Abstract
We determined the effects of changing ventilatory stimuli on the hypocapnia-induced apneic and hypopneic thresholds in sleeping dogs. End-tidal carbon dioxide pressure (PetCO2) was gradually reduced during non–rapid eye movement sleep by increasing tidal volume with pressure support mechanical ventilation, causing a reduction in diaphragm electromyogram amplitude until apnea/periodic breathing occurred. We used the reduction in PetCO2 below spontaneous breathing required to produce apnea (ΔPetCO2) as an index of the susceptibility to apnea. ΔPetCO2 was −5 mm Hg in control animals and changed in proportion to background ventilatory drive, increasing with metabolic acidosis (−6.7 mm Hg) and nonhypoxic peripheral chemoreceptor stimulation (almitrine; −5.9 mm Hg) and decreasing with metabolic alkalosis (−3.7 mm Hg). Hypoxia was the exception; ΔPetCO2 narrowed (−4.1 mm Hg) despite the accompanying hyperventilation. Thus, hyperventilation and hypocapnia, per se, widened the ΔPetCO2 thereby protecting against apnea and hypopnea, whereas reduced ventilatory drive and hypoventilation narrowed the ΔPetCO2 and increased the susceptibility to apnea. Hypoxia sensitized the ventilatory responsiveness to CO2 below eupnea and narrowed the ΔPetCO2; this effect of hypoxia was not attributable to an imbalance between peripheral and central chemoreceptor stimulation, per se. We conclude that the ΔPetCO2 and the ventilatory sensitivity to CO2 between eupnea and the apneic threshold are changeable in the face of variations in the magnitude, direction, and/or type of ventilatory stimulus, thereby altering the susceptibility for apnea, hypopnea, and periodic breathing in sleep.
There is considerable evidence that hypocapnia is a major contributor to the genesis of central apnea and periodic breathing during sleep in humans. First, studies using mechanical ventilation to lower PaCO2 reveal that non–rapid eye movement (NREM) sleep unmasks a highly sensitive apneic threshold induced by reductions in arterial carbon dioxide pressure (PaCO2) that were only 2–4 mm Hg less than eupneic PaCO2 (1–4). Second, the occurrence of central apneas in patients with Cheyne–Stokes respiration (CSR) was shown to be preceded by transient hyperpnea and hypocapnia (5–8); and in healthy subjects sleeping in hypoxia the apneic periods during periodic breathing coincided with reductions in end-tidal carbon dioxide pressure (PetCO2), which approximated the subjects' independently determined apneic threshold for Pco2 (1). In addition, the susceptibility to periodic breathing in hypoxia (9) or to CSR in congestive heart failure (CHF) (10–13) tended to coincide with relatively high degrees of ventilatory responsiveness to hypoxia and/or CO2. Third, the occurrence of CSR with central apnea in CHF patients was shown to correlate positively with the level of hypocapnia during eupneic breathing in wakefulness and/or sleep (6, 8, 14, 15). Finally, raising PaCO2 via augmented fraction of inspired CO2 (FiCO2) alleviates apnea and periodic breathing in healthy subjects sleeping in hypoxia (16), in patients with CHF and with CSR (5, 17–19), and in central apnea syndrome (20, 21).
Paradoxically, driving ventilation higher and reducing steady-state background PaCO2 can also alleviate central apnea and/or periodic breathing during sleep. For example, acetazolamide administration causes metabolic acidosis and hyperventilation, and this reduces the amount of periodic breathing during sleep in hypoxia (22). The magnitude of hypoxia-induced periodic breathing is also reduced over the time course of acclimatization to hypoxia as eupneic ventilation increases, and PaCO2 falls further with time (16, 23, 24). Similarly, theophylline or acetazolamide administration to CHF patients with CSR or central apnea alleviates much of the patients' apnea and periodicity even in the face of further reductions in spontaneous background PaCO2 (25–27). Finally, patients with interstitial lung disease do not show an increased prevalence of sleep apnea despite the presence of chronic hyperventilation (28).
Clearly, the compatibility of these two sets of observations would require that the apneic Pco2 threshold change in response to sustained hyperventilation. Furthermore, given the varying susceptibility to apnea and periodic breathing experienced with the various types of ventilatory stimuli, it would appear that not all types of stimuli affect the threshold in a similar way. To date, only limited findings are available on this question of a changing apneic threshold. Older studies used an extrapolation of the hypercapnic ventilatory response to zero V·e intercept to estimate the apneic threshold; but these extrapolations are highly uncertain because of the unknown shape of the CO2 response near and below eupnea (29, 30). Mechanical ventilation may be used to lower PaCO2 and provide a direct measure of the hypocapnia-induced apneic threshold (2, 4). Using variations of this technique in anesthetized rats, acute CO2 inhalation was shown to raise both eupneic and apneic threshold Pco2 and to increase the difference between them (31); and in sleeping healthy humans, mild hypoxia caused a slight hyperventilation and a narrowing of the difference between eupneic PetCO2 and the apneic threshold PetCO2 (32).
We examined the effects of several types of stimuli and inhibitors to respiratory motor output on the apneic threshold, using a mechanical ventilator in the pressure support mode to increase Vt, reduce PaCO2, and cause hypopnea and eventually apnea and periodic breathing in a chronically instrumented, sleeping dog model. We used the difference in PetCO2 between the prevailing spontaneous eupneic PetCO2 and the apneic (or hypopneic) threshold PetCO2 (i.e., ΔPetCO2 = Pco2 at apneic (or hypopneic) threshold − Pco2 spontaneous breathing) as an index of the susceptibility to apnea (or hypopnea). This index is useful in determining the susceptibility to apnea because a small ΔPco2 means that one is breathing very close to the apneic threshold and further transient increases in ventilation would cause apnea, whereas a large ΔPetCO2 means one is further from their apneic threshold during spontaneous eupnea and thus less susceptible to hypocapnic inhibition.
Studies were performed on six unanesthetized female mixed-breed dogs (20–25 kg) during NREM sleep. The dogs were trained to sleep in an air-conditioned (19–22° C) sound-attenuated chamber. Throughout all experiments, the dogs' behavior was monitored by an investigator seated within the chamber and also by closed-circuit television. The Animal Care and Use Committee of the University of Wisconsin approved the surgical and experimental protocols for this study.
