Effect of lung volume reduction surgery for emphysema on diaphragm function

Preoperative prediction of a successful outcome following lung volume reduction surgery (LVRS) for emphysema is imperfect. One mechanism could be improvement in respiratory muscle function yet controversy exists regarding the magnitude and mechanism of such an improvement. Therefore, we measured diaphragm strength in 18 patients before and after LVRS. Mean (S.D.) FRC fell from 6.53 to 5.40 l (p = 0.0001). Mean sniff transdiaphragmatic pressure increased from 76 to 87 cm H2O (14%, p < 0.03) and mean twitch transdiaphragmatic pressure (Tw Pdi) increased by 2.5 cm H2O at 3 months (12%, p = 0.03). There was a highly significant increase in twitch esophageal pressure (Tw Pes) (60%, p < 0.0001), which was maintained at 12 months (46% increase, p = 0.0004). No change was observed in quadriceps twitch tension in nine subjects in whom it was measured. After LVRS the ratio Tw Pes:Tw Pdi increased from 0.24 to 0.37 at 3 months (p = 0.0003) and 0.36 at 12 months (p = 008). Low values of Sn Pdi, Sn Pes, Tw Pes and a high RV/TLC ratio were the preoperative variables most predictive of improvement in shuttle walking distance. We conclude that LVRS improves diaphragm function primarily by alteration of lung volume. Patients with poor diaphragm function and high RV/TLC ratio preoperatively are most likely to benefit from the procedure.     

Contractile properties of the human diaphragm during chronic hyperinflation

BACKGROUND. In patients with chronic obstructive pulmonary disease (COPD) and hyperinflation of the lungs, dysfunction of the diaphragm may contribute to respiratory decompensation. We evaluated the contractile function of the diaphragm in well-nourished patients with stable COPD, using supramaximal, bilateral phrenic-nerve stimulation, which provides information about the strength and inspiratory action of the diaphragm. METHODS. In eight patients with COPD and five control subjects of similar age, the transdiaphragmatic pressure generated by the twitch response to phrenic-nerve stimulation was recorded at various base-line lung volumes, from functional residual capacity to total lung capacity, and during relaxation and graded voluntary efforts at functional residual capacity (twitch occlusion). RESULTS. At functional residual capacity, the twitch transdiaphragmatic pressure ranged from 10.9 to 26.6 cm of water (1.07 to 2.60 kPa) in the patients and from 19.8 to 37.1 cm of water (1.94 to 3.64 kPa) in the controls, indicating considerable overlap between the two groups. The ratio of esophageal pressure to twitch transdiaphragmatic pressure, an index of the inspiratory action of the diaphragm, was -0.50 +/- 0.05 in the patients, as compared with -0.43 +/- 0.02 in the controls (indicating more efficient inspiratory action in the patients than in the controls). At comparable volumes, the twitch transdiaphragmatic pressure and esophageal-to-transdiaphragmatic pressure ratio were higher in the patients than in normal subjects, indicating that the strength and inspiratory action of the diaphragm in the patients were actually better than in the controls. Twitch occlusion (a measure of the maximal activation of the diaphragm) indicated near-maximal activation in the patients with COPD, and the maximal transdiaphragmatic pressure was 106.9 +/- 13.8 cm of water (10.48 +/- 1.35 kPa). CONCLUSIONS. The functioning of the diaphragms of the patients with stable COPD is as good as in normal subjects at the same lung volume. Compensatory phenomena appear to counterbalance the deleterious effects of hyperinflation on the contractility and inspiratory action of the diaphragm in patients with COPD. Our findings cast doubt on the existence of chronic fatigue of the diaphragm in such patients and therefore on the need for therapeutic interventions aimed at improving diaphragm function.     

Effects of diaphragm activation on airway pressure generation during lower thoracic spinal cord stimulation

Lower thoracic spinal cord stimulation (SCS) results in the generation of large positive airway pressures. The potential effects of diaphragm co-activation during SCS were investigated in 10 anesthetized dogs. Diaphragm compound action potentials (CMAPs) were present during SCS at the T10 and T12 levels. In group 1, airway (Paw) and trans-diaphragmatic (Pdi) pressures were monitored during supramaximal SCS before and after phrenicotomy. In group 2, pressures were monitored before and after C2 section to evaluate the potential influence of supraspinal centers. Following phrenicotomy in group 1, the reduction in Pdi during SCS was associated with increases in Paw. In group 2, diaphragm CMAPs and active Pdi increased following C2 section, while Paw fell. Following phrenicotomy, Paw increased significantly. In intact animals therefore, changes in Paw during SCS are affected by the interaction between inhibitory and excitatory influences on diaphragm activation. We conclude that lower thoracic SCS results in substantial diaphragm co-activation and secondary reductions in airway pressure generation.     

