Effect of Lung Inflation on Pulmonary Resistance during NREM Sleep1-3

ROBERT L. BEGLE, SAFWAN BADR, JAMES B. SKATRUD, and JEROME A. DEMPSEY Introduction

During sleep,severalimportant changes in lung mechanics occur that have potentially deleterious effects on ventilation and gas exchange. First, upper airway resistance, which is normally minimal during wakefulness, increases considerably in normal persons and in snorers and patients with obstructive sleep apnea syndrome (OSAS) (1-3). The increase in upper airway resistance contributes to the hypoventilation and CO 2 retention noted in normal subjects during sleep (4). The other major change in lung mechanics is the decrease in FRC. It decreases because of the supine position (5), even while awake, and decreases further during sleep (6). The decrease in FRC probably contributes to sleep-related hypoxia in patients with abnormal lungs (7). In this study, we have investigated whether there is an interaction between resting lung volume and pulmonary resistance during sleep. The cause of the increased upper airway resistance during sleep has largely been attributed to decreased upper airway muscle activation (either absolute or in relation to diaphragm activation), which allows collapse of the supralaryngeal airway (8). However, we wondered if decreased FRC might also be playing a role. For example, it has been shown in awake subjects that upper airway resistance varies inverselywith lung volume (9). Although the effect is small- approximately 1.0em H 2 0 / L / s per liter of lung volume change - it represents about 30010 of total airway resistance. Because upper airway resistance is so small during wakefulness, this change has little clinicalor physiologic significance during eupneic breathing awake. However, given the larger upper airway resistance during sleep, a similar percentage change would be expected to have a significant effect on ventilation. If there is a lung-volume dependence of upper airway resistance present during sleep, what is the mechanism linking the two? In a broad sense, weconsidered that 854

SUMMARY Previous investigations have demonstrated an inverse relationship between lung volume and airway resistance in awake humans. Wewished to examine this relationship in the absence of conscious influences. We therefore studied eight healthy sUbjects who slept in a tank respirator. Hyperinflation was induced by continuous negative tank pressure while the sUbjects breathed spontaneously. Ventilation, pulmonary resistance (total pUlmonary resistance, seven sUbJects; upper airway resistance, one sUbJect),diaphragm and genioglossus electromyograms (EMGs), and sleep state were measured. During control NREM sleep, group mean maximal pulmonary resistance was 42.5 cm H20/Us (range, 17.4to 106.4 cm H20/Us). During steady-state hyperinflation (mean Increase in lung volume = 0.53 L), pulmonary resistance decreased 40% (range, -3 to -90%). Ventilation, sleep state, and end-tidal CO2 were unchanged. Inspiratory muscle EMG was Increased in two of two subjects during hyperinflation. Genioglossus EMGwas characterized by phasic and tonic activity during the control period in two of two SUbJects. Both components were decreased during steadystate hyperinflation. When lung volume was returned to baseline, pulmonary resistance and genioglossus EMG Increased to baseline levels. We conclude that alteration in lung volume within the tidal volume range significantly alters pulmonary resistance during NREM sleep. This Influence occurs independent of chemical stimuli or genloglossal muscle activity, and may be related to traction on neck structures caused by descent of mediastinal structures. AM REV RESPIR DIS 1990; 141:854-860

there weretwo possibilities. One was that there was an increase in neural drive to the upper airway muscles. We have demonstrated that inspiratory muscle EMG increases upon hyperinflation during sleep (10). Perhaps neural output to all inspiratory muscles, including the upper airway muscles (e.g., genioglossus), is tied to lung volume. If true, the fall in FRC during sleep would lead to a decrease in upper airway muscle drive, and thereby contribute to upper airway collapse. On the other hand, changes in lung volume may result in mechanical traction on structures of the upper airway, and thereby change upper airway resistance (11). Weemployed a tank respirator to cause negative pressure around the thorax to induce changes in lung volume. We hypothesized that during NREM sleep, (1) passive elevation of lung volume would decrease upper airway resistance, and (2) passive elevation of lung volume would augment neural output to the genioglossus. Methods Subject Selection Eight healthy, nonobese subjects (four female, four male) 19 to 26 yr of age and weighing 47 to 95 kg (mean 070 ideal body weight, 101;

range, 79 to 121) wereselected for study. None had any history of sleep disturbance, nocturnal apnea, hypersomnolence, or upper airway abnormality. Some reported a history of mild-to-moderate snoring. Regardless of their history, all were observed to snore at some point during the night.

Ventilation Minute ventilation and respiratory cycle timing were measured in the supine position with a tight-fitting face mask (Respironics,Monroeville, PA). The mask wasattached to a two-way breathing valve (dead space, 75 ml) with a heated pneumotachometer in the inspiratory line to measure inspiratory flow and timing parameters. Inspiratory volume was obtained by integrating the inspiratory flow signal and

(Received in original form May 25, 1989 and in revised form October 9, 1989) 1 From the Pulmonary Research Laboratory, Veterans Hospital, and the Departments of Medicine and Preventive Medicine, University of Wisconsin, Madison, Wisconsin. 2 Supported by the Veterans Administration Research Service and by SpecializedCenter of Research Grant No. IPOIHL42243-02 from the National Institutes of Health. 3 Correspondence and requests for reprints should be addressed to Robert L. Begle, M.D., William Beaumont Hospital, 3601WestThirteen Mile Road, Royal Oak, MI 48072-2793.