Our preparation required two surgical procedures performed under general anesthesia with strict sterile surgical techniques and appropriate postoperative analgesics and antibiotics. In the first procedure, a chronic tracheostomy was created; diaphragm EMG electrodes and a 5-lead EEG montage were also installed. After at least 3 weeks' recovery, a second procedure was required to install indwelling catheters in the abdominal aorta and abdominal vena cava. Catheters and electrode wires were tunneled subcutaneously to the cephalad portion of the dog's back where they were exteriorized. This chronically instrumented model is described in detail elsewhere (33).
The dogs breathed via a cuffed endotracheal tube (10.0 mm outer diameter; Shiley, Irvine, CA), which was inserted into the chronic tracheostomy. Airflow was measured via a heated pneumotachograph system (model 3700; Hans Rudolph, Kansas City, MO, and model MP-45-14-871; Validyne, Northridge, CA) connected to the endotracheal tube. The pneumotachograph was calibrated before each study with four known flows. Tracheal pressure (Ptr) was measured at a port in the endotracheal tube that was connected to a pressure transducer (model MP-45-14-871; Validyne) by means of 1.7 mm inner diameter high durometer PVC tubing (Abbott Laboratories, North Chicago, IL). The pressure transducer was calibrated before each study by applying six known pressures. Airway Po2 and Pco2 were monitored by a mass spectrometer (model MGA-1100; Perkin-Elmer, Norwalk, CT) through a second port in the endotracheal tube. Three to six 1-ml arterial samples were obtained at the start of each experiment from the aortic catheter and analyzed for pH, Po2, and Pco2 on a blood-gas analyzer (model ABL-505; Radiometer, Copenhagen, Denmark). The blood-gas analyzer was validated daily with dog blood tonometered with three different combinations of Po2 and Pco2 covering the range encountered in the experiments. Samples were corrected for both body temperature and systematic errors revealed by tonometry. The inspiratory and expiratory tubes of the ventilator were connected to the pneumotachograph using a Y-connector. A silent balloon valve was placed between the pneumotachograph and the Y-connector such that the dog could breath spontaneously from room air or abruptly switched to pressure support ventilation (PSV) by inflation of the balloon. All signals were digitized (128-Hz sampling frequency) and stored on the hard disk of a personal computer for subsequent analysis. Key signals were also recorded continuously on a polygraph (Gould ES 2000, Cleveland, OH or AstroMed K2G, West Warick, RI). All ventilatory and blood pressure data were analyzed on a breath-by-breath or beat-by-beat basis by means of custom analysis software developed in our laboratory.
Studies were performed over several days during NREM sleep. The animals were unrestrained during the experiments and the body position in which they chose to sleep was not restricted. The apneic threshold for CO2 was determined by means of PSV under five different steady-state levels of background ventilatory drive produced in various ways: normoxia (control), metabolic acidosis, metabolic alkalosis, hypoxia, and almitrine.
Dogs breathed room air spontaneously through the open port in the balloon valve (see Experimental Setup and Measurements). The mechanical ventilator (Veolar, Hamilton Medical) was set in the pressure support mode and the trigger sensitivity was set as low as possible (∼ 1.5 cm H2O). When the balloon was inflated and the low resistance shunt to the room sealed, the ventilator delivered preset levels of inspiratory pressure support whenever the trigger threshold was reached (i.e., a dog set its own frequency; increased pressure support resulted in increased Vt). The expiratory positive airway pressure was set at 0 cm H2O. Each pressure support level was maintained for 2 minutes and then the balloon was deflated and the dog was allowed to breathe spontaneously again. At least 2 minutes elapsed before another PSV trial was performed. PSV was increased in steps of 1–2 cm H2O (range 3–35 cm H2O depending on conditions) until apneas and periodic breathing were observed. Te was measured from the end of the inspiratory flow to the onset of the next EMGdi burst. Periodic breathing was identified visually by the presence of at least three cycles of hyperpnea and apnea as judged by EMGdi with a consistent periodicity (see Results). Further, the apnea lengths had to be at least three standard deviations greater than the baseline Te. The apneic threshold was taken to be the PetCO2 observed in the breath immediately preceding the start of periodic breathing. Once apneas were observed, PSV trials were repeated within 1–2 cm H2O PSV and the mean of these three to five periodic breathing trials taken as the apneic threshold PetCO2 for that animal in the specific condition under study.
A hypopneic threshold was determined from PSV trials in which periodic breathing did not occur. Hypopnea was defined as a reduction in the mean electrical activity (MEA) of the diaphragm EMG greater than two standard deviations from the mean MEA obtained during baseline eupnea of a given trial. The hypopneic threshold was taken to be the PetCO2 observed in the breath immediately preceding the start of a hypopneic breath. For each dog, each of the five experimental conditions (see below) provided an average of 32 (range 11–41) inspiratory efforts that reached a hypopneic threshold within the 2-minute observation period. The mean PetCO2 associated with the reduced EMGdi was used to represent the hypopneic threshold in that animal for that condition. Some trials at low levels of pressure support never achieved a hypopneic threshold. Trials in which there was a state change or sigh were excluded from analysis.
Acetazolamide (31.3–62.5 mg) was given orally 5 hours before the study. A stable metabolic acidosis and hyperventilation (pHa ∼ 7.34; PaCO2 ∼ 30 mm Hg) was achieved for the duration of the PSV trials (2–2.5 hours).
NaHCO3 solution was infused through the indwelling venous catheter at a rate of 0.08 mEq/kg/min for 45–60 minutes to achieve a stable metabolic alkalosis and hypoventilation (pHa ∼ 7.51; PaCO2 ∼ 44 mm Hg).
Moderate hypoxia (PetO2 ∼ 47 mm Hg) was applied by decreasing FiO2. PSV trials were begun after 10–15 minutes of hypoxic exposure. Hypoxia was maintained throughout the PSV trials (up to 1.5 hours).
Was used as a nonhypoxic peripheral chemoreceptor stimulus to ventilation based on evidence that carotid body denervation eliminated about two-thirds of the ventilatory response to almitrine in the awake cat (34) or sleeping dog (unpublished findings—author's laboratory) and that carotid body denervation plus vagotomy (and therefore aortic body denervation) completely eliminated the ventilatory response to almitrine in the anesthetized dog (35). An intravenous bolus of 14-ml (1 mg/ml) almitrine was administered over 60 seconds followed by a continuous infusion at a rate of 0.2–0.4 ml/minute for the duration of the PSV trials (1–1.5 hours). This bolus-infusion protocol resulted in a hyperventilation (PaCO2 ∼ 30 mm Hg) within 5 minutes, which remained stable for the duration of the PSV trials. PSV trials were begun 15 minutes after the start of almitrine infusion.