The effect of level of contraction on the electromyographic power spectrum of the diaphragm in pigs.

We investigated the relationship between the frequency components of myoelectric power spectra of the diaphragm and the level of diaphragmatic contraction in seven anaesthetized spontaneously breathing pigs. Electromyographic activity of the costal and crural portions of the diaphragm were recorded with fish-hook electrodes and the frequency-power spectra during inspiration were computed and expressed in terms of centroid frequency (fc). Diaphragmatic force was indirectly assessed as transdiaphragmatic pressure (Pdi) which was measured with balloon-catheter systems placed in the abdomen and oesophagus. The relationships between Pdi and costal and crural fc were assessed during brief (2 min) and incremental increases in diaphragmatic contraction, achieved by gradual occlusion of the inspiratory line of the breathing circuit. When Pdi was increased to 128, 191, 287 and 421% of the value measured during unobstructed breathing, costal and crural fc rose significantly in all animals because of an increase in the power of high-frequency components and a decline in the power of low-frequency components. Both costal and crural fc returned to control values within 5 min of the release of inspiratory occlusion. Our results indicate that the level of contraction is an important determinant of the diaphragmatic myoelectric power spectrum and should be taken into consideration when using power spectral analysis to diagnose diaphragmatic mechanical failure.     

Reproducibility of twitch and sniff transdiaphragmatic pressures

Twitch transdiaphragmatic pressure (Tw Pdi) measured with magnetic stimulation of the phrenic nerve is used to follow up patients and to assess the effect of clinical treatments on diaphragm function. However the reproducibility of Tw Pdi on different occasions has been little studied. We investigated 32 normal subjects, measuring Tw Pdi elicited by bilateral magnetic stimulation of the phrenic nerves on two to 14 occasions. Sniff transdiaphragmatic pressure (sniff Pdi) was also measured. The mean value of Tw Pdi and sniff Pdi were 28+/-5 and 134+/-24 cm H(2)O, respectively. The within subjects coefficient of variation was 11% for both Tw Pdi and sniff Pdi. We conclude that there is a variability of Tw Pdi and the variability of Tw Pdi is the same as that of sniff Pdi.     

The influence of lung volume on pharyngeal mechanics, collapsibility, and genioglossus muscle activation during sleep

STUDY OBJECTIVES: Previous studies in both awake and sleeping humans have demonstrated that lung-volume changes substantially affect upper-airway size and pharyngeal resistance and, thus, may influence pharyngeal patency. We sought to systematically investigate the isolated effects of lung-volume changes on pharyngeal collapsibility and mechanics and genioglossus muscle activation during stable non-rapid eye movement sleep. We hypothesized that lower lung volumes would lead to increased pharyngeal collapsibility, airflow resistance, and, in compensation, augmented genioglossus muscle activation. DESIGN: Nineteen normal individuals (age, 30.4 +/- 0.5 years; body mass index: 24.5 +/- 0.4 kg/m2) were studied during stable non-rapid eye movement sleep in a rigid head-out shell equipped with a variable positive/negative pressure attachment for manipulations of extrathoracic pressure and, thus, lung volume. SETTING: Sleep physiology laboratory. PARTICIPANTS: Normal healthy volunteers. INTERVENTIONS: N/A. MEASUREMENTS AND RESULTS: We measured change in end-expiratory lung volume (EELV) (magnetometers), genioglossus electromyogram (GGEMG) (intramuscular electrodes), pharyngeal pressure, and collapsibility of the pharynx in response to a brief pulse of negative pressure (-8 to -15 cm H2O) under the following conditions: (1) baseline, (2) increased EELV (+1 liter), and (3) decreased EELV (-0.6 liter). Reduced lung volumes led to increased inspiratory airflow resistance (7.54 +/- 2.80 cm H2O x L(-1) x s(-1) vs 4.53 +/- 1.05 cm H2O x L(-1) x s(-1), mean +/- SEM, P = 0.02) and increased genioglossus muscle activation (GGEMG peak 14.6% +/- 1.5% of maximum vs 8.6% +/- 1.5% of maximum, maximum P = 0.001) compared to baseline. The pharynx was also more collapsible at low lung volumes (4.3 +/- 0.5 cm H2O vs 5.4 +/- 0.6 cm H2O, P = 0.04). CONCLUSIONS: We conclude that upper-airway muscles respond to changes in lung volumes but not adequately to prevent increased collapsibility. These results suggest that lung volume has an important influence on pharyngeal patency during non-rapid eye movement sleep in normal individuals.     

