Reduced respiratory-related evoked activity in subjects with obstructive sleep apnea syndrome.

Midlatency respiratory-related evoked potentials were measured during wakefulness by using a 60-electrode array placed over the cortical region of the scalp. We studied the responses evoked by 200-ms pressure pulses at -5 and -10 cmH(2)O applied at inspiratory onset and during control tests (no pressure applied) in 14 subjects with obstructive sleep apnea syndrome (OSAS) and 18 normal subjects. Wavelet decomposition was used to smooth and dissect the respiratory-related evoked potentials in frequency and time in 8 frequency bands. After denoising, selected wavelet scales were used to reconstruct the respiratory-related evoked potentials, which were quantified by using global field power estimates. The time course of the global field power activity in OSAS subjects compared with normal subjects was significantly depressed in the period 55-70 ms poststimulus onset, a time when afferent traffic from upper airway receptors arrives in normal subjects. The reduced evoked response in subjects with OSAS suggests that these subjects receive less afferent input from upper airway mechanoreceptors. This may reflect reduced sensitivity of mechanoreceptors or reduced mechanoreceptor stimulation due to decreased upper airway compliance during wakefulness in OSAS.     

Respiratory-related cortical potentials evoked by inspiratory occlusion in humans

It has long been recognized that humans can perceive respiratory loads. There have been several studies on the detection and psychophysical quantification of mechanical load perception. This investigation was designed to record cortical sensory neurogenic activity related to inspiratory mechanical loading in humans. Inspiration was periodically occluded in human subjects while the electroencephalographic (EEG) activity in the somatosensory region of the cerebral cortex was recorded. The onset of inspiratory mouth pressure (Pm) was used to initiate signal averaging of the EEG signals. Cortical evoked potentials elicited by inspiratory occlusions were observed when C3 and C alpha were referenced to CZ. This evoked potential was not observed with the control (unoccluded) breaths. There was considerable subject variability in the peak latencies that was related to the differences in the inspiratory drive, as measured by occlusion pressure (P0.1). The results of this study demonstrate that neurogenic activity can be recorded in the somatosensory region of the cortex that is related to inspiratory occlusions. The peak latencies are longer than analogous somatosensory evoked potentials elicited by stimulation of the hand and foot. It is hypothesized that a portion of this latency difference is related to the time required for the subject to generate sufficient inspiratory force to activate the afferents mediating the cortical response.     

Effects of timing of inspiratory occlusion on cerebral evoked potentials in humans

Previous studies from these laboratories have shown that airway occlusion applied from the onset of inspiration or during midinspiration is associated with cerebral evoked potentials in human subjects. The hypothesis tested in the present study was that the more abrupt decrease in mouth pressure produced by midinspiratory occlusion will be associated with evoked potentials that have shorter peak latencies and greater peak amplitudes than those produced by occlusions from the onset of inspiration. The second objective of the present study was to determine whether there is bilateral projection of inputs from the respiratory system to the somatosensory cortex. Random presentation of 64 midinspiratory occlusions and 64 occlusions from the onset of inspiration was performed in eight subjects. The inspirations preceding the occlusions served as control. Evoked potentials were recorded from the scalp with electrode pairs Cz-C3 and Cz-C4. Reaction time to each type of occlusion was measured from the burst in electromyogram activity produced by contraction of the muscles encircling the eye. Each type of inspiratory occlusion was associated with evoked potentials that could be recorded bilaterally. The peak amplitudes of the evoked potentials recorded over the right cerebral hemisphere were significantly greater than those recorded from the left side. The peak amplitude was greater and the peak latency shorter for the evoked potentials produced by the midinspiratory occlusions. The results are consistent with the hypothesis that afferents mediating these potentials are stimulated by added loads to breathing and project bilaterally to the somatosensory cortex in humans.     

Consequences of periodic augmented breaths on tongue muscle activities in hypoxic rats.

