ratio, without going to the complexity of spectral frequency analysis.~ There are some limitations to the use of diaphragmatic EMG monitoring of fatigue. The diagnosis of fatigue cannot be made by inspection of the frequency spectrum alone, but depends on observing a changae in the spectrum to one of a lower-power density. It is important, therefore, to obtain a control ratio when the infant is stable on the
The Journal of Pediatrics November 1979
ventilator, and it can be reasonably assumed that the diaphragm is not fatigued. A further problem is the relationship of hypoxia to fatigue. The endurance of the diaphragm is markedly reduced by hypoxia. 12Thus during weaning, when there is unexpected hypoxia, there is a rapid fall in the H / L ratio. This occurred during this study in one infant who became severely hypoxic during feeding. Therefore it is preferable to have continuous monitoring of the Po2 with a transcutaneous electrode. Within these limitations, our results suggest that monitoring of the diaphragmatic E M G may be of benefit in directing the weaning process. E M G changes precede the deterioration of blood gas values, give a continuous output, and are technically simpler than monitoring end tidal CO2. There are compelling reasons for discontinuing mechanical ventilation in infants as quickly as possible. Applying high pressures to a lung with a nonuniform compliance must overdistend the most compliant (presumably the healthiest) areas of lung. The immediate risk is pneumothorax or pneumomediastinum; the long-term risk is bronchopulmonary dysplasia. In addition, the diaphragm decreases activity during mechanical ventilation. An inactive muscle rapidly loses oxidative capacity and hence fatigues more readily.1~. 16 Thus the least ventilatory assistance that is provided, the better for both the lung and the diaphragm. Currently, however, the least ventilatory assistance required is judged by pushing the weaning process to the point of clinical deterioration. As diaphragmatic E M G evidence precedes clinical deterioration, it may permit faster weaning with fewer crises. REFERENCES
1. Knill RL, Andrews W, Bryan AC, and Bryan MH: Respiratory load compensation in infants, J Appl Physio140:357, 1976. 2. Muller N, Gulston (3, Cade D, Whitton J, Froese AB, Bryan MH, and Bryan ~ : Diaphragmatic muscle fatigue in the newborn, J Appl Physiol 46:688, 1979. 3. MolbechS: Average percentage force at repeated maximal isometric muscle contractions at different frequencies, in Communications from The Testing and Observations Institute of Danish National Association for Infantile Paralysis, No. 16, 1963. 4. Keens TG, Bryan AC, Levison H, and Ianuzzo CD: Development of fatigue-resistant muscle fibers in human ventllatory muscles, J Appl Physlol 44:909, 1978. 5. Kad~fors R, Kaiser E, and Petersen I: Dynamic spectrum analysis of myopotentials and special reference to muscle fatigue, Electromyography 8:39, 1968. 6. Kogi K, and Hakamada T: Slowing of surface electromyogram and muscle strength in muscle fatigue, Res Inst Sci Lab 60:27, 1962. 7. Gross D, Grassino A, Ross WD, and Macklem PT: Electro-
Volume 95 Number 5, part 1
8.
9.
10. 11.
Diaphragmatic muscle fatigue
myogram pattern of diaphragmatic fatigue, J'Appl Physiol 46:1, 1979. Rigatto H, and Brady JP: Periodic breathing and apnea in pre-term infants, I. Evidence for hypoventilation possibly due to central respiratory depression, Pediatrics 50:202, 1972. Prechtl HFR, Van Eykern LA, and O'Brien M J: Respiratory muscle EMG in newborns: a non-intrusive method, Early Hum Dev 1:265, 1977. Muller N, Volgyesi G, Becket L, Bryan MH, and Bryan AC: Diaphragmatic muscle tone, J Appl Physiol (in press). Schweitzer TW, Fitzgerald JW, Bowden JA, and LynneDavies P: Spectral analysis of human inspiratory diaphragmatic electromyograms, J Appl Physiol 46:152, 1979.