Cartography of human diaphragmatic innervation: preliminary data

In humans, anatomy indicates that the phrenic nerve mainly arises from the C4 cervical root, with variable C3 and C5 contributions. How this translates into functional innervation is unknown. The diaphragm response to electrical stimulation of C3, C4 and C5 was described in three patients undergoing surgical laryngeal reinnervation with an upper phrenic root (surface chest electrodes at anterior, lateral and posterior sites; oesophageal and gastric pressures (Pes and Pga) to derive transdiaphragmatic pressure (Pdi)). Anatomically, the phrenic nerve predominantly originated from C4. Phrenic stimulation elicited motor responses at the three sites in the three patients, as did C4 stimulation. It produced Pdi values of 9, 11, and 14cmH(2)O in the three patients, respectively, vs. 9, 9, and 7cmH(2)O for C4. C3 stimulation produced modest Pdi responses, whereas C5 stimulation could produce Pdi responses close to those observed with C4 stimulation. These singular observations confirm the dominance of C4 in diaphragm innervation but suggest than C5 can be of importance.     

Novel method for transdiaphragmatic pressure measurements in mice

The diaphragm muscle (DIAm) is responsible for breathing and determines the ability to generate both ventilatory and non-ventilatory behaviors. Size limitations of the mouse make transdiaphragmatic pressure (Pdi) measurement using a dual balloon system untenable. Adult C57BL/6J mice (n = 8) and C57BL/6 × 129 (n = 9), underwent Pdi measurements using solid-state pressure catheters spanning the thoracic and abdominal surfaces of the DIAm. Measurements were conducted during eupnea, hypoxia (10% O₂)–hypercapnia (5% CO₂), chemical airway stimulation (i.e., sneezing), spontaneously occurring deep breaths, sustained tracheal occlusion, and bilateral phrenic nerve stimulation. There was a difference in the Pdi generated across the range of ventilatory and non-ventilatory behaviors (p = 0.001). No difference in Pdi across behaviors was evident between mouse strains (p = 0.161). This study establishes a novel method to determine Pdi across a range of DIAm behaviors in mice that may be useful in evaluating conditions associated with reduced ability to perform expulsive, non-ventilatory behaviors.     

Effect of expiratory resistive loading on inspiratory and expiratory muscle fatigue

Inspiratory and expiratory pressures are increased during expiratory resistive loading (ERL). We asked whether ERL elicits inspiratory as well as expiratory muscle fatigue. On four separate days, seven male subjects underwent ERL to task failure. Subjects maintained respiratory frequency at 15 breaths min⁻¹, expiratory gastric pressure at 40% or 60% of maximum, and expiratory duty cycle at 0.4 or 0.7 (ERL_40%0.4, ERL_40%0.7, ERL_60%0.4, ERL_60%0.7). Inspiratory and abdominal muscle contractility was assessed before and up to 30 min after ERL by measuring transdiaphragmatic twitch pressure (P_di,tw) and gastric twitch pressure (P_ga,tw) in response to magnetic nerve stimulation. After each trial, P_di,tw and P_ga,tw were reduced below baseline values (−9 to −15% for P_di,tw and −15 to −22% for P_ga,tw; P < 0.05). The severity of diaphragm fatigue was unaffected by expiratory pressure production or expiratory duty cycle, whereas extending the expiratory duty cycle increased the severity of abdominal muscle fatigue. In conclusion, ERL elicits contractile fatigue of the diaphragm and the abdominal muscles.     

Measuring end-expiratory lung volume and pulmonary mechanics to detect early lung function impairment in rabbits