Nonchemical effects of PSV have been shown to include a significant reduction in the amplitude of respiratory motor output in sleeping humans (36). In the present study, the nonchemical effects of PSV on diaphragm electromyogram (EMGdi) and breath timing were quantified in two ways. First, we analyzed the effects of the first breath of pressure support in all PSV trials, assuming that any observed effects on EMGdi would have occurred before any change in chemoreceptor Pco2. Second, we conducted 29 PSV trials in four dogs in which PetCO2 was prevented from falling by increasing FiCO2.
Sleep state was determined from the polygraph record of each study. NREM sleep was defined as a synchronized low-frequency (< 10 Hz) EEG associated with an absence of rapid eye movements. EEG arousal was defined as desynchronization and speeding (> 10 Hz) of the EEG for more than 3 seconds. Any PSV trials in which the dog changed sleep state were excluded from further analysis.
The group means for all dogs were compared by means of one-way repeated measures analysis of variance (ANOVA) with the Tukey post hoc test for multiple comparisons. Differences were considered significant if p ⩽ 0.05.
Typical examples of the progressive inhibitory effects of PSV on EMGdi amplitude and timing are shown for the same dog in Figures 1A–1C
Figure 1. (A) Polygraph record of one pressure support trial (7 cm H2O) in which apnea/periodicity was not achieved. Note that this level of PSV increased Vt and decreased PetCO2 slightly; EMGdi amplitude was reduced at the onset of pressure support and throughout the trial, Te increased slightly but breathing pattern showed no tendency toward periodicity or apnea. (B) Polygraph record of one pressure support trial (9 cm H2O) in which PetCO2 is further reduced and EMGdi amplitude is reduced to greater than 2 SD below eupneic baseline control on most breaths; but apnea and periodic breathing are not present. Note that after ∼ 2 minutes 15 seconds of continued pressure support in this trial, EMGdi amplitude was reduced on a single inspiration to such an extent (< 30% of baseline eupnea) that the spontaneous inspiratory effort was insufficient to produce enough pressure to trigger the ventilator to deliver the desired volume. (C) Polygraph record of one pressure support trial (11 cm H2O) in which apnea/periodicity was achieved (same dog, same test session as A). Note that a reduced EMGdi and inspiratory effort on the seventh ventilator cycle was insufficient to trigger a ventilator breath. Clear periodicity developed after the ninth ventilator cycle. Arrow marks the PetCO2 considered to be the apneic threshold.
First, in Figure 1A note that EMGdi amplitude was reduced ∼ 30% with no change in Ti or Te on the first breath of pressure support. This reduction in EMGdi is an example of a non-CO2 related, neuromechanical inhibition (also see Figure 4 for mean values and Figure 5 for further evidence of neuromechanical inhibition during PSV). In Figure 1B, conducted at a higher level of pressure support and increased Vt and with further reductions in PetCO2, EMGdi mean electrical activity (MEA) was now reduced further to greater than 2 SD or 40–70% below eupnea in the majority of breaths. Te was prolonged consistently but only by ∼ 30% or 1 second greater than baseline spontaneous eupnea Te. In Figure 1C, conducted at 11 cm H2O pressure support, PetCO2 was reduced 4–5 mm Hg below baseline eupnea. A greatly reduced EMGdi and inspiratory effort that were insufficient to trigger the ventilator are shown at ∼ 1 minute, 15 seconds into this trial, again with little prolongation of Te. Thereafter, regularly appearing apneas (Te > 3 × baseline eupnea) and periodic, cluster-type breathing was observed for the remainder of the trial. This abrupt transition from a gradually reduced amplitude of EMGdi with little change in breath timing, to substantial Te prolongation and periodic breathing pattern was a consistent feature of achieving the apneic threshold during progressive hypocapnia via PSV.
When the dogs breathed room air spontaneously during NREM sleep, they maintained typical canine values of ventilatory and blood gas variables (V·i = 3.3 L/min; PetCO2 = 38.6 mm Hg; pHa = 7.378; [HCO3−] = 21.8 mEq/L; Table 1)
Condition | N | Ti (seconds) | Te (seconds) | fb (b/minute) | Vt (L) | V·i (L/minute) | PaCO2 (mm Hg) | PaO2 (mm Hg) | pHa | [HCO3−]a (mEq/L) |
|---|---|---|---|---|---|---|---|---|---|---|
| Control, normoxic, normocapnic | 6 | 1.45 (0.06) | 3.52 (0.75) | 13.0 (1.4) | 0.26 (0.07) | 3.3 (0.6) | 38.2 (1.9) | 106 (6.4) | 7.378 (0.018) | 21.8 (0.4) |
| Met. Acidosis | 6 | 1.35 (0.33) | 2.95 (1.2) | 16.5 (6.7) | 0.29 (0.12) | 4.1 (0.7) | 30.9 (2.1) | 117 (5.0) | 7.341 (0.023) | 16.4 (1.8) |
| Met. Alkalosis | 5 | 1.61 (0.29) | 3.56 (0.63) | 12.3 (2.2) | 0.24 (0.07) | 2.9 (0.5) | 44.2 (2.1) | 101 (5.2) | 7.507 (0.021) | 34.6 (1.7) |
| Hypoxia | 6 | 1.26 (0.26) | 2.02 (0.61) | 20.0 (5.4) | 0.23 (0.06) | 4.4 (1.0) | 31.0 (1.7) | 47 (1.1) | 7.428 (0.02) | 20.0 (0.7) |
| Almitrine | 3 | 1.06
(0.07) | 1.29
(0.05) | 26.1
(0.3) | 0.20
(0.03) | 5.2
(0.9) | 29.5
(1.1) | 122
(0.8) | 7.427
(0.013) | 19.1
(1.3) |
Figure 2. Changes in the ΔPetCO2 required to achieve the hypopneic and apneic thresholds in dogs during NREM sleep (mean ± SEM) during normal control conditions (normoxic, normocapnia), metabolic acidosis, metabolic alkalosis, hypoxia, and peripheral chemoreceptor stimulation with almitrine during NREM sleep (mean ± SEM). Also see Tables 2 and 3.
