American Journal of Respiratory and Critical Care Medicine

We determined whether prolonged complete inactivation of the human diaphragm results in atrophy and whether this could be prevented by brief periods of electrical phrenic nerve stimulation. We studied a subject with high spinal cord injury who required removal of his left phrenic nerve pacemaker (PNP) and the reinstitution of positive-pressure ventilation for 8 mo. During this time, the right phrenic nerve was stimulated 30 min per day. Thickness of each diaphragm (tdi) was determined by ultrasonography. Maximal tidal volume (Vt) was measured during stimulation of each diaphragm separately. After left PNP reimplantation, Vt and tdi were measured just before the resumption of electrical stimulation and serially for 33 wk. On the previously nonfunctioning side, there were substantial changes in Vt (from 220 to 600 ml) and tdi (from 0.18 to 0.34 cm). On the side that had been stimulated, neither Vt nor tdi changed appreciably (Vt from 770 to 900 ml; tdi from 0.25 to 0.28 cm). We conclude that prolonged inactivation of the diaphragm causes atrophy which may be prevented by brief periods of daily phrenic nerve stimulation.

Immobilization of peripheral skeletal muscle results in a loss of muscle mass with atrophy typically occurring over a 4 to 6-wk period (1). In rats, controlled positive-pressure mechanical ventilation for as little as 48 h is accompanied by diaphragm atrophy as well as by a reduction in diaphragm contractile properties (2). In humans, however, the occurrence of disuse atrophy of the diaphragm as a consequence of mechanical ventilation has not been demonstrated. Furthermore, the amount of stimulation of the diaphragm that would be needed to prevent atrophy is unclear.

To evaluate whether brief periods of electrical stimulation prevent diaphragm atrophy in humans, we assessed diaphragm mass and function in a patient with a high spinal cord injury whose ventilation was usually sustained with bilateral phrenic nerve pacemakers. Unilateral infection of one pacemaker receiver site in the abdominal wall necessitated pacemaker removal and the institution of passive mechanical ventilation for 8 mo. This study was designed to determine if atrophy of the diaphragm connected to the remaining pacemaker could be prevented by brief periods of electrical stimulation and if atrophy of the unstimulated side occurred during the 8 mo of inactivity.

Subject Selection

A 49-yr-old high tetraplegic man with unilateral dysfunction of a phrenic nerve pacemaker was identified for our study. Ten years prior to study, he suffered a spinal cord injury that resulted in complete motor paralysis below C2. One year after injury, bilateral phrenic nerve pacemakers (Model S232G; Avery Laboratories Inc., Commack, NY) were implanted, and positive-pressure mechanical ventilation was discontinued. Seven years after implantation, infection occurred at the site of the left pacemaker receiver, although it continued to function normally. That is, infection in the abdominal pouch in which the receiver resided did not affect the function of the pacemaker or the nerve to which it was attached. The infection eventually led to removal of the receiver and the reinstitution of positive-pressure mechanical ventilation. The subject received positive-pressure ventilation for 8 mo during which the functional right side was paced 30 min per day to determine if this would prevent atrophy. The minimum electrical stimulus that generated a maximal tidal volume was used to pace the functional side (amplitude = 2.8 mA, rate = 10, pulse width = 150 microseconds, frequency = 20 Hz, and inspiratory time = 1.3 s).

Five days after pacemaker reimplantation on the left side, bilateral phrenic nerve pacing was restarted (30 min/d). Over the course of 6 wk, pacing time was gradually increased to 24 h/d.


Determination of tidal volume. A Wright spirometer (Mark 8; Ferraris Medical Limited, London, UK) calibrated with a graduated syringe was attached to the subject's tracheostomy tube to measure tidal volume. The subject was suctioned prior to testing and the cuff of the tracheostomy tube was inflated. On two occasions, the subject had an uncuffed tracheostomy tube in place. During measurements on those days, the tracheostomy tube flange was sealed to the skin with petrolatum and the subject wore a noseclip and voluntarily closed his mouth. Each phrenic nerve was stimulated separately. Signal amplitude was increased from zero until a tidal volume could be measured (considered to be the threshold amplitude). Thereafter, amplitude was increased in increments of 0.2 to 0.5 mA. At each new amplitude, three tidal volumes were measured and an average obtained. Amplitude of stimulation was increased until no further increases in tidal volume occurred. This volume was considered to represent the maximal tidal volume.

Diaphragm ultrasound. The technique of diaphragm ultrasound for measuring diaphragm thickness has been described previously (3). Briefly, the diaphragm was visualized at the zone of apposition of the diaphragm to the rib cage using a 7.5-10 MHZ annular transducer (Ultramark 9: HDI; Advanced Technology Laboratories, Bothell, WA). The diaphragm was identified by the presence of two echogenic lines representing the pleura and peritoneum overlying the muscle (which is less echogenic). Thickness of the diaphragm (tdi) was measured as the distance (in centimeters) between the outer edges of the echogenic lines. Measurements were obtained on both the right (probe placed at the 8th intercostal space, midaxillary line) and left (8th intercostal space anterior axillary line) sides with the subject studied supine. These sites were chosen because they allowed the best visualization of the diaphragm. Subsequent diaphragm thickness measurements were made identically. All measurements were done at end-expiration.

Protocol. Baseline tdi was measured 5 d after implantation of the left phrenic nerve pacemaker (i.e., before retraining had begun), and followed over time (4, 9, 23, and 33 wk after surgery). Maximal tidal volumes were determined in the supine position for the right and left sides on the same days that the diaphragm ultrasounds were performed. Tidal volumes at 5 d and at 4 wk and 9 wk after surgery were determined with a cuffed tracheostomy tube, whereas tidal volumes at 23 and 33 wk were determined with an uncuffed tracheostomy tube as described above.