We investigated whether end-expiratory lung volume (EELV) or lung mechanical parameters are more sensitive for the detection of a compromised gas exchange during bronchoconstriction and after surfactant depletion. EELV was determined via SF(6) multiple breath wash-outs in mechanically ventilated rabbits while a positive end-expiratory pressure (PEEP) of 1, 3 or 7 cm H(2)O was maintained. Airway resistance (R(aw)) and parenchymal elastance (H) were estimated from the pulmonary input impedance measured at each PEEP level by means of forced oscillations. Measurements were repeated during i.v. methacholine (MCh) infusions and following lung injury induced by saline lavage. MCh induced marked elevations in R(aw), with no significant change in EELV or H at any PEEP. After lavage, the severity of hypoxia was reflected systematically in significant decreases in EELV at all PEEP levels (-42+/-13%, -26+/-4%, and -18+/-5% at 1, 3 and 7 cm H(2)O, respectively), whereas compromised gas exchange was not associated with consistent changes in the mechanical parameters at a PEEP of 7 cm H(2)O (20+/-9% and 14+/-9% in R(aw) and H, respectively; p=0.2). We conclude that R(aw) is the only sensitive indicator for the detection of a compromised lung function during MCh infusions, whereas the estimation of EELV is necessary to follow the progression of a lung injury when a high PEEP level is applied.     

Influence of diaphragm and rib cage muscle fatigue on breathing during endurance exercise

Inspiratory muscle fatigue (IMF) can develop during exhaustive exercise and cause tachypnea or rapid shallow breathing. We assessed the effects of rib cage muscle (RCM-F) and diaphragm fatigue (DIA-F) on breathing pattern and respiratory mechanics during high-intensity endurance exercise. Twelve healthy subjects performed a constant-load (85% maximal power) cycling test to exhaustion with prior IMF and a cycling test of similar intensity and duration without prior IMF (control). IMF was induced by resistive breathing and assessed by oesophageal and gastric twitch pressure measurements during cervical magnetic stimulation. Both RCM-F and DIA-F increased RCM and abdominal muscle force production during exercise compared to control. With RCM-F, tidal volume decreased while it increased with DIA-F. RCM-F was associated with a smaller increase in end-expiratory oesophageal pressure (i.e. decrease in lung volume) than DIA-F. These results suggest that RCM-F and not DIA-F is associated with rapid shallow breathing and that lowering the operating lung volume with DIA-F may help to preserve diaphragmatic function.     

Diaphragm motor unit recruitment in rats

We hypothesized that considerable force reserve exists for the diaphragm muscle (DIAm) to generate transdiaphragmatic pressures (Pdi) necessary to sustain ventilation. In rats, we measured Pdi and DIAm EMG activity during different ventilatory (eupnea and hypoxia (10% O₂)–hypercapnia (5% CO₂)) and non-ventilatory (airway occlusion and sneezing induced by intranasal capsaicin) behaviors. Compared to maximum Pdi (Pdi_max generated by bilateral phrenic nerve stimulation), the Pdi generated during eupnea (21 ± 2%) and hypoxia–hypercapnia (28 ± 4%) were significantly less (p < 0.0001) than that generated during airway occlusion (63 ± 4%) and sneezing (94 ± 5%). The Pdi generated during spontaneous sighs was 62 ± 5% of Pdi_max. Relative DIAm EMG activity (root mean square [RMS] amplitude) paralleled the changes in Pdi during different ventilatory and non-ventilatory behaviors (r² = 0.78; p < 0.0001). These results support our hypothesis of a considerable force reserve for the DIAm to accomplish ventilatory behaviors. A model for DIAm motor unit recruitment predicted that ventilatory behaviors would require activation of only fatigue resistant units.     

Assessing effects of PEEP and global expiratory lung volume on regional electrical impedance tomography

Objective: This study was performed to assess the value of electrical impedance tomography (EIT) as an indicator of tidal (V_T) and end expiratory lung volume (EELV).

Methods: EIT measurements were performed in seven healthy piglets during constant tidal volume ventilation at incremental and decremental positive end-expiratory pressure (PEEP) levels. Tidal impedance changes were calibrated to volume using V_T calculated from flow at the airway opening. Simultaneously, calibrated respiratory inductive plethysmography was used to measure EELV changes, and used as a reference standard.

Results: EIT systematically underestimated both V_T and EELV changes when EELV deviated from the level at which it was calibrated. Calculated over the entire pressure–volume curve, EIT systematically underestimated V_T by 28 ml, with a precision from ?16 to 72 ml. EELV was systemically underestimated by 406 ml, with a precision of ?38 to 849 ml. Nonlinear recruitment in the ventral regions of the lungs was the main cause of this underestimation.

Conclusions: Tidal and end-expiratory changes in pulmonary impedance reflect corresponding changes in lung volume, but the increasing underestimation with increasing lung volume should be taken into account in the analysis of EIT data.     