855

EFFECT OF LUNG INFLATION ON PULMONARY RESISTANCE

calibrating with a water-seal spirometer. Precautions weretaken to assure a leak-free mask seal by monitoring CO 2 around the border of the mask.

Pulmonary Resistance Resistance of the total airways (and tissues) was measured as total pulmonary resistance (RL) in seven subjects, and upper airway resistance (Ruaw) was measured in one subject. RL was measured at peak esophageal pressure measured with a 10 cm latex ballooncatheter swallowed into the lower esophagus. Using the technique of von Neergaard and Wirz (12), lung elastic pressure is subtracted from total intrapleural pressure to obtain the resistive pressure. RL is calculated by dividing resistive pressure by flow. Ruaw was measured at peak pharyngeal pressure (Subject 8). The tip of a transducertipped catheter (Millar Instruments Inc., Houston, TX) was placed in the lower pharynx just above the vocal cords. Proper positioning was confirmed with a fiberoptic laryngoscope. Resistance of the upper airway segment was calculated from the difference between mask pressure and peak pharyngeal pressure (= resistive pressure), divided by inspiratory flow. Because total RLduring sleep predominantly reflects Rua~ (1), resistance data in Subject 8 werecombined with RLmeasurements in the other seven subjects. Change in functional residual capacity (dFRC) was measured from inductance plethysmography (Respitraces; Ambulatory Monitoring, Ardsley,NY) in the nonresetting mode. It was calibrated using the isovolume technique in conjunction with an Ohio 840 rolling seal spirometer (Ohio Instruments, Madison, WI) (13). The ability of the Respitrace to accurately measure dFRC has been previously shown (10). Respiratory Muscle Activity Inspiratory EMG (EMGinsp) was recorded from two bipolar surface electrodes (3M Red Dot; 3M Company, St. Paul, MN) placed 2 to 4 cm above the right costal margin in the anterior axillary line. One pair was positioned at the point of percussed dullness at total lung capacity, and another pair was positioned at the point of percussed dullness at FRC. (Our previous work demonstrated similar changes in activity with either pair during hyperinflation [10].) The electrode pair with the best signal-to-noise ratio was chosen for analysis. In most cases, inspiratory muscle activity was contaminated by expiratory muscle activity. Inspiratory muscle activity was reported only in subjects without expiratory contamination. Genioglossus EMG was recorded from two stainless steel wires implanted perorally into the base of the tongue using 25-gauge hypodermic needles (14). The raw EMG signals were amplified, band-pass-filtered from 5 to 10,000 Hz (Grass model 7D polygraph; Grass Instruments', Quincy, MA), full-waverectified, and integrated with a Paynter filter with a time constant of 100 ms. Peak genio-

glossal phasic activity was measured from a common baseline chosen for each trial. Tonic genioglossal activity was measured as a percentage change from baseline activity. For inspiratory muscle and genioglossus EMG, the slope of the phasic component of the moving time average was calculated (breath-bybreath) by dividing the peak activity by the duration of the increasing activity.

Results

Effect of Hyperinflation on Ventilation Fifty-six hyperinflation trials (range, 1 to 13 per subject) increased lung volume 0.29 ± 0.10 L by the fifth hyperinflated breath, and 0.53 ± 0.10 L during the steady state. (When normalized for body size, the mean acute change in FRC was Snoring 11070 of predicted FRC, and the mean A snorogram was obtained from a micro- steady-state change in FRC was 20% of phone placed in the mask. predicted FRC.) As seen in table 1, tidal volume and minute ventilation were preFiberoptic View A 4-mm laryngoscope and videocamera serveddespite hyperinflation. Inspiratory recorded the upper airway during hyperinfla- time shortened from 1.75 to 1.60 s (p < tion in two subjects during sleep and during 0.05), and expiratory time lengthened from 2.10 to 2.48 s (p < 0.01). Frequency, wakefulness. P0 2 , and Pe02 remained unchanged.

Chemical Stimuli End-tidal CO 2 and O2 were measured either from an accessory port in the mask or from a nasal cannula (Beckman LB-3 and OM-ll; Beckman Instruments, Fullerton, CA).