This study was designed to investigate the influence of hypoxia-evoked augmented breaths (ABs) on respiratory-related tongue protrudor and retractor muscle activities and inspiratory pump muscle output. Genioglossus (GG) and hyoglossus (HG) electromyogram (EMG) activities and respiratory-related tongue movements were compared with peak esophageal pressure (Pes; negative change in pressure during inspiration) and minute Pes (Pes x respiratory frequency = Pes/min) before and after ABs evoked by sustained poikilocapnic, isocapnic, and hypercapnic hypoxia in spontaneously breathing, anesthetized rats. ABs evoked by poikilocapnic and isocapnic hypoxia triggered long-lasting (duration at least 10 respiratory cycles) reductions in GG and HG EMG activities and tongue movements relative to pre-AB levels, but Pes was reduced transiently (duration of <10 respiratory cycles) after ABs. Adding 7% CO(2) to the hypoxic inspirate had no effect on the frequency of evoked ABs, but this prevented long-term declines in tongue muscle activities. Bilateral vagotomy abolished hypoxia-induced ABs and stabilized drive to the tongue muscles during each hypoxic condition. We conclude that, in the rat, hypoxia-evoked ABs 1) elicit long-lasting reductions in protrudor and retractor tongue muscle activities, 2) produce short-term declines in inspiratory pump muscle output, and 3) are mediated by vagal afferents. The more prolonged reductions in pharyngeal airway vs. pump muscle activities may lead to upper airway narrowing or collapse after spontaneous ABs.     

Effect of positive nasal pressure on upper airway pressure-flow relationships

To determine the influence of changes in nasal pressure (Pn) on airflow mechanics in the upper airway, we examined the effect of elevations in Pn on upper airway resistance and critical pressure (Pcrit) during stage I/II sleep in six patients with obstructive sleep apnea. When Pn was elevated above a Pcrit, periodic occlusions of the upper airway were eliminated and inspiratory airflow limitation was demonstrated by the finding that inspiratory airflow (VI) became maximal (VImax) and independent of fluctuations in hypopharyngeal pressure (Php) when Php fell below a specific Php (Php'). As Pn was elevated, VI vs. Php demonstrated 1) marked decreases in early and late inspiratory resistances from 75.9 +/- 34.7 and 54.6 +/- 19.0 to 8.0 +/- 1.7 and 7.6 +/- 1.6 cmH2O.l-1.s (P less than 0.05), respectively, and 2) increases in early and late inspiratory Php' to levels that exceeded Pcrit by 3.0 +/- 0.6 and 3.1 +/- 0.7 cmH2O, respectively, at the highest level of Pn applied (P less than 0.01). This latter finding suggests that elevations in Pn result in increases in Pcrit. We suggest that elevations in Pn produce distinct alterations in upper airway resistance and collapsibility, which may influence oppositely the level of airflow through the upper airway during sleep.     

Posterior cricoarytenoid muscle activity during wakefulness and sleep in normal adults

Six normal adults were studied 1) to compare respiratory-related posterior cricoarytenoid (PCA) muscle activity during wakefulness and sleep and 2) to determine the effect of upper airway occlusions during non-rapid-eye-movement (NREM) sleep on PCA activity. A new electromyographic technique was developed to implant hooked-wire electrodes into the PCA by using a nasopharyngoscope. A previously described technique was used to induce upper airway occlusions during NREM sleep (Kuna and Smickley, J. Appl. Physiol. 64: 347-353, 1988). The PCA exhibited phasic inspiratory activity during quiet breathing in wakefulness and sleep in all subjects. Discounting changes in tonic activity, peak amplitude of PCA inspiratory activity during stage 3-4 NREM sleep decreased to 77% of its value in wakefulness. Tonic activity throughout the respiratory cycle was present in all subjects during wakefulness but was absent during state 3-4 NREM sleep. In this sleep stage, PCA phasic activity abruptly terminated near the end of inspiration. During nasal airway occlusions in NREM sleep, PCA phasic activity did not increase significantly during the first or second occluded effort. The results, in combination with recent findings for vocal cord adductors in awake and sleeping adults, suggest that vocal cord position during quiet breathing in wakefulness is actively controlled by simultaneously acting antagonistic intrinsic laryngeal muscles. In contrast, the return of the vocal cords toward the midline during expiration in stage 3-4 NREM sleep appears to be a passive phenomenon.     