[More] [Minimize]Condition (N) | P applied (cm H2O) | Vt (L) and Vt (% of baseline) | ΔPetCO2 (mm Hg) | MEAdi (% of baseline) | Apnea Duration (s) and Apnea Duration (% of baseline) | ΔV·a /ΔPetCO2 (L/minute/mm Hg < eupnea) |
|---|---|---|---|---|---|---|
| Normoxic | 14.8 ± 2.7 | 0.55 ± 0.09 | −5.1 ± 0.8 | 70 ± 22 | 10.0 ± 1.6 | 0.68 ± 0.11 |
| Control, 6 | 207 ± 50 | 286 ± 70 | ||||
| Met. | 24.5 ± 3.9† | 1.2 ± 0.84† | −6.7† ± 0.8 | 57 ± 19 | 9.0 ± 2.7 | 0.70 ± 0.07 |
| Acidosis, 6 | 421 ± 171† | 348 ± 116 | ||||
| Met. | 9.0 ± 0† | 0.40 ± 0.07 | −3.7† ± 1.0 | 69 ± 19 | 8.4 ± 1.5 | 0.84 ± 0.24 |
| Alkalosis, 5 | 174 ± 43 | 239 ± 52 | ||||
| Hypoxia, 6 | 15.7 ± 4.1 | 0.56 ± 0.08 | −4.1†± 0.8 | 76 ± 10 | 7.5 ± 2.1 | 1.07 ± 0.25† |
| 249 ± 59 | 356 ± 104 | |||||
| Almitrine, 4 | 27.0 ± 4.0† | 0.94 ± 0.19 | −5.9† ± 0.8 | 74 ± 24 | 6.3 ± 1.2 | 0.77 ± 0.08 |
| 438 ± 24† | 470 ± 89† |
Condition (N) | P applied (cm H2O) | ΔPetCO2 (mm Hg) | Hypopnea Te (s) and Hypopnea Te (% of baseline) | MEAdi (% of baseline) | Minute ∫ EMGdi (% of baseline) |
|---|---|---|---|---|---|
| Normal, control (6) | 10.9 ± 2.7 | −4.5 ± 1 | 4.6 ± 1.2 | 62 ± 7 | 53 ± 5 |
| 138 ± 18 | |||||
| Met. | 18.6 ± 4.5* | −5.9 ± 2.2 | 4.3 ± 2.8 | 58 ± 8 | 51 ± 11 |
| Acidosis, 6 | 138 ± 41 | ||||
| Met. | 7.4 ± 0.8 | −3.2 ± 0.8* | 4.1 ± 0.3 | 66 ± 4 | 66 ± 15 |
| Alkalosis, 5 | 116 ± 17 | ||||
| Hypoxia, 6 | 9.4 ± 2.5 | −2.1 ± 1* | 2.6 ± 0.9 | 64 ± 12 | 64 ± 19 |
| 127 ± 21 | |||||
| Almitrine, 4 | 26.5 ± 6.7* | −4.6 ± 1.1 | 2.3 ± 0.9* | 56 ± 6 | 51 ± 11 |
| 169 ± 48 |
After a stable metabolic acidosis had been achieved (pH= 7.341 versus 7.378; [HCO3−]= 16.4 versus 21.8 mEq/L; Table 1) all dogs manifested stable hyperventilation (V·i = 4.8 versus 3.4 L/minute for control; PetCO2 = 28.2 versus 38.6 mm Hg) indicating a substantial increase in background ventilatory drive during NREM sleep. To reach the apneic threshold in this background of increased drive, an increase in V·i (via increased Vt from PSV) sufficient to reduce the PetCO2 to 21.5 mm Hg was required. Thus, ΔPetCO2 was −6.7 ± 0.8 mm Hg, which represented a significant widening over that in control conditions (Figure 2). The average gain of the slope of the ventilatory reduction in response to hypocapnia below eupnea was not significantly different from normoxic eupnea (Table 2). The ΔPetCO2 for hypopnea with metabolic acidosis was widened (Table 3) to a similar extent as was the ΔPetCO2 for the apneic threshold.
After a stable metabolic alkalosis had been achieved in NREM sleep (pHa = 7.507 versus 7.378; [HCO3−] = 34.6 versus 21.8 mEq/L); Table 1) all dogs hypoventilated (V·i = 2.9 versus 3.4 L/minute; PetCO2 = 43.7 versus 38.6 mm Hg; Figure 3)
Figure 3. (A) Mean data showing the ΔPetCO2 from spontaneous ventilation to the apneic threshold as a function of the spontaneous PetCO2 achieved during all background ventilatory drive conditions. Note the inverse relationship between spontaneous PetCO2 and ΔPetCO2 for all conditions except hypoxia. (B) Mean slope of the ventilatory response to reduced Pco2 below eupnea (ΔV·a/ΔPetCO2) for all five experimental conditions. Note the large increase in gain during hypoxia in contrast to the near-control values observed during acidosis and almitrine at comparable levels of reduced spontaneous PetCO2 during spontaneous eupnea. Diamonds = normal, squares = acidosis, triangles = alkalosis, circles = hypoxia, × = almitrine. * indicates a mean value for ΔPetCO2 or ΔV·a/ΔPetCO2 significantly different from control.
[More] [Minimize]Moderate hypoxia during NREM sleep (PetO2 = 47.2 mm Hg) resulted in stable hyperventilation and respiratory alkalosis in all dogs (V·i = 4.6 versus 3.4 L/minute; PetCO2 = 31.0 versus 38.6 mm Hg; pHa = 7.428 versus 7.378; [HCO3−] = 20.0 versus 21.8 mEq/L; Table 1) indicating a substantial increase in ventilatory drive. To reach the apneic threshold in this background of increased drive an increase in V·i via increased Vt from PSV sufficient to reduce the PetCO2 to 26.9 mm Hg was required; i.e., relative to control there was a significantly narrowed ΔPetCO2 in all dogs (mean = −4.1 ± 0.8 mm Hg; Figure 2). The average gain of the slope of the ventilatory reduction in response to hypocapnia was increased 57 ± 37% greater than control (p < 0.05; Table 2). Hypoxia also caused a narrowing of the group mean hypopneic threshold ΔPetCO2 to −2.1 ± 1 mm Hg (p < 0.05)(see Table 3); a reduction which was more than 50% below normoxic control conditions and substantially greater than the average 20% fall in the apneic threshold ΔPetCO2.