797
12. Roussos CS, and Macldem PT: Diaphragmatic fatigue in man, J Appl Physiol 43:189, 1977. 13. Cobb S, and Forbes A: Electromyographic studies of muscular fatigue in man, Am J Physiol 65:234, 1923. 14. Mortimer JT, Magnusson R, and Petersen I: Conduction velocity in ischemic muscle: effect on EMG frequency spectrum, Am J Physiol 219:1324, 1970. 15. Booth FW: Time cot~rse of muscular atrophy during immobilization of hindlimbs in rats, J Appl Physiol 43:656, 1977. 16. Rifenberick DH, Gamble JG, and Max SR: Response of mitochondrial enzymes to decreased muscular activity, Am J Physiol 225:1295, 1973.
Copyright information The appearance of a code at the bottom of the first page of an original article in this JOURNAL indicates the copyright owner's consent that copies of the article may be made for personal or internal use, or for the personal or internal use of specific clients. This consent is given on the condition, however, that the copier pay the stated per copy fee through the Copyright Clearance Center, Inc., P.O. Box 765, Schenectady, N.Y. 12301, (518) 374-4430, for copying beyond that permitted by Sections 107 or 108 of the U.S. Copyright Law. This consent does not extend to other kinds of copying, such as copying for general distribution, for advertising or promotional purposes, for creating new collective works, or for resale.
TheJournalofPEDIATRICS
793
The consequences of diaphragmatic muscle fatigue in the newborn infant We have previously demonstrated that diaphragmatic muscle fatigue can be diagnosed in infants from spectral frequency analysis of the surface diaphragmatic eleetromyogram. This requires a digital computer, but the analysis takes several days. Spectral frequency changes, however, can be accurately reflected by band pass filtering and expressing the ratio of high-frequency power to low-frequency power. A fall in this ratio of greater than 20% indicates muscle fatigue. Using a simple analog device to obtain this ratio permits the results to be immediately available; we, have used this method to study weaning from mechanical ventilators in ten infants. With a successful weaning step there is no significant change in the ratio, whereas an unsuccessful weaning step invariably leads to a decrease in the ratio of greater than 20%, which precedes C02 retention and clinical deterioration. These data indicate that diaphragmatic muscle fatigue plays an important role in the infant's response to lung disease9 Monitoring of the high~low frequency ratio may be helpful in weaning infants from assisted ventilation.
Nestor Muller, M.D., George Volgyesi, P. Eng., M. Heather Bryan, M.D.,* and A. Charles Bryan, M.D., Toronto, Ont., C a n a d a
THE WORK OF BREATHING is often high in lung disease because the respiratory muscles have to overcome high resistance a n d / o r low compliance to expand the lungs. The work is even higher in infants because the highly compliant rib cage does not provide a stable platform for the muscles. When the diaphragm attempts to create large negative pleural pressures, the rib cage is sucked in (paradoxical respiration) and thus a substantial fraction of the force of the diaphragm is dissipated in distorting the rib cage rather than effecting volume exchange?' 2 Tachypnea is characteristic of lung disease in infancy, and the faster the repetition rate of muscle contraction, the shorter is its endurance time? Histochemically the infant diaphragm is poorly equipped to sustain high work loads, Keens et al 4 have shown that there are relatively few fatigue,resistant high-oxidative fibers in the infant, particularly in the preterm infant's diaphragm. This combination of high work, rapid breathing, and low-oxidative capacity suggests that respiratory muscle From the Research Institute, The Hospital for Sick Children. *Reprint address: The Hospitalfor Sick Children, 555 University Ave., Toronto, Ont., Canada, M5G 1X8.
0022-3476/79/110793+05500.50/0 9 1979 The C. V. Mosby Co.