Time of application of negative pressure pulses and upper airway muscle activity

Effect of upper airway pressure changes on thoracic inspiratory muscles has been shown to depend on the time of application during the breathing cycle. The present study was designed to investigate the importance of the time of application of upper airway negative pressure pulses on upper airway muscles. The upper airway was functionally isolated into a closed system in 24 anesthetized spontaneously breathing rabbits. Negative pressure pulses were applied in early (within the first 200 ms) and late (greater than or equal to 200 ms) inspiration, while electromyograms (EMG) of the diaphragm (Dia), genioglossus (GG), alae nasi (AN), and/or posterior cricoarytenoid (PCA) muscles were simultaneously monitored. When negative pressure pulse was applied in early inspiration, the increase in GG activity was greater [0.49 +/- 0.37 to 4.24 +/- 3.71 arbitrary units (AU)] than when negative pressure was applied in late inspiration (0.44 +/- 0.29 to 2.64 +/- 3.05 AU). Similarly, increased activation of AN (2.63 +/- 1.01 to 4.26 +/- 1.69 AU) and PCA (3.46 +/- 1.16 to 6.18 +/- 2.93 AU) was also observed with early inspiratory application of negative pressure pulses; minimal effects were seen in these muscles with late application. An inhibitory effect on respiratory timing consisting of a prolongation in inspiration (TI) and a decrease in peak Dia EMG/TI was observed as previously reported. These results indicate that the time of application of negative pressure during the breathing cycle is an important variable in determining the magnitude of the response of upper airway muscles.(ABSTRACT TRUNCATED AT 250 WORDS)     

Effect of inspiratory muscle strength training on inspiratory motor drive and RREP early peak components.

This study investigated the effect of inspiratory muscle strength training (IMST) on inspiratory motor drive [mouth occlusion pressure at 0.1 s (P(0.1))] and respiratory-related evoked potentials (RREP). It was hypothesized that, if IMST increased inspiratory muscle strength, inspiratory motor drive would decrease. If motor drive were related to the RREP, it was further hypothesized that an IMST-related decrease in drive would change RREP latency and/or amplitude. Twenty-three subjects received IMST at 75% of their maximal inspiratory pressure (Pi(max)) with the use of a pressure threshold valve. IMST consisted of four sets of six breaths daily for 4 wk. P(0.1) and the RREP were recorded before and after IMST. Posttraining, Pi(max) increased significantly by 36.0 +/- 2.7%. P(0.1) decreased significantly by 21.9 +/- 5.2%. The increase in Pi(max) was significantly correlated to the decrease in P(0.1). RREP peaks P(1a), N(f), P(1), and N(1) were identified pre- and post-IMST, and there was no difference in either amplitude or latency for those peaks. These results demonstrate that high-intensity IMST significantly increased Pi(max), decreased P(0.1), but did not change the RREP.     

Influence of upper airway negative pressure reflex on response to airway occlusion in sleeping infants

Artificially produced upper airway suction inhibits the diaphragm in animals and infants; however, the effects of spontaneously generated suction in humans are unknown. We studied nine tracheostomized infants because separation of the upper from the lower airway allowed us to channel suction created by an occluded inspiratory effort to both upper and lower airways (upper + lower airway occlusions) or to the lower airway only (lower airway occlusion). The tracheostomy airway was briefly occluded at end expiration during quiet sleep. In upper + lower airway occlusions, peak airway pressure of the first occluded breath was less negative and rate of pressure decrease slower than that of lower airway occlusions, indicating that upper airway suction inhibits thoracic inspiratory muscles. The threshold for this response was less than or equal to 4 cmH2O suction pressure. The effect on inspiratory time was variable. A decrease in slope of the inspiratory pressure waveform occurring at approximately 0.12 s after inspiration onset was more marked in upper + lower airway occlusions. We conclude that infants have an upper airway reflex response to inspiratory pressure that alters not only the peak and slope but also the shape of the inspiratory pressure waveform.     

Respiratory-related evoked potentials elicited by inspiratory occlusions in double-lung transplant recipients.

This study investigated the role of lung vagal afferents in the respiratory-related evoked potential (RREP) response to inspiratory occlusions by using double-lung transplant recipients as a lung denervation model. Evoked potential recordings in response to inspiratory occlusions were obtained from 10 double-lung transplant (DLT) recipients with normal lung function and 12 healthy control (Nor) subjects under the attend, ignore, and unoccluded conditions. Results demonstrated that early-latency RREP components (P(1), P(1a), N(f), and N(1)) were not significantly different between the DLT and the Nor groups. The late-latency RREP component (P(3)) was identifiable in all DLT subjects during the attend trial. However, P(3) latency was significantly longer in the DLT group compared with the Nor group. The zero-to-peak amplitude of P(3) was also significantly smaller in the DLT group than that in the Nor group during the attend trial. These results suggest that lung vagal afferents were not essential to elicit RREP responses, but may contribute to the cognitive processing of respiratory stimuli.     