Sleep Stage Electroencephalogram (EEG), electrooculogram, and EMG were recorded (Grass model 7D) as an index of sleep stage according to the criteria of Rechtschaffen and Kales (15). Protocol Subjects slept supine in a tank respirator (Emerson respirator, Cambridge, MA). Hyperinflation was gradually induced by lowering tank pressure with a vacuum cleaner over a period of 5 to 10 breaths. In two trials in Subject 8, a sudden change was induced over a period of 1 to 2 breaths. After a steady state was achieved (defined by constant ventilatory pattern and PETe02 - usually 2 to 3 min), tank pressure was released and lung deflation occurred. Measurements were made during four phases of each trial as follows: mean of 10 breaths at control FRC immediately preceding hyperinflation (FRC-I), the fifth breath after initiation of acute hyperinflation (HA), mean of 10breaths during steady-state hyperinflation (HS), and mean of 10breaths after deflation (FRC-2). The range of pressures used was 4 to 19cm H 20 (mean ± SEM = 8.1 ± 0.25 em H 20 ). Only data collected during steady-state NREM sleep were used. Baseline awake measurements werealso made with the patients lying supine in the tank respirator. Hyperinflation was not performed during wakefulness because of erratic breathing responses shown by most subjects to this stimulus. Data Analysis A paired t test was used to compare means in the control to steady-state ventilatory data. Because of the broad variability in resistance measurements during the control period (greater than sixfold), Wilcoxon's signed rank test was used to compare resistance measurements during the four phases of the hyperinflation trials (16).

Resistance and NREM Sleep Mean RL during wakefulness was 6.1 em H 20/LIs (range, 3.5 to 17.3 em H 20/LIs) (table 2 and figure 1). During NREM sleep, mean resistance for the group increased to 42.3 cm H 20/L/s (range, 17.4 to 106.4 em H 20/LIs). Resistance and Hyperinflation Pulmonary resistance decreased in all eight subjects during acute hyperinflation (mean decrease, 11.9 cm H 20/L/s; range, 2.2 to 44.3 em H 20/L/s; p < 0.005), and remained below baseline in seven of the eight subjects during steadystate hyperinflation (mean decrease, 17.1 ern H 20/LIs; p < 0.005). When lung volume was returned to its preinflation value, resistance increased toward its baseline NREM level in six of seven subjects. An example of the effect of sudden hyperinflation on upper airway resistance is shown in figure 2. The fall in resistance occurs within the first few breaths upon increase in FRC and remains constant throughout the period of hyperinflation. When lung volume is returned to baseline FRC, resistance increases to preinflation values. Snoring is eliminated during hyperinflation and recurs upon deflation. EEG (not shown) was Stage III and remained unchanged throughout the intervention. Despite the overall fall in RL with hyperinflation noted, there was a lack of proportionality between the magnitude of hyperinflation and the fall in resistance observed. It can be seen in table 2 that a mean hyperinflation of 0.29 L decreased resistance in eight of eight subjects. However, despite a further mean increment of 0.24 L, resistance fell further in only three subjects. It remained the same or increased slightly in four, and returned

856

B£GLE, BADR, SKATRUD, AND DEMPSEY

TABLE 1 MEAN VENTILATOAY CHANGES WITH HYPEAINFLATlON 6FAC (L) Subject No.

VT (L)

TI (s)

TE (s)

VI (Llmin)

f (breaths/min)

Trials (n)

HA

HS

FAC·l

HS

FAC·l

HS

FAC·l

HS

FAC·l

HS

FAC·l

HS

5 1 11 7 10 4 5 13 7 1.5

0.08 0.25 0.49 0.31 0.21 0.33 0.28 0.40 0.29 0.04

0.41 0.66 0.86 0.49 0.42 0.45 0.49 0.49 0.53 0.05

0.39 0.64 0.40 0.52 0.41 0.22 0.74 0.46 0.47 0.05

0.42 0.63 0.39 0.52 0.43 0.25 0.74 0.45 0.48 0.05

2.08 1.91 1.48 2.15 1.34 1.71 1.67 1.63 1.75 0.09

1.70 1.64 1.31 1.99 1.31 1.69 1.53 1.60 1.60' 0.07

1.80 2.53 2.43 2.33 1.84 1.91 2.17 1.79 2.10 0.10

2.70 2.77 2.60 2.76 2.64 2.06 2.31 2.01 2.48t 0.10

15.5 13.5 15.3 13.5 18.9 16.6 15.6 17.5 15.8 0.6

13.6 13.6 15.3 12.6 15.2 16.0 15.6 16.6 14.8 0.5

6.0 8.6 5.5 7.0 7.8 3.6 11.6 8.0 7.3 0.8

5.7 8.6 5.5 6.6 6.5 3.9 11.7 7.5 7.0 0.8

1 2 3 4 5 6 7 8 Mean SEM

PO,

Pco,

(mmHg)

(mmHg)

FAC·l

HS

FRC·l

HS

96.7 81.1 95.9 97.5

97.5 103.5 92.6 101.0

43.1

41.9

110.8 98.4 91.1 95.9 2.9

104.7 103.4 89.4 98.9 1.9

42.7 41.8 38.9 45.1 45.0 42.8 0.7

41.4 40.8 40.0 45.4 45.3 42.4 0.7

Definition of abbreviations: 6.FRC = change in end-expiratory lung volume during hyperinflation; FRC-l = baseline lung volume; HA = fillh breath of acute hyperinflation; HS hyperinflation; VT = tidal volume; TI = inspiratory duration; TE = expiratory duration; f = breathing frequency; VI = inspiratory minute ventilation. • p < 0.05 comparedwith value durin9 FRC-l. t p < 0.01 compared with value during FRC-l.

to baseline in one. In figure 3, for example, the maximal fall in total RL occurs by Breath 8, at which point only about one-fifth of the ultimate increase in FRC has been attained. To further investigate the relationship between the magnitude of hyperinflation and degree of resistance change, and to eliminate transient influences during acute hyperinflation, regression analysis was performed between the magnitude of steady-statehyperinflation and steadystate fall in resistance. Regression lines were calculated for each subject's trials as wellas for all of the trials for the group as a whole. No significant relationship was found.