Infusion of almitrine bismesylate (see Methods) during NREM sleep resulted in a stable hyperventilation in all dogs (V·i = 5.2 versus 3.4 L/minute; PetCO2 = 28.8 versus 38.6 mm Hg; pHa = 7.427 versus 7.378; [HCO3−] = 19.1 versus 21.8 mEq/L; Table 1), indicating a substantial increase in ventilatory drive. To reach the apneic threshold in this background of increased drive, an increase in V·i via increased Vt from PSV sufficient to reduce the PetCO2 to 22.9 mm Hg was required; i.e., relative to control there was a significantly widened ΔPetCO2 in all dogs (mean = −5.9 ± 0.8 mm Hg versus −5.1 ± 0.8 control (p < 0.05; Figure 2). The average gain of the ventilatory response to CO2 below eupnea was unchanged from control (Table 2). During almitrine administration, the ΔPetCO2 from eupnea to hypopnea was unchanged from control.
As summarized in Figure 3A, increased ventilatory drives of nonhypoxic origin tended to widen the ΔPetCO2 required for apnea whereas decreased ventilatory drive tended to narrow it. Hypoxia was the exception; the ΔPetCO2 was narrower than control and much narrower than one would predict from the substantial accompanying hyperventilation. The slope of the fall in V·a per mm Hg reduction in PetCO2 below eupnea (Figure 3B) was unchanged from control with metabolic acidosis and almitrine, increased slightly but not significantly above control with metabolic alkalosis, and increased significantly (∼ 40% greater than control) with hypoxia.
PSV, per se, decreased the amplitude of the integrated EMGdi 20 ± 2% during the first PSV breath and prolonged Te an average of 32 ± 18% (i.e., 1.1 ± 0.6 seconds > eupneic control; Figure 4)
Figure 4. Nonchemical effect of PSV on EMGdi mean electrical activity (MEAdi) and Te obtained from the first pressure support breath of all PSV trials. Note that unloading caused a 20 ± 2% decrease in the overall group mean MEAdi and a 32 ± 18% increase in Te (relative to baseline spontaneous breathing) within the first ventilator breath. Baseline control Te averaged 2.9 seconds, thus the mean increase in Te was 0.9 second. Data from all pressure support levels for a given background ventilatory drive condition are combined, as the absolute level of pressure support was not correlated with the magnitude of reduction in EMGdi amplitude.
[More] [Minimize]
Figure 5. Polygraph record of an isocapnic trial of PSV. Note the immediate and persistent fall in EMGdi with little change in breath timing during PSV. The arrow indicates the first pressure support breath. FiCO2 was increased during PSV to prevent hypocapnia.
[More] [Minimize]Our findings show that the hypocapnia-induced apneic threshold and, more importantly, the proximity of the spontaneous Pco2 to the apneic threshold Pco2 below eupnea (i.e., ΔPetCO2) changed in proportion to the background ventilatory drive. The exception to this finding was hypoxia, during which the ventilatory sensitivity to CO2 below eupnea was markedly increased and ΔPetCO2 was reduced significantly below control despite a substantial increase in ventilatory drive and a reduction in spontaneous PaCO2. In turn, this effect of hypoxia on ΔPetCO2 was not attributable to an imbalance between peripheral and central chemoreceptor stimulation per se; to the contrary, nonhypoxic peripheral chemoreceptor stimulation caused an increase in ΔPetCO2 similar to that attending the hyperventilation of metabolic acidosis. The relevance of these findings to causes of apnea during sleep is discussed below.
During NREM sleep chemoreceptors provide the dominant control over central respiratory motor output (1, 37). In order for apnea or hypopnea to develop during sleep, sufficient sensitivity of ventilatory responses and an adequate ventilatory stimulus must be present to produce first, a transient ventilatory overshoot above eupneic levels, commonly in response to changes in sleep state and/or airway resistance (30, 38); and subsequently, sufficient sensitivity to CO2 below eupnea must be present in order for hypocapnia to depress central respiratory motor output and cause apnea or hypopnea. Modelers of the ventilatory control system have identified two types of gain which predict predisposition to periodic breathing, namely: “controller gain”, which indicates the strength of the chemoreflex and is conventionally tested by the steady-state or transient ventilatory response to increased CO2; and “plant gain” or the amplification with which any ventilatory overshoot is translated into a reduction in PaCO2. The product of these two gains or “loop gain” is considered a sensitive determinant of unstable ventilation (30).
ΔPco2 may be viewed as an integral part of the “controller gain” component because it defines the chemoreceptor gain to hypocapnia below eupnea; whereas controller gain, as currently conventionally defined, speaks only to the susceptibility to ventilatory overshoot and only to one potential determinant (i.e., sensitivity to hypercapnia) of the overshoot. Thus, the closer the proximity of eupneic Pco2 to the apneic or hypopneic threshold Pco2 (i.e., the narrower the ΔPco2) the more likely it is that additional transient perturbations in ventilation above eupnea will precipitate significant ventilatory undershoots. As mentioned previously, in addition to the change in ΔPco2 with changing ventilatory drive, any change in the susceptibility to apnea would also be determined in part by the magnitude of the further increase in ventilation required to achieve this ΔPco2 (i.e., “plant gain”).
Both of these determinants of apnea are shown together in Figure 6
Figure 6. Two principal determinants of apnea: (A) the ΔPetCO2 from spontaneous breathing to the apneic threshold; and (B) the change in ventilation above eupnea required to lower PetCO2 sufficiently to reach the apneic threshold (ΔPetCO2), are shown during conditions of control, metabolic acidosis and alkalosis, and in hypoxia in the dog (present study) and in normoxia and hypoxia in the sleeping human (32). The iso-metabolic line describing the relationship between PaCO2 and V·a is theoretical and was constructed using an assumed constant V·co2 of 150 ml/minute for both species and the measured PetCO2 during spontaneous eupneic breathing and at the apneic thresholds (PaCO2 = V·co2/V·a • k). For example, in the dog under control conditions the ΔPetCO2 was 5.1 mm Hg below spontaneous eupnea and the increase in V·a from spontaneous eupnea required to reach the apneic threshold was 0.5 L/minute. The diagonal dashed lines join eupneic and apneic points, and their slopes indicate the gain below eupnea of the ventilatory response to hypocapnia in each condition. Note the increased slope of the V·a:PaCO2 relationship between eupnea and the apneic threshold in hypoxia in both dogs and humans, whereas these slopes remained unchanged from control with metabolic acidosis and alkalosis.