Pressure-diameter relationships of the upper airway in awake supine subjects

In awake supine normal subjects, dimensional changes of the oropharyngeal airway were measured during exposure to negative intraluminal pressures. The pressure was generated 1) "actively" by subjects inspiring against an externally occluded airway or 2) "passively" by external suction at the mouth during voluntary glottic closure with no inspiratory effort. Airway dimensions were imaged with X-ray fluoroscopy and anteroposterior diameters measured at levels corresponding to cervical vertebra 3 and 4 (C3 and C4). Cephalad axial displacement of the hyoid bone (CDHY) was also measured. During the "active" maneuver, airway diameters and position were maintained at resting levels despite airway pressure up to -15 cmH2O. In contrast, during the passive maneuver at -15 cmH2O, C3 was only 15 +/- 9% and C4 only 47 +/- 8% of control; CDHY was 5.6 +/- 1.8 mm. In three subjects airway wall apposition occurred and persisted until an active inspiratory effort. We conclude that, in the absence of inspiratory effort, negative oropharyngeal airway pressures result in marked narrowing and cephalad displacement of the upper airway, even during wakefulness. Therefore, our data suggest that the complex interaction of upper airway and thoracic muscle activity is critical in determining the effective compliance and patency of the upper airway, which is readily collapsible even in normal subjects.     

Assessment of upper airway stabilizing forces with the use of phrenic nerve stimulation in conscious humans.

Phrenic nerve stimulation (PNS) applied at end-expiration allows the investigation of passive upper airway (UA) dynamic during wakefulness. Assuming that phasic UA dilating/stabilizing forces should modify the UA properties when twitches are applied during inspiration, we compared the UA dynamic responses to expiratory and inspiratory twitches (2 s and 200 ms after expiratory and inspiratory onset, respectively) in nine men (mean age 28 yr). This procedure was repeated with a 2-cm mouth opening provided with a closed mouthpiece. The percentage of flow-limited (FL) twitches was significantly higher when PNS was realized during expiration than during inspiration. Maximal inspiratory flow (Vi(max)) of FL twitches was significantly higher for inspiratory twitches (1,383 +/- 42 and 1,185 +/- 40 ml/s). With mouth aperture, Vi(max) decreased with an increase in the corresponding pharyngeal resistance values, and the percentage of twitch with a FL regimen increased but only for inspiratory twitches. We conclude that 1) UA dynamics are significantly influenced by the inspiratory/expiratory timing at which PNS is applied, 2) the improvement in UA dynamic properties observed from expiratory to inspiratory PNS characterizes the overall inspiratory stabilizing effects, and 3) mouth aperture alters the stability of UA structures during inspiration.     

Involvement of pharyngeal muscles in compensatory responses of the respiration system to inspiration resistance load]

The present study was undertaken to test the hypothesis that recruitment of upper airway muscles in loaded breathing is a result of integration of peripheral chemoreceptor and pulmonary mechanoreceptor afferents. Experiments were performed in spontaneously breathing tracheostomized anesthetized rabbits. It had been studied the effects of inspiratory resistive loading to EMG activity of genioglossus muscle. In the intact rabbits the peak value and duration of inspiratory activity of genioglossus increased in loading. Imposition of resistive load in vagotomized animals did not evoke alteration in inspiratory activity of genioglossus in the first loaded breath. Hyperoxia decreased the response of genioglossus muscle to inspiratory loading and vagatomy. We conclude that hypoxic stimulation of peripheral chemoreceptors and decrease in volume-related afferent activity from pulmonary stretch receptors are major mechanisms of the upper airway muscle recruiment in inspiratory resistive loading.     