Inspiratory Muscle and Genioglossus EMGs The effect of hyperinflation on inspiratory muscle EMGs and resistance in two subjects is summarized in table 3.

The slope of the inspiratory muscle EMG remained unchanged, whereas resistance decreased during acute hyperinflation. During steady-state hyperinflation, inspiratory muscle activity increased 11 to 14070, whereas resistance showed little further change. Peak inspiratory muscle EMG showed identical changes. Genioglossus EMGs were recorded in two subjects during hyperinflation during NREM sleep (table 4). Both phasic and tonic activity were present during control. In both subjects during steadystate hyperinflation, tonic activity fell 52 and 10070, and phasic slope fell 42 and 22070. Peak activity fell 48 and 32070. Inspiratory time (table 1) fell I to 8070. Upon return to control lung volume, phasic and tonic activity increased (figure 4).

Site of Resistance Change with Hyperinflation In two subjects the pharynx and larynx

= steady-state

wereobserved during hyperinflation with a fiberoptic laryngoscope. During wakefulness, the retropalatal airway was widely patent in both subjects. In Subject 3, the airway remained large during NREM sleep, and did not change appreciably with hyperinflation. This was the subject in whom hyperinflation had no discernible effect on resistance during steady-state hyperinflation. In Subject I (figure 5) the retropalatal airway narrowed during NREM sleep. It increased during steady-state hyperinflation and then decreased again upon return to control FRC. Discussion

We have demonstrated that elevation of resting lung volume during NREM sleep decreases RL and snoring in normal snorers. We believe that the site of the decrease in resistance is the upper airway for three reasons. First, we actually mea-

TABLE 2

"T

MEAN CHANGES IN AIRWAY MECHANICS DURING HYPERINFLATION WHILE IN NREM SLEEP

0 0

100

Resistive Pressure (cm H2O)

Aesistance (cm H2O/LIs)'

SUbject Trials No. (n) Awake FAC·l 1 5 1 2 11 3 4 7 5 10 4 6 7 5 13 8 Group mean SEM

5.0 5.0 4.9 3.5 3.5 17.3 5.0 4.3 6.1 1.5

106.4 65.1 18.9 22.2 17.4 54.9 31.4 21.9 42.3t 11.1

HA 62.1 46.7 16.7 18.0 12.9 44.4 26.1 16.2 30.4t 6.5

HS

FAC-2 FRC·l

11.1 101.3 54.1 18.3 18.3 16.0 20.7 14.0 15.9 49.8 56.0 29.2 21.9 22.1 16.7 25.:11":1: 37.6 5.9 11.0

21.5 22.2 8.6 7.9 8.1 8.5 12.4 7.7 12.1 2.2

HA

HS

Flow

(Lis)

FAC-2 FRC·l

17.1 6.7 20.1 20.1 17.9 8.9 8.3 8.4 6.9 7.2 7.9 5.5 7.5 7.0 8.6 9.0 8.2 11.9 10.9 9.9 6.1 7.8 6.8 10.8t 8.81:1: 10.3 1.4 1.8 1.6

0.21 0.34 0.46 0.36 0.47 0.16 0.40 0.35 0.34 0.04

HA

HS

0.28 0.60 0.43 0.33 0.52 0.49 0.45 0.43 0.66 0.40 0.19 0.18 0.42 0.45 0.43 0.37 0.42tO.41 0.05 0.04

• Total pUlmonary resistancefor Subjects 1 to 7; upper airway resistance for Subject 8. comparedwith value to the lell. :j: p < 0.005 comparedwith value at FAC-l.

0

::::::: 60

•

ON

FRC·2 0.21 0.45 0.35 0.47 0.16 0.41 0.35 0.34 0.04

Definition of abbreviations: FRC-2 = return to baseline lung volume aller hyperinflation. For other definitions, see table 1.

t p < 0.005

0

~

I

50

13 ~40

~

a

j30 20 10 0

• 0

0

•

A"",," Baseline

i

NREMSleep Baseline

•

.

\)

0

•

•

• ~

• ~0

Hyperinflation Hyperinflation S1h8reoth S1eody-State

NREM S_p Baseline

Fig. 1, The influence of lung volume and sleep state on pulmonary resistance (Resistance = total pulmonary resistance in Subjects 1 to 7 and upper airway resistance in Subject 8). Each symbol represents the mean of all trials within a subject at the indicated condition. Dash indicates group mean.