Maximal activation of the human diaphragm but not inspiratory intercostal muscles during static inspiratory efforts

It is widely held that transdiaphragmatic pressure is a reliable index of the extent of central activation of the diaphragm but the maximal voluntary transdiaphragmatic pressure is lower during inspiratory than expulsive efforts. To determine whether the diaphragm is fully activated during the two manoeuvres supramaximal stimuli were delivered to both phrenic nerves during maximal efforts. No discernible twitch was evoked during 30-55% of attempted maximal efforts with either voluntary manoeuvre. Thus the difference in maximal transdiaphragmatic pressure between the manoeuvres must reflect changes in chest-wall geometry or mechanics rather than in the phrenic motor outflow. Inspiratory intercostal muscle activity was consistently submaximal during maximal inspiratory efforts.     

An implantable impedance pneumograph monitor for detection of diaphragm contraction and airway obstruction during diaphragm pacing

Impedance pneumography signals were characterised during diaphragm pacing using stimulating and recording electrodes placed on the abdominal surface of the diaphragm. These measurements were useful for the detection of muscle contraction without confounding effects from stimulus artifacts. Impedance pneumography signals were measured using 23 epimysial electrodes implanted in seven dogs with 1-5 experiments on each electrode. The polarity of the change in impedance associated with diaphragm pacing differed for each recording electrode and its configuration. Thirty-four of 57 cases produced increased impedance, 11 produced decreased impedance and the remaining 12 depended on the level of diaphragm activation. Impedance pneumography signals were useful for detecting complete airway obstruction. The mean difference between the impedance measured during open and obstructed airway conditions was 80% of the open airway impedance signal. The difference between open and obstructed airway impedance measurements was a mean of 2.3 times larger with a recording electrode on the same hemidiaphragm as the stimulating electrode, compared to an electrode placed on the opposite hemidiaphragm (p < 0.05, paired t test, four dogs). In addition, the differences between open and completely obstructed airways were a mean of 2.8 times larger when the second recording electrode was placed on the thorax at the fifth intercostal space, compared to the ninth intercostal space (p < 0.05, two-factor ANOVA, one dog, two replicates). It was concluded that impedance pneumograph circuitry could be incorporated into an existing diaphragm pacer using electrodes placed on the diaphragm to provide valuable measurements of the function of the device.     

Diaphragmatic effects of selective resection of the upper phrenic nerve root in dogs

The aim of this study was to evaluate the effects on the diaphragm of upper phrenic nerve root resections in dogs. During laryngeal reinnervation, selective resections of the upper phrenic nerve root (C5) were performed unilaterally (right side, n=7; Group A) and bilaterally (n=6; Group B) and compared to non denervated animals (n=5). After 8 months, a diaphragmatic evaluation was performed: X-ray, EMG, transdiaphragmatic pressure (Pdi) after ipsi- and bilateral tetanic stimulation of the phrenic nerves and a bilateral histological study of five hemidiaphragmatic regions. EMG alterations were significantly more severe in Group B than in Group A, for the left (p<0.05) and right hemidiaphragms (p<0.01). No differences in the X-rays were noted between the three groups. The Pdi of the three groups after occlusion and phrenic nerve stimulations (unilateral and bilateral) were not statistically different. Histological data demonstrated that there were no differences in fibre irregularity, predominant fibre type or fibrosis between the three groups. Macroscopic and microscopic atrophy, which was mainly present on the anterior regions of the hemidiaphragms, was significantly higher in Group B than in Group A and undenervated dogs (p<0.05). In conclusion, resection of the upper phrenic nerve root of one phrenic nerve (right side) have limited effect on the diaphragm in dogs. However, resection of the upper phrenic nerve root on both sides resulted in a significant effect on the EMGs and histology of the entire diaphragm without any significant consequences on transdiaphragmatic pressure.     

An implantable impedance pneumograph monitor for detection of diaphragm contraction and airway obstruction during diaphragm pacing

Impedance pneumography signals were characterised during diaphragm pacing using stimulating and recording electrodes placed on the abdominal surface of the diaphragm. These measurements were useful for the detection of muscle contraction without confounding effects from stimulus artifacts. Impedance pneumography signals were measured using 23 epimysial electrodes implanted in seven dogs with 1–5 experiments on each electrode. The polarity of the change in impedance associated with diaphragm pacing differed for each recording electrode and its configuration. Thirty-four of 57 cases produced increased impedance, 11 produced decreased impedance and the remaining 12 depended on the level of diaphragm activation. Impedance pneumography signals were useful for detecting complete airway obstruction. The mean difference between the impedance measured during open and obstructed airway conditions was 80% of the open airway impedance signal. The difference between open and obstructed airway impedance measurements was a mean of 2.3 times larger with a recording electrode on the same hemidiaphragm as the stimulating electrode, compared to an electrode placed on the opposite hemidiaphragm (p<0.05, paired t test, four dogs). In addition, the differences between open and completely obstructed airways were a mean of 2.8 times larger when the second recording electrode was placed on the thorax at the fifth intercostal space, compared to the ninth intercostal space (p<0.05, two-factor ANOVA, one dog, two replicates). It was concluded that impedance pneumograph circuitry could be incorporated into an existing diaphragm pacer using electrodes placed on the diaphragm to provide valuable measurements of the function of the device.     