Respiratory changes in nasal muscle length

Respiratory changes in alae nasi muscle length were recorded using sonomicrometry in pentobarbital sodium-anesthetized tracheostomized dogs spontaneously breathing 100% O2. Piezoelectric crystals were inserted via small incisions into the alae nasi of 11 animals, and bipolar fine-wire electrodes were inserted contralaterally in nine of the same animals. The alae nasi shortened during inspiration in all animals. The mean amount of shortening was 1.33 +/- 0.22% of resting length (LR), and the mean velocity of shortening during the first 200 ms was 4.60 +/- 0.69% LR/S. The onset of alae nasi shortening preceded inspiratory flow by 77 +/- 18 ms (P less than 0.002), at which time both alae nasi shortening and the moving average of electromyographic (EMG) activity had reached approximately one-third of their peak values. In contrast, there was a relative delay in alae nasi relaxation relative to the decay of alae nasi EMG at the end of expiration. Single-breath airway occlusions at end expiration changed the normally rounded pattern of alae nasi shortening and moving average EMG to a late-inspiratory peaking pattern; both total shortening and EMG were increased by similar amounts. The onset of vagally mediated volume-related inhibition of alae nasi shortening occurred synchronously with the onset of inhibition of alae nasi EMG; both occurred at lung volumes substantially below tidal volume. These results indicate that the pattern of inspiratory shortening of this nasal dilating muscle is reflected closely in the pattern of EMG activity and that vagal afferents cause substantial inhibition of alae nasi inspiratory shortening.     

Upper airway dilating forces during wakefulness and sleep in dogs

We measured the pressure within an isolated segment of the upper airway in three dogs during wakefulness (W), slow-wave sleep (SWS) and rapid-eye-movement (REM) sleep. Measurements were taken from a segment of the upper airway between the nares and midtrachea while the dog breathed through a tracheostoma. These pressure changes represented the sum of respiratory-related forces generated by all muscles of the upper airway. The mean base-line level of upper airway pressure (Pua) was -0.5 +/- 0.03 cmH2O during W, increased by a mean of 2.1 +/- 0.2 cmH2O during SWS, and was variable during REM sleep. The mean inspiratory-related phasic change in Pua was -1.2 +/- 0.1 cmH2O during wakefulness. During SWS, this phasic change in Pua decreased significantly to a mean of -0.9 +/- 0.1 cmH2O (P less than 0.05). During REM sleep, the phasic activity was extremely variable with periods in which there were no fluctuations in Pua and others with high swings in Pua. These data indicate that in dogs the sum of forces which dilate the upper airway during W decreases during SWS and REM sleep. The consistent coupling between inspiratory drive and upper airway dilatation during wakefulness persists in SWS, but is frequently uncoupled during REM sleep.     

Cerebral cortical respiratory-related evoked potentials elicited by inspiratory occlusion in lambs.

Respiratory-related evoked potentials (RREP) elicited by inspiratory mechanical loads have been recorded in humans. Early RREP peaks were hypothesized to be generated by activation of neurons in the somatosensory cortex. An animal model was developed to test this hypothesis in chronically instrumented, awake, spontaneously breathing lambs. Electrocorticogram (ECoG) was recorded bilaterally with ball electrodes on the dural surface over the somatosensory region. Inspiratory occlusions were presented through a face mask or endotracheal tube as interruptions of inspiration. Occlusion-elicited evoked potentials were obtained by computer-signal averaging the ECoG activity. A short-latency positive peak was observed bilaterally in the averaged occlusion-elicited evoked potentials in all animals breathing with the facemask and 5 of 8 lambs with the endotracheal tube. Postmortem identification of the electrode location demonstrated that the ECoG was recorded in the caudal-lateral portion of the somatosensory cortex. These results demonstrate that inspiratory occlusion elicits an evoked potential in the somatosensory cortical region of awake, spontaneously breathing lambs. The lamb cortical RREP is similar to human RREP.     

Effect of intestinal afferent stimulation on pattern of respiratory muscle activation

The purpose of the present study was to examine the reflex effects of mechanical stimulation of intestinal visceral afferents on the pattern of respiratory muscle activation. In 14 dogs anesthetized with pentobarbital sodium, electromyographic activity of the costal and crural diaphragm, parasternal intercostal, and upper airway respiratory muscles was measured during distension of the small intestine. Rib cage and abdominal motion and tidal volume were also recorded. Distension produced an immediate apnea (11.16 +/- 0.80 s). During the first postapneic breath, costal (43 +/- 7% control) and crural (64 +/- 6% control) activity were reduced (P less than 0.001). In contrast, intercostal (137 +/- 11%) and upper airway muscle activity, including alae nasi (157 +/- 16%), genioglossus (170 +/- 15%), and posterior cricoarytenoid muscles (142 +/- 7%) all increased (P less than 0.005). There was greater outward rib cage motion although the abdomen moved paradoxically inward during inspiration, resulting in a reduction in tidal volume (82 +/- 6% control) (P less than 0.005). Postvagotomy distension produced a similar apnea and subsequent reduction in costal and crural activity. However, enhancement of intercostal and upper airway muscle activation was abolished and there was a greater fall in tidal volume (65 +/- 14%). In conclusion, mechanical stimulation of intestinal afferents affects the various inspiratory muscles differently; nonvagal afferents produce an initial apnea and subsequent depression of diaphragm activity whereas vagal pathways mediate selective enhancement of intercostal and upper airway muscle activation.     