Data analysis. Relationships between time and tidal volumes, and time and diaphragm thickness were analyzed using linear regression (4). A p value of less than 0.05% was considered statistically significant.

The relationship between tidal volume and magnitude of electrical stimulation of the diaphragm 5 d after pacemaker reimplantation and prior to retraining of the diaphragm is shown in Figure 1. As expected, as the magnitude of stimulation was increased, tidal volume increased and then reached a maximum. Thereafter, further increases in stimulus magnitude did not result in increases in tidal volume. The stimulus-response curve is typical for phrenic nerve pacemakers. The maximal tidal volume of the previously functional side was greater than that produced by the newly reimplanted pacemaker (220 versus 770 ml, Figure 2).

Prior to retraining, the diaphragm thickness was less on the unstimulated side than on the side that was stimulated 30 min per day (0.18 versus 0.25 cm). Resumption of electrical stimulation of the previously unstimulated side resulted in significant increases in both maximal tidal volumes (220 to 600 ml, p < 0.05) (Figure 2), and diaphragm thickness over time (0.18 versus 0.34 cm, p < 0.05) (Figure 3). By contrast, despite the reinstitution of continuous phrenic nerve pacing, the hemidiaphragm that had been stimulated intermittently for 8 mo did not increase significantly in maximal tidal volume (from 770 to 900 ml, p > 0.10) (Figure 2) or thickness (from 0.25 to 0.28 cm, p > 0.20) (Figure 3).

The thickness and maximal tidal volume of the hemidiaphragm unstimulated for an 8-mo period were less than those of the hemidiaphragm stimulated 30 min/d. Thickness of the unstimulated hemidiaphragm (0.18 cm) was less than what is observed typically in normal persons, and the thickness of the hemidiaphragm stimulated for only 30 min/d (0.25 cm) was well within the normal range (5). Reinstitution of continuous phrenic nerve pacing produced a training effect only on the side unstimulated for 8 mo. The side stimulated for only 30 min/d for 8 mo was unaffected by the resumption of continuous stimulation. On the basis of these results, we conclude that diaphragm atrophy occurred only on the unstimulated side after prolonged complete disuse. These findings are consistent with data in chronic unilateral denervation of the diaphragm in which the paralyzed side was atrophied compared with the normally functioning side (6). Whether passive stretch of the paralyzed side of the subjects in that study (6) or of the unstimulated side in our subject affected the extent of atrophy is unclear, and we do not know of information in the literature that bears directly on this question.

Our results have relevance to patients undergoing mechanical ventilation in whom it has been postulated that unloading of the respiratory muscles and reduction in diaphragm EMG activity may lead to atrophy and consequent difficulties in weaning. It remains unclear precisely how much diaphragm activation is required to prevent this. Qin and colleagues (7) demonstrated that brief periods of tibialis muscle stimulation (30 min/d, 5 d/wk) were sufficient to prevent atrophy during 3 wk of hindlimb immobilization. Our data, and those of Qin and colleagues (7) suggest that substantial functional integrity of skeletal muscle can be preserved with relatively little activation. Pertinent in this regard is the recognition that in patients ventilated with the assist-control mode of ventilation, work performed by respiratory muscles often equals or exceeds the work expected in spontaneously breathing normal subjects (8). In the light of our data and those of Qin and colleagues (7), it appears plausible that this work would be sufficient to preserve diaphragm function and possible that concerns about respiratory muscle atrophy caused by disuse are overstated. If atrophy occurs, critical illness polyneuropathy, nutritional deficiency, and a catabolic state may be factors more important than disuse.

Supported by a DVA Merit Review Grant.

1. Booth F. W.Time course of muscular atrophy during immobilization of hindlimbs in rats. J. Appl. Physiol.431977656661
2. Le Bourdelles G., Viires N., Broczowski J., Seta N., Pavlovic D., Aubier M.Effects of mechanical ventilation on diaphragmatic contractile properties in rats. Am. J. Respir. Crit. Care Med.1531996787793
3. McCool F. D., Conomos P., Benditt J. O., Cohn D., Sherman C. B., Hoppin F. G.Maximal inspiratory pressures and dimensions of the diaphragm. Am. J. Respir. Crit. Care Med.155199713291334
4. Rosner, B. 1995. Fundamentals of Biostatistics, 4th ed. Duxbury Press, Belmont, MA. 443–469.
5. McCool F. D., Benditt J. O., Conomos P., Anderson L., Sherman C. B., Hoppin F. G.Variability of diaphragm structure among healthy individuals. Am. J. Respir. Crit. Care Med.155199713231328
6. Gottesman E., McCool F. D.Ultrasound evaluation of the paralyzed diaphragm. Am. J. Respir. Crit. Care Med.155199715701574
7. Qin L., Appell H. J., Chan K. M., Maffuli N.Electrical stimulation revents immobilization atrophy in skeletal muscle of rabbits. Arch. Phys. Med. Rehabil.781997512517
8. Marini J. J., Rodriguez R. M., Lamb V.The inspiratory work of patient-initiated mechanical ventilation. Am. J. Respir. Crit. Care Med.1341986902909
Correspondence and requests for reprints should be addressed to F. Dennis McCool, M.D., Pulmonary Division, Memorial Hospital, 111 Brewster St., Pawtucket, RI 02860.


No related items
American Journal of Respiratory and Critical Care Medicine

Click to see any corrections or updates and to confirm this is the authentic version of record