[More] [Minimize]With hypoxia-induced hyperventilation, (see Figure 6B) the protective effect of the greater further increase in ventilation required to produce a given reduction in PaCO2 (because of the shift in position on the iso-metabolic hyperbola) was offset by a narrowed ΔPco2 below eupnea. This was true in both the sleeping dog and especially in the human, in whom acute mild hypoxia barely increased spontaneous ventilation and reduced the PaCO2 only 2–3 mm Hg, but ΔPco2 was reduced to less than one-half normoxic control conditions (32). Interestingly, in both the human and dog, hypoxia narrowed the ΔPco2 for hypopnea (below normoxic control) substantially more than the ΔPco2 for apnea (see reference 32 and Figure 2).
There are two potential reasons why the ΔPco2 from spontaneous ventilation to the apneic threshold changes with changing background ventilatory drive. First, if only eupneic ventilation increased and PaCO2 fell, i.e., with no change in true ventilatory responsiveness to CO2 below eupnea (see below), then the ΔPetCO2 must widen simply because a greater reduction in Pco2 is required to cause the greater reduction in ventilation to produce apnea. This change in spontaneous eupneic ventilation was the major reason for the widened ΔPco2 that accompanied the hyperventilation during metabolic acidosis or almitrine and the narrowed ΔPco2 found during the hypoventilation of metabolic alkalosis (see Figure 6).
Second, an alteration in the gain of the slope of the fall in ventilation in response to the reduced Pco2 below eupnea (ΔV·a/ΔPetCO2) will also affect the magnitude of the ΔPco2. These changes in sensitivity may even offset the above-mentioned effects of changes in background spontaneous ventilation. For example, despite the comparable levels of hyperventilation accompanying metabolic acidosis and hypoxia, the ΔPco2 was significantly greater than control in the former and yet less than control in the latter. This difference in ΔPetCO2 occurred because the ΔV·a/ΔPetCO2 below eupnea was unchanged with metabolic acidosis and markedly increased (to 50% > control) with hypoxia.
Several investigators have observed that CHF patients with CSR have chronically reduced eupneic PaCO2, both awake and asleep (5–8, 15). We note that not all studies show chronic hyperventilation in patients with CHF and CSR (39, 40). Based on these correlative data it has been suggested that the reduced eupneic PaCO2 in these patients is moved closer to their apneic threshold and therefore renders the patient more susceptible to apnea and periodic breathing. To the contrary, our results show that increases in ventilatory drive (via nonhypoxic stimuli) that cause hyperventilation protect against apnea and periodic breathing by both widening the ΔPco2 and requiring a greater further increase in V·a for a given ΔPco2 (see Figure 6). Alternatively, the decreased ventilatory drive and an increased eupneic PaCO2 that accompanied metabolic alkalosis sensitized the ventilatory depression in response to hypocapnia. For these same reasons, the present findings also explain why adding more ventilatory drive in subjects already experiencing apnea and/or periodic breathing (by means of theophylline or acetazolamide) results in a significant lessening or even elimination of apnea and periodic breathing (25–27). Furthermore, perhaps even the reported effects of acutely increasing PaCO2 via added FiCO2 on removing central sleep apnea and stabilizing breathing (5, 16, 17, 20, 21) may be due primarily to the nonspecific effects on ΔPetCO2 of increasing background ventilatory drive and ventilation rather than to a raised PaCO2, per se.
Whether or not increased ventilatory drive and/or reduced eupneic PaCO2 protects against and/or eliminates apnea and periodic breathing on the one hand or may actually create conditions favorable to apnea on the other, may well depend on the specific stimulus causing the hyperventilation. We found that metabolic acidosis and pharmacologic stimulation of (primarily) the peripheral chemoreceptors (via almitrine) increased ΔPetCO2 and protected against apnea, whereas hypoxia did not. The cause of increased ventilatory drive in CHF patients with CSR has been correlated (cross-sectionally and over time) to increased pulmonary-capillary wedge pressure (40, 41). In turn, this increased drive to breathe is likely mediated by stimulation of pulmonary C-fiber endings secondary to pulmonary vascular congestion and edema in CHF (42). Perhaps hyperventilation caused primarily by sensory input from these receptors would—like hypoxia—increase the ventilatory sensitivity to hypocapnia, and therefore reduce the ΔPco2 required to render the patient more susceptible to further transient ventilatory overshoots, which would lead to apnea, hypopnea, or periodic breathing. The effects of several types of specific ventilatory stimuli on the apneic threshold need to be determined. Recent findings in CHF patients with CSR do show a significant reduction in the ΔPetCO2 below eupnea (43).
When carotid chemoreceptor stimulation—relative to that at the level of the medullary chemoreceptors—becomes the dominant sensory input to the respiratory controller, this is believed to precipitate instability in breathing pattern. Several lines of evidence are commonly cited to support this hypothesis, including the periodic breathing produced in humans with carotid body stimulation via hypoxia (16, 44), adenosine infusion (45)—which has both a peripheral stimulatory and central depressant effect (46, 47), or in anesthetized animals via blockade of either the carotid chemoreceptor inhibitory neurotransmitter, dopamine (48), or the medullary chemoreceptors (49). Our present findings support the idea that hypocapnic hypoxia enhances the susceptibility to apnea and hypopnea by reducing ΔPetCO2. An additional question posed in the present study was whether this apneic-producing effect of hypoxic hypocapnia was attributable to the imbalance between peripheral and central chemoreception.
We determined the effects of an imbalance between peripheral and central chemoreceptors on the apnea and hypopneic thresholds ΔPetCO2 by using a nonhypoxic peripheral chemoreceptor stimulant, almitrine (35—also see Methods), to cause hyperventilation. Like hypoxia, almitrine increases the slope of the ventilatory and/or the carotid sinus nerve response to hypercapnia in most studies in humans and cats (50–52); although some indirect evidence suggests that hypoxia and almitrine might not have the same site and/or mechanism of stimulation of the carotid chemoreceptors (50). We found that, unlike hypoxia, almitrine-induced hyperventilation: (a) widened the ΔPetCO2 similar to that found during metabolic acidosis-induced hyperventilation; and (b) caused a small increase in the ΔV·a/ΔPetCO2 response slope below eupnea which was only about one-fifth that observed in hypoxia (see Table 2). Furthermore, the reduced plant gain accompanying the hyperventilation induced by almitrine combined with an increased ΔPetCO2 required to reach the apneic or hypopneic thresholds would mean considerable protection against transient hypopneas or apneas in sleep.