9fatigue must play a major role in the infant's response to respiratory disease9 Muscle fatigue produces characteristic changes in the electromyogram of skeletal muscles. A fatiguing load produces a fall in the high-frequency power and rise in the low-frequency power of the E M G prior to exhaustion?' ~
See related article, part 2, p. 844. Abbreviations used CPAP: continuous positive airway pressure electromyogram EMG: high/low frequency ratio H/L: intermittent mandatory ventilation IMV: arterial carbon dioxide tension Paeo.~ PETco2 end tidal carbon dioxide tension transcutaneous oxygen tension tCpo2 electrocardiogram ECG: Gross et al 7 have shown that these changes can be detected from surface electrodes recording the diaphragmatic EMG. We have demonstrated that the diaphragmatic EMG spectrum of the infant is the same as that of adult skeletal muscle, and that power spectral shifts character-
Vol. 95, No. 5, part 1, pp. 793-797
Muller et aL
794
The Journal of Pediatrics November 1979
Table
Infant gestational age (wk) 34 26 27 26 36 27 28 27
25 26
Study age (days) 5 31 3 4 12 30 1 18 5 17 17 17 15 3 3
Control (ram Hi)* , Po 2 [ Pco~ 46 73 49 42 44 53 50 53 52 81 82 72 49 50 54
43 42 43 38 44 45 33 40 43 42.8 • 43.5 • 44.5 + 43.5 • 44.8 • 45.0 •
0.8 1.1 0.7 1.1 0.8 1.8
Ventilatory assistance CPAP,-~ CPAP~ CPAP~ --->CPAP~ IMV~o--~ CPAP:. CPAP~ --->CPAP~ !MV~ ---->CPAP~ CPAP~ --->CPAP~ IMV~5~ CPAP5 IMV~ --~ CPAP3 IMV~o~ CPAP~ IMVI0 ~ CPAP, IMV~0--~CPAP~ IMV~0--->CPAP, IMV~--~ CPAP, CPAP,---> CPAP~ CPAP, ~ CPAP~
Time to fatigue (rain) 6 -40 1 31 22 1 13 1 6 5 3
Maximal .fall in H / L (%) 9.6 11.4 12.8 24.0 12.1 70.0 35.7 24.5 42.5 37.5 31.1 65.5 29.2 47.7 63.4
After H/L fall (ram Hi)* I
Po2 66 54 78 54 45 60 51
82 80 65 53 48 52
I Pco~_ 46 40 41 51 38 56 59 Apnea Apnea 54.7 • 52.1 • 53.5 • 49.7 • 56.4 • 59.4 •
Recovery time of H/L (rain) -3
1.3 1.2 1.6 1.4 1.1 2.4
32 5 1 6 1 1 4 1 3 5
CPAP ~ Continuous positive airway pr~ssure in cm H._,O;IMV = intermittent mandatory ventilation in breaths/minute. *Pa%, Pac% in infants 1-7, LP%, PETe%(mean _+SD of breaths) in the others.
istic of fatigue can be detected in normal infants during R E M sleep and in infants with cardiopulmonary disease? These studies were done using a digital computer. Since the results were not immediately available, they were not clinically useful. W e have therefore developed a simple analog device to indicate immediately and continuously the spectral frequency changes in the diaphragm of infants during weaning from mechanical ventilation.