857

EFFECT OF WNG INFLATION ON PULMONARY RESISTANCE

q. rt~

sured Ruaw in one subject (Subject 8), and the changes noted were similar to those in the other subjects. Second, given the relatively small contribution of the lower airways to total pulmonary airway resistance during sleep (1), the magnitude of decrease in total airway resistance we observed with hyperinflation must have occurred within the upper airway. Third, we endoscopically documented the increase in upper airway dimensions in one subject whose steady-state resistance fell (Subject I), and the absence of change in airway dimension in the one patient whose airway resistance did not fall (Subject 3). These findings extend those made during previous investigations in awake subjects. Measurements of airway resistance in upright subjects who voluntarilychange lung volume show small (approximately 1 em H 20/L/s per liter change in lung volume) inverse changes in upper and total airway resistance (9, 17-20). Similarly, pharyngeal cross-sectional area (measured by acoustic reflectance) decreases 23 to 36070 in upright normal subjects when they exhale from TLC to residual volume (21). An even larger decrease (54%) is noted in obese patients with OSAS when they exhale to residual volume (22). The studies listed above relied on voluntary changes in lung volume. Changing posture from upright to supine decreases lung volume passively, and a few studies have also looked at the effect of this passive change in lung volume on upper airway patency. Anch and coworkers (3) measured the change in supraglottic resistance accompanying a change from upright to supine. They showed a 25% increase in normal subjects and a 42% increase in obese patients with OSAS. Changes in pharyngeal cross-sectional area are uncertain. Fouke and Strohl (23) found a 17% decline, whereas Brown and coworkers (24) found no significant change.

O[

tr;~[ ;~

0

~

0

s i~

:s

1 LL

5

-5 -10

05 0

-.j 05rw~WlWNN'JINVWV OUI ~20

~s

~£

g

10

0 ~Baseline

FRC-+-----

Fig. 2. Effect of a sudden decrease in tank pressure (PT) on end-tidal Pe02 (PETC02) , pharyngeal pressure (Pph), flow, end-expiratory level and tidal volume (VT), and upper airway resistance (Ruaw). This is followed by a sudden return of tank pressure to the control level.

'O!lO

50

,_FAt NREM_

Fig. 3. Effect of sleep state and gradlJal hyperinflation on esophageal pressure (Peso), flow, end-expiratory level and tidal volume (VT), snoring (Phono), and total pulmonary resistance (RTP).

TABLE 3 EFFECT OF HYPERINFLATION ON INSPIRATORY EMG* Slope (units) Subject No.

6 7

Resistance

~FRC

(cm H2 0 /LIs)

(L)

Trials (n)

FRC-l

HA

HS

FRC-l

HA

HS

HA

HS

4 5

3.6 ± 0.2 3.5 ± 0.1

3.7 ± 0.4 3.4 ± 0.3

4.0 ± 0.2 4.0 ± 0.2

55.6 ± 0.6 31.5 ± 3.0

47.7 ± 1.7 24.5 ± 2.1

49.8 ± 2.1 22.1 ± 0.4

0.33 ± 0.07 0.28 ± 0.06

0.45 ± 0.07 0.49 ± 0.04

For definition of abbreviations, see table 1. • Values are mean ± SEM.

858

BEGlE, BADR, SKATRUD, AND DEMPSEY

TABLE 4 EFFECT OF HYPERINFLATION ON GENIOGLOSSUS EMG'

Subject No. 1

8

Trials (n) 3 6

Tonic Activity (Ola of FRC-1) FRC-1 100 100

HA 77 ± 8 92 ± 2

HS 25 ± 3 82 ± 1

Phasic Slope

Phasic Peak

~FRC

(units)

(units)

(L)

FRC-1

HA

HS

FRC-1

HA

HS

HA

HS

2.4 ± 0.2 14.7 ± 1.6

2.7 ± 0.2 12.4 ± 2.4

1.4 ± 0.3 11.5 ± 0.8

17.0 ± 2.3 93.8 ± 1.4

13.7 ± 1.7 80.0 ± 1.00

8.8 ± 1.7 63.6 ± 0.8

0.06 ± 0.02 0.16 ± 0.04

0.25 ± 0.10 0.36 ± 0.03

For definition of abbreviations, see table 1. • Values are mean ± SEM.

Although changing position passively changes lung volume, other effects (e.g., gravitational effects on mobile upper airway structures, subtle changes in headneck alignment) occur that may affect upper airway resistance. Two recent reports have examined the effect of changing lung volume while in the supine position. Both werein awake subjects. Using a chest cuirass, Fouke and Strohl (23) found no change in pharyngeal cross-sectional area when FRC was increased approximately 450ml in normal supine subjects. Series and colleagues (25), on the other hand, have reported a 25070 reduction in upper airway resistance with hyperinflation in a tank respirator using 5 to to em H 2 0 . Because both of these studies wereperformed in awake subjects, variability of results is not surprising given the foreign sensations engendered by these techniques. For example, we noted erratic changes in breathing patterns and end-expiratory lung volume when our subjects were exposed to constant pressure changes during wakefulness, Although we could have coached them to adopt a uniform breathing pattern, we felt that this would just impose one more artificial influence on their response to

hyperinflation. The results of our study are unique in that volume-dependence of upper airway resistance has been demonstrated with passive changes in the lung volume, and that it occurs during sleep.