So, the contrast of the apnea-producing effects of hypoxia versus the stabilizing effects of hypocapnic hyperventilation induced via nonhypoxic peripheral chemoreceptor stimulation—means characteristics of hypoxia itself, rather than simply an imbalance between peripheral and medullary chemoreceptor stimuli—is likely responsible for most of the marked enhancement of the CO2 sensitivity below eupnea. Our study provides no information on the source of this hypoxic effect, which enhanced the gain of the ventilatory inhibition in response to hypocapnia. It may reflect a unique sensory input from the carotid chemoreceptors that is specific to hypoxia (50); or a vasodilatory effect of hypoxia on the cerebral circulation that would promote apnea and hypopnea by lowering medullary chemoreceptor Pco2 at any given arterial Pco2 (53).
The ΔPco2 below spontaneous eupneic PaCO2 required for apnea or significant reductions in EMGdi amplitude, as determined using pressure support ventilation, is an index of the propensity for apnea and hypopnea during sleep. Our findings have shown that the ΔPco2 during sleep changes in proportion to a changing background ventilatory drive. The exception was hypoxia during which ΔPco2 was reduced to less than normoxic, normocapnic control despite an increase in background ventilatory drive. In contrast with some current concepts concerning causes of central sleep apnea, our findings suggest that alveolar hyperventilation or (nonhypoxic) peripheral chemoreceptor stimulation (in combination with systemic and central alkalosis) all coincided with a widened ΔPetCO2 below eupnea, thereby protecting against (rather than enhancing) the occurrence of central apnea, hypopnea, and unstable breathing. In contrast, reduced ventilatory drive and the accompanying hypoventilation precipitated apnea and hypopnea.
The many contributions of Kathleen S. Henderson are gratefully acknowledged. We acknowledge the Servier Pharmaceutical Company for supplying the almitrine.
| 1. | Skatrud JB, Dempsey JA. Interaction of sleep state and chemical stimuli in sustaining rhythmic ventilation. J Appl Physiol 1983;55:813–822. |
| 2. | Henke KG, Arias A, Skatrud JB, Dempsey JA. Inhibition of inspiratory muscle activity during sleep. Chemical and nonchemical influences. Am Rev Respir Dis 1988;138:8–15. |
| 3. | Datta AK, Shea SA, Horner RL, Guz A. The influence of induced hypocapnia and sleep on the endogenous respiratory rhythm in humans. J Physiol 1991;440:17–33. |
| 4. | Meza S, Mendez M, Ostrowski M, Younes M. Susceptibility to periodic breathing with assisted ventilation during sleep in normal subjects. J Appl Physiol 1998;85:1929–1940. |
| 5. | Lorenzi-Filho G, Rankin F, Bies I, Bradley TD. Effects of inhaled carbon dioxide and oxygen on Cheyne-Stokes respiration in patients with heart failure. Am J Respir Crit Care Med 1999;159:1490–1498. |
| 6. | Mortara A, Sleight P, Pinna GD, Maestri R, Prpa A, La Rovere MT, Cobelli F, Tavazzi L. Abnormal awake respiratory patterns are common in chronic heart failure and may prevent evaluation of autonomic tone by measures of heart rate variability. Circulation 1997;96:246–252. |
| 7. | Bradley TD, Floras JS. Pathophysiologic and therapeutic implications of sleep apnea in congestive heart failure. J Card Fail 1996;2:223–240. |
| 8. | Naughton M, Benard D, Tam A, Rutherford R, Bradley TD. Role of hyperventilation in the pathogenesis of central sleep apneas in patients with congestive heart failure. Am Rev Respir Dis 1993;148:330–338. |
| 9. | Lahiri S, Maret K, Sherpa MG. Dependence of high altitude sleep apnea on ventilatory sensitivity to hypoxia. Respir Physiol 1983;52:281–301. |
| 10. | Wilcox I, McNamara SG, Dodd MJ, Sullivan CE. Ventilatory control in patients with sleep apnoea and left ventricular dysfunction: comparison of obstructive and central sleep apnoea. Eur Respir J 1998;11:7–13. |
| 11. | Javaheri S. A mechanism of central sleep apnea in patients with heart failure. N Engl J Med 1999;341:949–954. |
| 12. | Solin P, Roebuck T, Johns DP, Walters EH, Naughton MT. Peripheral and central ventilatory responses in central sleep apnea with and without congestive heart failure. Am J Respir Crit Care Med 2000;162:2194–2200. |
| 13. | Xie A, Rutherford R, Rankin F, Wong B, Bradley TD. Hypocapnia and increased ventilatory responsiveness in patients with idiopathic central sleep apnea. Am J Respir Crit Care Med 1995;152:1950–1955. |
| 14. | Javaheri S, Corbett WS. Association of low PaCO2 with central sleep apnea and ventricular arrhythmias in ambulatory patients with stable heart failure. Ann Intern Med 1998;128:204–207. |
| 15. | Hanly P, Zuberi N, Gray R. Pathogenesis of Cheyne-Stokes respiration in patients with congestive heart failure. Relationship to arterial PCO2. Chest 1993;104:1079–1084. |
| 16. | Berssenbrugge A, Dempsey J, Iber C, Skatrud J, Wilson P. Mechanisms of hypoxia-induced periodic breathing during sleep in humans. J Physiol 1983;343:507–526. |
| 17. | Steens RD, Millar TW, Su X, Biberdorf D, Buckle P, Ahmed M, Kryger MH. Effect of inhaled 3% CO2 on Cheyne-Stokes respiration in congestive heart failure. Sleep 1994;17:61–68. |
| 18. | Green JA. Clinical studies on respiration. IV. Some observations on Cheyne-Stokes respiration. Arch Intern Med 1933;52:454–463. |
| 19. | Anthony AJ, Cohn AE, Steele JM. Studies on Cheyne-Stokes respiration. J Clin Invest 1932;11:1321–1341. |
| 20. | Xie A, Rankin F, Rutherford R, Bradley TD. Effects of inhaled CO2 and added dead space on idiopathic central sleep apnea. J Appl Physiol 1997;82:918–926. |
| 21. | Badr MS, Grossman JE, Weber SA. Treatment of refractory sleep apnea with supplemental carbon dioxide. Am J Respir Crit Care Med 1994; 150:561–564. |