METHODS We monitored ten intubated premature newborn infants during 15 studies w h e n they were being switched from IMV to CPAP or w h e n CPAP was reduced. Their gestational age and age at the time of the study are shown in Table I. The study included infants with respiratory distress syndrome, patent ductus arteriosus and congestive heart failure, and pneumonia. Respiratory movements were monitored using anteroposterior magnetometer pairs placed in the midline, one at the level of the nipple, and the other 1 cm above the umbilicus.' In seven studies arterial Pac% was measured; in the others end-tidal Pco~ was monitored continuously at the endotracheal tube using a Beckman LB~ CO~ analyzer adapted for the low respiratory flow rates present in newborn babies. 8 The diaphragmatic E M G was recorded using Beckman surface electrodes in the right sixth and seventh interspaces or subcostally, between the mid-clavicular and midaxillary lines. 2 These electrodes were secured to t h e skin with double adhesive tape rings. The accepted
interelectrode impedance was below 20 kohm. The E M G signal was led to a hfgh input impedance, high c o m m o n mode rejection E M G preamplifier* placed in the incubator close to the infant. F r o m there it was taken to a Honeywell Accudata 134A biomedical amplifier* with band pass filters of 15-1000 Hz for further conditioning. To analyze the diaphragmatic frequency spectrum it is necessary to gate the superimposed E C G signal in a method similar to the one described by Prechtl et aP and ourselves? ~ Briefly, the E M G signal was introduced into two channels. In channel one it was led to a so-called bucket brigade circuit which delayed this signal for a period of up to 50 milliseconds, In channel 2, the E C G artifact in the E M G signal acted as a trigger for a pulse generator, the output ~ which reached the delayed E M G signal in channel one just prior to the cardiac artifact and blocked out any signal for the duration o f the pulse. The blocking pulse had a variable time setting so that it was possible to suppress the entire E C G signal. The gated E M G signal was then fed into two-band pass filters: a high-frequency one (150-350 Hz) and a low-frequency one (20-,46.7.Hz). These band pass filters have a similar frequency, range to those used by Kadefors et al 5 and by Gross et al. 7 The rectified output of each filter passed through a similar electronic switch into an averager circuit. W h e n the switch was closed the averager developed a voltage proportional to the average value of the band o f frequencies. W h e n the switch opened (during the *Honeywell Test Instruments Division, P.O. Box 5227, Denver, CO 80217.
Volume 95 Number 5, part 1
Diaphragmatic muscle fatigue
795
HIGH/LOW FREQUENCY ,(percent normal)
::::::
: ?
:::
,
::
;:
--:,
:
::::;;:;:
Figure. The tracing shows the H/L frequency ratio monitored from the diaphragmatic EMG, the PETeo2,rib cage and abdominal movement. The infant has been weaned from IMV to CPAP alone and the results show the progressive fall in the H/L ratio followed by rising PETe%.Recovery of the H/L ratio is rapid when IMV is restarted (arrow), while PExc% shows a slower return to normal.
QRS complex) the averager maintained its output at the level that existed just prior to the opening of the switch. The gated and averaged output of the high-frequency filter was then electronically divided by the similarly processed output of the low-frequency filter to provide a continous high-to-low frequency ratio, which was shown on a visual display. Since there are small changes in the frequency spectrum at the beginning of inspiration," the analysis of the H / L ratio was done ,at midinspiration as judged from the magnetometer tracing when the baby was breathing quietly without gross body movements. Control values for the H / L ratio were obtained from the mean of 10 breaths when the infants were stable just before any changes were made in the ventilatory mode. In order to assess the stability of this control value, 10 to 20 sections of 10 breath intervals were analyzed while the infant was stable. Following ventilatory mode changes, similar 10 breath intervals were analyzed. All mean values were expressed as a percentage of the initial control H / L ratio. The H / L ratio measurements did not influence clinical management. The usual practices of nursing care and r e p e a t e d blood gas measurements were continued. Because diaphragmatic muscle fatigue occurs rapidly in the presence of hypoxia, 12 oxygenation was monitored
using a transcutaneous Po~ electrode (Radiometer). Initially the infants were stable m inspired oxygen concentrations ranging from 25 to 40%. During the studies the inspired oxygen was altered in order to maintain a stable transcutaneous Po2 reading. None of the infants required increases of more than 20% inspired oxygen. All data were recorded on a Hewlett-Packard strip chart recorder or a Brush pen recorder. Statistical analysis was done using the unpaired t test. RESULTS During the 15 studies while the infants were stable and breathing on assisted ventilation, the mean H/L ratio did not vary more than 10% (the coefficient of variation ranged from 1.1 to 4.5%). When assisted ventilation was altered, we observed two different patterns. In four infants, two going from IMV to CPAP and two in whom CPAP was reduced, the H / L ratio did not fall more than 12.8% from the control value at any time throughout the studies, which continued for two to five hours (Table). These infants had normal blood gas values throughout the study and remained stable for at least 12 hours afterward. In 11 studies, after assisted ventilation had been altered, the H / L ratio fell progressively to levels of 20% or more
796
Muller et al.