Upper airway muscle activity during nasal occlusion in newborn babies

Submental electromyorgams (SM EMG) were recorded from 20 preterm babies (gestational age 30 +/- 2 wk, postmenstrual age at study 35 +/- 2 wk) (mean +/- SD) and 3 full-term infants (7-14 days old). SM EMG was evaluated during eupnea and brief experimental airway occlusion. Phasic inspiratory SM EMG was rarely seen during eupnea. SM EMG tended to increase on the first occluded effort, although this increase was not statistically significant in most babies. All infants showed progressive breath-by-breath augmentation of phasic SM EMG during occlusions in rapid-eye-movement (REM) as well as quiet (QS) sleep; phasic increases in SM EMG were similar during REM and QS occlusions in the majority (16/22) of babies. Periods of airway closure were detected during 24 occlusions in 5 infants; phasic SM EMG was reduced on these occasions. The results are consistent with the idea that recruitment of upper airway muscles contributes to the stability of the airway of the preterm human.     

Neuromuscular and mechanical responses to inspiratory resistive loading during sleep

The purposes of this study were 1) to characterize the immediate inspiratory muscle and ventilation responses to inspiratory resistive loading during sleep in humans and 2) to determine whether upper airway caliber was compromised in the presence of a resistive load. Ventilation variables, chest wall, and upper airway inspiratory muscle electromyograms (EMG), and upper airway resistance were measured for two breaths immediately preceding and immediately following six applications of an inspiratory resistive load of 15 cmH2O.l-1 X s during wakefulness and stage 2 sleep. During wakefulness, chest wall inspiratory peak EMG activity increased 40 +/- 15% (SE), and inspiratory time increased 20 +/- 5%. Therefore, the rate of rise of chest wall EMG increased 14 +/- 10.9% (NS). Upper airway inspiratory muscle activity changed in an inconsistent fashion with application of the load. Tidal volume decreased 16 +/- 6%, and upper airway resistance increased 141 +/- 23% above pre-load levels. During sleep, there was no significant chest wall or upper airway inspiratory muscle or timing responses to loading. Tidal volume decreased 40 +/- 7% and upper airway resistance increased 188 +/- 52%, changes greater than those observed during wakefulness. We conclude that 1) the immediate inspiratory muscle and timing responses observed during inspiratory resistive loading in wakefulness were absent during sleep, 2) there was inadequate activation of upper airway inspiratory muscle activity to compensate for the increased upper airway inspiratory subatmospheric pressure present during loading, and 3) the alteration in upper airway mechanics during resistive loading was greater during sleep than wakefulness.     

Comparison of volume changes in the upper airway and thorax

To describe the mechanical cycles of the upper and lower portions of the respiratory system, we measured volume change in and out of the isolated upper airway in 13 anesthetized dogs and compared volume changes in the upper airway with tidal volume change during spontaneous respiratory efforts. During inspiration the onset and peak increase in volume into the upper airway preceded the onset and peak of inspiratory tidal volume by 84 +/- 8 and 638 +/- 47 ms, respectively. The volume cycle of the upper airway was nearly complete by the end of inspiratory airflow into the thorax. With progressive hypercapnia there was an increase in the change in both upper airway volume and tidal volume but the temporal sequence was preserved. End-expiratory tracheal occlusion increased the volume change in the isolated upper airway at any level of CO2; however, the effect was disproportionately greater at low rather than at high levels of CO2. Following hyperventilation-induced apnea, a change in volume in the upper airway and thorax occurred on the first inspiratory effort. In most animals at lower levels of CO2, the percent change in upper airway volume with inspiration was relatively less than tidal volume, but the reverse was true at higher levels of CO2. These differences represent dissimilarities in the mechanical forces occurring as the result of upper airway and chest wall muscle contraction during inspiration.