| 22. | Sutton JR, Houston CS, Mansell AL, McFadden MD, Hackett PM, Rigg JR, Powles AC. Effect of acetazolamide on hypoxemia during sleep at high altitude. N Engl J Med 1979;301:1329–1331. |
| 23. | White DP, Gleeson K, Pickett CK, Rannels AM, Cymerman A, Weil JV. Altitude acclimatization: influence on periodic breathing and chemoresponsiveness during sleep. J Appl Physiol 1987;63:401–412. |
| 24. | Weil JV. Sleep at high altitude. Clin Chest Med 1985;6:615–621. |
| 25. | DeBacker WA, Verbraecken J, Willemen M, Wittesaele W, DeCock W, Van deHeyning P. Central apnea index decreases after prolonged treatment with acetazolamide. Am J Respir Crit Care Med 1995;151:87–91. |
| 26. | Javaheri S, Parker TJ, Wexler L, Liming JD, Lindower P, Roselle GA. Effect of theophylline on sleep-disordered breathing in heart failure. N Engl J Med 1996;335:562–567. |
| 27. | White DP, Zwillich CW, Pickett CK, Douglas NJ, Fidley MD, Weil JV. Central sleep apnoea: improvement with acetazolamide therapy. Arch Intern Med 1982;142:1816–1819. |
| 28. | Perez-Padilla R, West P. Lertzman, Kryger MH. Breathing during sleep in patients with interstitial lung disease. Am Rev Respir Dis 1985;132:224–229. |
| 29. | Cunningham DJC, Robbins PA, Wolff CB. Integration of respiratory responses to changes in alveolar partal pressures of CO2 and O2 and in arterial pH. In: Fishman AP, editor. Handbook of physiology: the respiratory system. Vol. II. Bethesda: American Physiological Society; 1986. p. 475–528. |
| 30. | Khoo MCK. Determinants of ventilatory instability and variability. Respir Physiol 2000;122:167–182. |
| 31. | Boden AG, Harris MC, Parkes MJ. Apneic threshold for CO2 in the anesthetized rat: fundamental properties under steady-state conditions. J Appl Physiol 1998;85:898–907. |
| 32. | Xie A, Skatrud JB, Dempsey JA. Effect of hypoxia on the hypopnoeic and apnoeic threshold for CO2 in sleeping humans. J Physiol 2001;535:269–278. |
| 33. | Satoh M, Eastwood PR, Smith CA, Dempsey JA. Nonchemical elimination of inspiratory motor output via mechanical ventilation in sleep. Am J Respir Crit Care Med 2001;163:1356–1364. |
| 34. | Gautier H, Bonora M. Effects of hypoxia and respiratory stimulants in conscious intact and carotid denervated cats. Bull Eur Physiopathol Respir 1982;18:565–582. |
| 35. | Laubie M, Diot F. Etude pharmacologique de l'action stimulante respiratoire du S. 2620. Role des chemorecepteurs carotidiens et aortiques. J Pharmacol (Paris) 1972;3:363–374. |
| 36. | Wilson CR, Satoh M, Skatrud JB, Dempsey JA. Non-chemical inhibition of respiratory motor output during mechanical ventilation in sleeping humans. J Physiol 1999;518:605–618. |
| 37. | Phillipson EA, Bowes G. Control of breathing during sleep. In: Cherniack NS, Widdicombe JG, editors. Handbook of physiology: the respiratory system. Bethesda: American Physiological Society; 1986. p. 649–689. |
| 38. | Khoo MC, Koh SS, Shin JJ, Westbrook PR, Berry RB. Ventilatory dynamics during transient arousal from NREM sleep: implications for respiratory control stability. J Appl Physiol 1996;80:1475–1484. |
| 39. | Javaheri S. Central sleep apnea-hypopnea syndrome in heart failure: prevalence, impact, and treatment. Sleep 1996;19:S229–231. |
| 40. | Solin P, Bergin P, Richardson M, Kaye DM, Walters EH, Naughton MT. Influence of pulmonary capillary wedge pressure on central apnea in heart failure. Circulation 1999;99:1574–1579. |
| 41. | Tkacova R, Liu PP, Naughton MT, Bradley TD. Effect of continuous positive airway pressure on mitral regurgitant fraction and atrial natriuretic peptide in patients with heart failure. J Am Coll Cardiol 1997;30:739–745. |
| 42. | Roberts AM, Bhattacharya J, Schultz HD, Coleridge HM, Coleridge JC. Stimulation of pulmonary vagal afferent C-fibers by lung edema in dogs. Circ Res 1986;58:512–522. |
| 43. | Xie A, Skatrud JB, Puleo DS, Rahko PS, Dempsey JA. Apnea–hypopnea threshold for CO2 in patients with congestive heart failure. Am J Respir Crit Care Med 2002;165:1245–1250. |
| 44. | Khoo MC, Anholm JD, Ko SW, Downey R III, Powles AC, Sutton JR, Houston CS. Dynamics of periodic breathing and arousal during sleep at extreme altitude. Respir Physiol 1996;103:33–43. |
| 45. | Gleeson K, Zwillich CW. Adenosine infusion and periodic breathing during sleep. J Appl Physiol 1992;72:1004–1009. |
| 46. | Eldridge FL, Millhorn DE, Kiley JP. Respiratory effects of a long-acting analog of adenosine. Brain Res 1984;301:273–280. |
| 47. | Eldridge FL, Millhorn DE, Kiley JP. Antagonism by theophylline of respiratory inhibition induced by adenosine. J Appl Physiol 1985;59: 1428–1433. |
| 48. | Lahiri S, Smatresk N, Pokorski M, Barnard P, Mokashi A, McGregor KH. Dopaminergic efferent inhibition of carotid body chemoreceptors in chronically hypoxic cats. Am J Physiol 1984;247:R24–R28. |
| 49. | Cherniak N, Longobordo GS. Periodic breathing in sleep. In: Saunders NA, Sullivan CE, editors. Sleep and breathing, 2nd ed. New York: Marcel Dekker; 1994. p. 157–190. |
| 50. | Lahiri SA, Mokashi W, Huang AK. Sherpa DiGiulio C. Stimulus interaction between CO2 and almitrine in the cat carotid chemoreceptors. J Appl Physiol 1989;67:232–238. |
| 51. | Stradling JR, Barnes P, Pride NB. The effects of almitrine on the ventilatory response to hypoxia and hypercapnia in normal subjects. Clin Sci Lond 1982;63:401–404. |
| 52. | Stanley NN, Galloway JM, Flint KC, Campbell DB. Increased respiratory chemosensitivity induced by oral almitrine in healthy man. Br J Dis Chest 1983;77:136–146. |
| 53. | Brown DL, Lawson EE. Brain stem extracellular fluid pH and respiratory drive during hypoxia in newborn pigs. J Appl Physiol 1988;68: 1055–1059. |