Effect of expiratory positive airway pressure on sleep disordered breathing

STUDY OBJECTIVES: We sought to determine the effect of expiratory positive airway pressure on end expiratory lung volume (EELV) and sleep disordered breathing in obstructive sleep apnea patients. DESIGN: Observational physiology study PARTICIPANTS: We studied 10 OSA patients during sleep wearing a facial mask. We recorded 1 hour of NREM sleep without treatment (baseline) and 1 hour with 10 cm H2O EPAP in random order, while measuring EELV and breathing pattern. RESULTS: The mean EELV change between baseline and EPAP was only 13.3 mL (range 2-25 mL). Expiratory time was significantly increased with EPAP compared to baseline 2.64 +/- 0.54 vs 2.16 +/- 0.64 sec (P = 0.002). Total respiratory time was longer with EPAP than at baseline 4.44 +/- 1.47 sec vs 3.73 +/- 0.88 sec (P = 0.3), and minute ventilation was lower with EPAP vs baseline 7.9 +/- 4.17 L/min vs 9.05 +/- 2.85 L/min (P = 0.3). For baseline (no treatment) and EPAP respectively, the mean apnea+hypopnea index (AHI) was 62.6 +/- 28.7 and 56.8 +/- 30.3 events per hour (P = 0.4). CONCLUSION: In OSA patients during sleep, the application of 10 cm H2O EPAP led to prolongation of expiratory time with only marginal increases in FRC. These findings suggest important mechanisms exist to avoid hyperinflation during sleep.     

Structure-activity relationships in rodent diaphragm muscle fibers vs. neuromuscular junctions

The diaphragm muscle (DIAm) is a highly active muscle of mixed fiber type composition. We hypothesized that consistent with greater activation history and proportion of fatigue-resistant fibers, neuromuscular transmission failure is lower in the mouse compared to the rat DIAm, and that neuromuscular junction (NMJ) morphology will match their different functional demands. Minute ventilation and duty cycle were higher in the mouse than in the rat. The proportion of fatigue-resistant fibers was similar in the rat and mouse; however the contribution of fatigue-resistant fibers to total DIAm mass was higher in the mouse. Neuromuscular transmission failure was less in mice than in rats. Motor end-plate area differed across fibers in rat but not in mouse DIAm, where NMJs displayed greater complexity overall. Thus, differences across species in activation history and susceptibility to neuromuscular transmission failure are reflected in the relative contribution of fatigue resistant muscle fibers to total DIAm mass, but not in type-dependent morphological differences at the NMJ.     

Effect of hypophosphatemia on diaphragmatic contractility in patients with acute respiratory failure

We studied the effects of hypophosphatemia on diaphragmatic function in eight patients with acute respiratory failure who were artificially ventilated. Their mean serum phosphorus level was 0.55 +/- 0.18 mmol per liter (normal value, 1.20 +/- 0.10). The contractile properties of the diaphragm were assessed by measuring the transdiaphragmatic pressure generated at functional residual capacity during bilateral supramaximal electrical stimulation of the phrenic nerves. Diaphragmatic function was evaluated in each patient before and after correction of hypophosphatemia, which was achieved by administration of 10 mmol of phosphorus (as KH2PO4) as a continuous infusion for four hours. After phosphate infusion, the mean serum phosphorus level increased significantly (1.33 +/- 0.21 mmol per liter, P less than 0.0001). The increase in serum phosphorus was accompanied by a marked increase in the transdiaphragmatic pressure after phrenic stimulation (17.25 +/- 6.5 cm H2O as compared with 9.75 +/- 3.8 before phosphate infusion, P less than 0.001). Changes in the serum phosphorus level and transdiaphragmatic pressure were well correlated (r = 0.73). These results strongly suggest that hypophosphatemia impairs the contractile properties of the diaphragm during acute respiratory failure, and they emphasize the importance of maintaining normal serum inorganic phosphate levels in such patients.