Limitations First, we changed lung volume using only one technique. Is it possible that the fall in RL that we measured was due to an artifact of this technique? Our major concern was that a change in the degree of head flexion might result from hyperinflation in the tank respirator, and that this would change Ruaw (19, 26, 27). However, we observed no change in head flexion during the hyperinflations. In order to eliminate the possibility of other artifacts related to the tank respirator, we attempted to change FRC by tilting subjects on a rocking bed. Unfortunately, the subjects uniformly failed to sleep. Second, genioglossusEMGs weremeasured in only two subjects. Although both subjects showed the same response, these data werenot the main focus of this investigation. Additional trials would be helpful to confirm these results and to more fully characterize the response (e.g., time course, etc.) to hyperinflation.

:-+------~Increased

FRC:--------1I-Boseline FRC-i

Fig. 4. The influence of hyperinflation on genioglossus EMG(EMGgg), end-expiratorylevel and tidal volume (VT), and end-tidal Peo, (PETeo.).

Implications It is not surprising, given the complexity of the intervention, that there was no correlation between the degree of hyperinflation and the amount of resistancechange. First of all, there was a narrow range of ilFRC in most subjects, and only a small number of trials in the two subjects with the broadest range. Second, because we did not measure absolute lung volume, the starting point for each trial may have varied as the subject's FRC spontaneously varied. Third, minimal hyperinflation may provide just enough dilating force to overcome surface forces that render the pharyngeal mucosa "sticky" (28), whereas further hyperinflation has little effect on airway diameter. In an earlier work, we demonstrated that hyperinflation during sleep augmented inspiratory muscle activity (to). Although the mechanism underlying this response is unknown, Golgi tendon organ activation during muscle stretch has been suggested (29).Although genioglossal muscle length is not changed with hyperinflation, we had speculated that genioglossal activity might also increase with hyperinflation as part of a general augmentation of drive to inspiratory muscles. Instead, genioglossal activity decreased (table 4 and figure 4). A decrease in Pe0 2 cannot be implicated. Many animal species show a volumerelated inhibition of genioglossal and other upper airway muscles (30). Yetthis inhibition in animals is vagally mediated, and vagal influences appear to be less active in adult humans (31). Interestingly, another study has shown inhibition of genioglossal activity in patients with application of nasal continuous positive airway pressure (CPAP) (32). Although nasal CPAP causes hyperinflation in these patients, stimulation of inhibitory airway receptors or hypocapnia may have mediated the effect. Finally, genioglossus EMG may have fallen because of the reduction in load (33). As noted above, we demonstrated a

859

EFFECT OF LUNG INFLATION ON PULMONARY RESISTANCE

CONTROL

Fig. 5. View through a fiberoptic bronchoscope from just above the posterior edge of the soft palate. Left, control lung volume; center, steady-statehyperinflation; right, return to control lung volume.

HYPERINFLATION

CONTROL

Anterior

Posterior

preservation of tidal volume along with augmentation of inspiratory muscle EMG during hyperinflation during NREM sleep (10). During hyperinflation, the diaphragm shortens and becomes a less effective pressure generator (34). Because tidal volume did not fall, ,we attributed its preservation exclusively to the augmented drive to the diaphragm. However, the fall in RLthat occurs with the hyperinflation that we have demonstrated in the current study indicates that this load reduction must also be taken into account when explaining the preservation of tidal volume. The importance of increased Ruaw on ventilation during NREM sleep was recently demonstrated in another group of normal subjects (4). Reduction of resistance by 38070 during helium-oxygen breathing led to a 7% increase in tidal volume despite a 20% fall in inspiratory muscle EMG. Why did RL fall in the presence of decreased upper airway dilator muscle activity? One possibility is that the shortened diaphragm contracted with less power, and therefore less "suction" pressure was exerted on the upper airway (34, 35). If this were the case, mean inspiratory flow (VT/TI) would have decreased or remained unchanged; instead, it increased. The second possibility is that other upper airway dilator muscles were activated, whereas the genioglossus was inhibited. However, because activity of the genioglossus parallels that of other upper airway dilator muscles, this seems unlikely (30, 36). A third possibility is that hyperinflation led to relatively greater inhibition of supraglottic constrictor muscles than of dilator muscles (37). Because we did notmeasure EMG from this muscle, we can only speculate that this