below the control value (P < 0.001). The time observed for the fall to occur varied from infant to infant (Table). There was a significant rise in the arterial Pco~ in three infants when blood gas values were determined after the development of fatigue. Two babies developed severe apneic spells and were put on assisted ventilation before arterial blood gases were obtained (Table)~ In the six studies in which end-tidal CO2 was monitored, there was no significant change in PETCO~ when the H / L ratio remained within 10% of the control. When the H / L ratio fell 20% or more below the control value, there was a progressive and significant rise in PETCO2 (P < 0.001). The maximum levels are listed in the Table. One study is shown in the Figure. There was a steady fall in the H / L ratio with a slower rise in PETco~ after weaning from IMV. The time lapse between the H / L ratio fall of over 20% and the rise in PETe% varied in the infants from 0.5 to 10 minutes. The time for recovery from fatigue, after assisted ventilation had been reinstituted, varied in the infants studied (Table). The longest recovery time of 32 minutes occurred in an infant with the largest fall (70%) in the H / L ratio (Table). DISCUSSION The changes in the E M G spectrum with muscle fatigue have been known since the beginning of this century. 1~ These changes are best described by spectral frequency analysis, using either multiple band pass filters or fast Fourier transform analysis. The shift in the E M G power to lower frequencies with a fatiguing load is well established?' ~ The reasons for these changes are still conjectural. Mortimer et al TM showed that the changes were related to a decrease in the conduction velocity within the muscle fiber and suggested that this was due to the accumulation of metabolic end products within the muscle. However, when a muscle is presented with a fatiguing load there is an almost immediate increase in love-frequency power and a gradual decrease in the high-frequency power? Thus the changes predict eventual exhaustion, and the application of this technique to the diaphragm was a major step. 7 We established that the spectrum of the infant diaphragm was the same as for any other skeletal muscle, 2 and that the essential information could be obtained from a high and a low band pass ratio, without going to the complexity of spectral frequency analysis.~ There are some limitations to the use of diaphragmatic EMG monitoring of fatigue. The diagnosis of fatigue cannot be made by inspection of the frequency spectrum alone, but depends on observing a changae in the spectrum to one of a lower-power density. It is important, therefore, to obtain a control ratio when the infant is stable on the
The Journal of Pediatrics November 1979
ventilator, and it can be reasonably assumed that the diaphragm is not fatigued. A further problem is the relationship of hypoxia to fatigue. The endurance of the diaphragm is markedly reduced by hypoxia. 12Thus during weaning, when there is unexpected hypoxia, there is a rapid fall in the H / L ratio. This occurred during this study in one infant who became severely hypoxic during feeding. Therefore it is preferable to have continuous monitoring of the Po2 with a transcutaneous electrode. Within these limitations, our results suggest that monitoring of the diaphragmatic E M G may be of benefit in directing the weaning process. E M G changes precede the deterioration of blood gas values, give a continuous output, and are technically simpler than monitoring end tidal CO2. There are compelling reasons for discontinuing mechanical ventilation in infants as quickly as possible. Applying high pressures to a lung with a nonuniform compliance must overdistend the most compliant (presumably the healthiest) areas of lung. The immediate risk is pneumothorax or pneumomediastinum; the long-term risk is bronchopulmonary dysplasia. In addition, the diaphragm decreases activity during mechanical ventilation. An inactive muscle rapidly loses oxidative capacity and hence fatigues more readily.1~. 