mechanism may have played a role in lowering resistance. However, we can say that the other major site of airflow limitation caused by constrictor activity, the laryngeal aperture, showed no change during hyperinflation in the two subjects examined endoscopically. We feel the most likely explanation for the fall in RLduring hyperinflation is that the descent of the diaphragm and increased negative pleural pressure at endexpiration resulted in caudal motion of upper airway structures as demonstrated by van de Graff (11). In his experiments with anesthetized dogs, he showed that Ruaw decreased with caudal motion of the neck structures, and that this decrease was independent of upper airway muscle innervation. Reduced pharyngeal wall compliance, increased lumen diameter, and traction on the hyoid bone are some of the possible mechanisms by which caudal traction might reduce Ruaw. The clinical significance of the volumedependence of RL probably depends on the baseline condition of the upper airway and the extent of the fall in FRC that normally occurs during sleep. We did not measure the fall in FRC in our subjects during sleep, but if we assume a fall similar to what Hudgel and Devadatta (6) found, it was probably around 200 to 300 ml. With hyperinflation, we increased FRC to awake or slightly above awake levels, and RL decreased toward normal in seven of the eight subjects, substantially so in Subject 1. Even if one excludes Subject 1, 22% of the increase in resistance that occurred during NREM sleep was reversed when FRC was increased to awake levels. In addition, snoring was obliterated in all subjects except Subject 3. In patients with OSAS, the significance

of this relationship might be even greater. First of all, patients with OSAS have smaller pharyngeal cross-sectional areas at FRC than do normal subjects (independent of body weight) (22, 38). Second, those who are obese (constituting 50% of patients with OSAS) have a greater volume-dependency of cross-sectional area than do both obese patients without OSAS and normal subjects (21, 22, 24). As noted above, the FRC of normal subjects decreases during sleep (6). In obese patients with OSAS, FRC has also been noted to vacillate along with other respiratory events during periods of repetitive apneas (39). FRC is at its minimum at the end of the apneic phase. It increases to its maximal value during the first few breaths ofthe ventilatory phase, and then decreases progressively until the next ventilatory phase begins. Although the primary factor relieving the obstruction is undoubtedly the increased drive to the genioglossus (and presumably other dilator muscles) that accompanies the arousal at the end of the apnea, the increase in FRC undoubtedly contributes to the fall in resistance. As seen in Subject 1, a considerable fall in resistance may occur simply as a result of a change in FRe. Finally, a preliminary study has reported a 12% reduction in apneas and hypopneas in three patients with OSAS during hyperinflation (500 ml) in a poncho-type respirator (40). In summary, we have demonstrated that modest hyperinflation results in a 17to 90% reduction in RLand abolishes snoring during NREM sleep in normal subjects. Ventilatory compensation for hyperinflation occurs because of reduction in RL and augmentation of inspiratory muscle activity. The most likely explanation for the decrease in resistance is traction on the upper airway structures exerted by descent of the diaphragm and mediastinal structures. Reduction in upper airway constrictor muscle activity may also playa role. Elevation of lung volume has variable effects on inspiratory muscle activity. Inspiratory muscle activity is augmented, whereas genioglossal activity is inhibited. Reflex mechanisms are likely responsible for these changes since they occurred in the absence of changes in chemical stimuli. The efficacy of nasal CPAP in OSAS is partially due to the hyperinflation that accompanies this mode of therapy. References 1. Hudge) ow, Martin RJ, Johnson B, Hill P. Mechanics of the respiratory system and breathing

860 pattern during sleep in normal humans. J Appl Physiol 1984; 56:133-7. 2. Warner GJ, Skatrud JB, Dempsey JA. Effect of hypoxia-induced periodic breathing on upper airway obstruction during sleep. J Appl Physiol 1987; 62:2201-11. 3. Anch AM, Remmers JE, Bunce H. Supraglottic airway resistance in normal subjects and patients with occlusive sleep apnea. J Appl Physiol 1982; 53:1158-63. 4. Skatrud JB, Dempsey JA, Badr S, Begle RL. Effect of airway impedance on CO 2 retention and respiratory muscle activity during NREM sleep. J Appl Physiol 1988; 65:1676-85. 5. Agostoni E, Hyatt RE. Static behavior of the respiratory system. In: Macklem PT, Mead J, eds. Handbook of physiology. Section 3: The respiratory system. Bethesda: Williams & Wilkins, 1986; 120. 6. Hudgel DW, Devadatta P. Decrease in functional residual capacity during sleep in normal humans. J Appl Physiol 1984; 57:1319-22. 7. Hudgel DW, Martin RJ, Capehart M, Johnson B, Hill P. Contribution of hypoventilation to sleep oxygen desaturation in chronic obstructive pulmonary disease. J Appl Physiol1983; 55:669-77. 8. Harper RM, Sauerland EK. The role of the tongue in sleepapnea. In: Guilleminault C, ed. Sleep apnea syndromes. New York: Alan R. Liss, 1978; 219-34. 9. Hyatt RE, Wilcox RE. Extrathoracic airway resistance in man. J Appl Physiol 1961; 16:326-30. 10. Begle RL, Skatrud JB, Dempsey JA. Ventilatory compensation for changes in functional residual capacity during sleep. J Appl Physiol 1987; 62:1299-1306. 11. van de Graff WB. Thoracic influence on upper airway patency.J Appl Physiol1988; 65:2124-31. 12. von Neergaard J, Wirz K. Die messung der stromungswiderstande in den atemwegen des menschen, insbesondere bei asthma und emphysem. Z KlinMed 1927; 105:51-82. 13. Konno K, Mead J. Measurement ofthe separate volume changes of rib cage and abdomen during breathing. J Appl Physiol 1967; 22:407-22. 14. Sauerland EK, Mitchell SP. Electromyographic activity of intrinsic and extrinsic muscles of the human tongue. TexRep BioI Med 1975;33:445-55.