16 Thus the least ventilatory assistance that is provided, the better for both the lung and the diaphragm. Currently, however, the least ventilatory assistance required is judged by pushing the weaning process to the point of clinical deterioration. As diaphragmatic E M G evidence precedes clinical deterioration, it may permit faster weaning with fewer crises. REFERENCES
1. Knill RL, Andrews W, Bryan AC, and Bryan MH: Respiratory load compensation in infants, J Appl Physio140:357, 1976. 2. Muller N, Gulston (3, Cade D, Whitton J, Froese AB, Bryan MH, and Bryan ~ : Diaphragmatic muscle fatigue in the newborn, J Appl Physiol 46:688, 1979. 3. MolbechS: Average percentage force at repeated maximal isometric muscle contractions at different frequencies, in Communications from The Testing and Observations Institute of Danish National Association for Infantile Paralysis, No. 16, 1963. 4. Keens TG, Bryan AC, Levison H, and Ianuzzo CD: Development of fatigue-resistant muscle fibers in human ventllatory muscles, J Appl Physlol 44:909, 1978. 5. Kad~fors R, Kaiser E, and Petersen I: Dynamic spectrum analysis of myopotentials and special reference to muscle fatigue, Electromyography 8:39, 1968. 6. Kogi K, and Hakamada T: Slowing of surface electromyogram and muscle strength in muscle fatigue, Res Inst Sci Lab 60:27, 1962. 7. Gross D, Grassino A, Ross WD, and Macklem PT: Electro-
Volume 95 Number 5, part 1
8.
9.
10. 11.
Diaphragmatic muscle fatigue
myogram pattern of diaphragmatic fatigue, J'Appl Physiol 46:1, 1979. Rigatto H, and Brady JP: Periodic breathing and apnea in pre-term infants, I. Evidence for hypoventilation possibly due to central respiratory depression, Pediatrics 50:202, 1972. Prechtl HFR, Van Eykern LA, and O'Brien M J: Respiratory muscle EMG in newborns: a non-intrusive method, Early Hum Dev 1:265, 1977. Muller N, Volgyesi G, Becket L, Bryan MH, and Bryan AC: Diaphragmatic muscle tone, J Appl Physiol (in press). Schweitzer TW, Fitzgerald JW, Bowden JA, and LynneDavies P: Spectral analysis of human inspiratory diaphragmatic electromyograms, J Appl Physiol 46:152, 1979.
797
12. Roussos CS, and Macldem PT: Diaphragmatic fatigue in man, J Appl Physiol 43:189, 1977. 13. Cobb S, and Forbes A: Electromyographic studies of muscular fatigue in man, Am J Physiol 65:234, 1923. 14. Mortimer JT, Magnusson R, and Petersen I: Conduction velocity in ischemic muscle: effect on EMG frequency spectrum, Am J Physiol 219:1324, 1970. 15. Booth FW: Time cot~rse of muscular atrophy during immobilization of hindlimbs in rats, J Appl Physiol 43:656, 1977. 16. Rifenberick DH, Gamble JG, and Max SR: Response of mitochondrial enzymes to decreased muscular activity, Am J Physiol 225:1295, 1973.
Copyright information The appearance of a code at the bottom of the first page of an original article in this JOURNAL indicates the copyright owner's consent that copies of the article may be made for personal or internal use, or for the personal or internal use of specific clients. This consent is given on the condition, however, that the copier pay the stated per copy fee through the Copyright Clearance Center, Inc., P.O. Box 765, Schenectady, N.Y. 12301, (518) 374-4430, for copying beyond that permitted by Sections 107 or 108 of the U.S. Copyright Law. This consent does not extend to other kinds of copying, such as copying for general distribution, for advertising or promotional purposes, for creating new collective works, or for resale.