BEGlE, BADR, SKATRUD, AND DEMPSEY

15. Rechtschaffen A, Kales A. A manual of standardized terminology, techniques, and scoring system for sleep stages of human subjects. Bethesda: National Institutes of Health, 1968. (NIH publication no. 204). 16. Colton T. Statistics in medicine. Boston: Little, Brown and Co., 1974; 219-21. 17. Briscoe WA, DuBois AB. The relationship between airway resistance, airway conductance and lung volume in subjects of different age and body size. J Clin Invest 1958; 37:1279-85. 18. Brunes L, Holmgren A. Total airway resistance and its relationship to body size and lung volumes in healthy young women. Scand J Clin Lab Invest 1966; 18:316-24. 19. Spann RW, Hyatt RE. Factors affecting upper airway resistance in conscious man. J Appl Physiol 1971; 31:708-12. 20. Skoogh BE. Normal airways conductance at different lung volumes. Scand J Clin Lab Invest 1973; 31:429-41. 21. Brown to, Zamel N, Hoffstein V. Pharyngeal cross-sectional area in normal men and women. J Appl Physiol 1986; 61:890-5. 22. Hoffstein V, Zamel N, Phillipson EA. Lung volume dependence of pharyngeal cross-sectional area in patients with obstructive sleep apnea. Am Rev Respir Dis 1984; 130:472-5. 23. Fouke J, Strohl K. Effect of position and lung volume on upper airway geometry. J Appl Physiol 1987; 63:375-80. 24. Brown IG, McClean R, Boucher R, Zamel N, Hoffstein V. Changes in pharyngeal cross-sectional area with posture and application of continuous positive airway pressure in patients with obstructive sleep apnea. Am Rev Respir Dis 1987; 136: 628-32. 25. Series F, Desmeules M, Couture J, Cormier Y. Effects of pulmonary inflation, expiratory positive airway pressure (EPAP) and CPAP with and without pulmonary inflation on upper airways resistance in normal man (abstract). Am Rev Respir Dis 1989; 139:A79. 26. Greene DG, Elam JO, Dobkin AB, Studley CL. Cinefluorographic study of hyperextension of the neck and upper airway potency. JAMA 1961; 176:570-3. 27. Lhstro G, Stanescu D, Dooms G, Rodenstein

D, Veriter C. Head position modifies upper airway resistance in men. J Appl Physiol1988; 64:1285-8. 28. Miki H, Hida W, Kikuchi Y, et al. Effect of hypoglossal stimulation and pharyngeal lubrication on the opening of obstructed upper airway (abstract). Am Rev Respir Dis 1988; 137:AI25. 29. Green M, Mead J, Sears TA. Muscle activity during chest wall restriction and positive pressure breathing in man. J Appl Physiol1978; 35:283-300. 30. Van Lunteren E, Strohl KP. The muscles of the upper airways. In: Widdecombe JH, ed. Clinics in Chest Medicine. Philadelphia: W.B. Saunders, 1986; 174-9. 31. Polacheck J, Strong R, Arens J, DaviesC, Metcalf I, Younes M. Phasic vagal influence on inspiratory motor output in anesthetized human subjects. J Appl Physiol 1980; 49:609-19. 32. Strohl KP, Redline S. Nasal CPAP therapy, upper airway muscle activation, and obstructive sleep apnea. Am Rev Respir Dis 1986; 134:555-8. 33. Patrick GB, Strohl KP, Rubin SB, Altose MD. Upper airway and diaphragm muscle responses to chemical stimulation and loading. J Appl Physiol 1982; 53:1133-7. 34. Braun NMT, Arora NS, Rochester DF. Forcelength relationship of the normal human diaphragm. J Appl Physiol 1982; 53:405-12. 35. Brouillette lIT, Thach BT. Control of genioglossus muscle inspiratory activity. J Appl Physiol 1980; 49:801-8. 36. Mathew OP. Upper airway negative pressure effects on the respiratory activity of upper airway muscles. J Appl Physiol 1984; 56:500-5. 37. Sauerland EK, Orr WC, Hairston LE. EMG patterns of oropharyngeal muscles during respiration in wakefulness and sleep. Electromyogr Clin Neurophysiol 1981; 21:307-16. 38. Bradley TD, Brown IG, Grossman RF, et al. Pharyngeal size in snorers, nonsnorers, and patients with obstructive sleep apnea. N Engl J Med 1986; 315:1327-31. 39. Aronson RM, Alex CG, Onal E, Lopata M. Changes in end-expiratory lung volume during sleep in patients with occlusive apnea. J Appl Physiol 1987; 63:1642-7. 40. Series F, Cormier Y, Lampron N, LaForge J. Influence of lung volume on sleep apnea (abstract). Am Rev Respir Dis 1987; 135:A134.