INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

There is a well known hyperbolic relationship between airway resistance and lung volume (1); as lung volume increases airway resistance falls. In asthma it is known that lung volumes rise because of gas trapping or persistent activity of the inspiratory muscle throughout expiration (4). This rise in lung volume is an important adaptive response that serves to ameliorate the increase in airway resistance and defends airway patency. The mechanism by which lung volume influences airway resistance is generally believed to be the tethering effect of the parenchyma attachments to the walls of the airways, a function that is termed interdependence (5).

Previously, Ballard and colleagues (6) have shown that lower pulmonary resistance (Rlp) progressively increases during the night in patients with nocturnal asthma. Moreover, most, but not all, of this increase in Rlp was shown to be due to the sleep state. In a subsequent study (7), we showed that thoracic gas volume (Vtg) fell during sleep and that arousal was associated with a rise in Vtg. At that point we had concluded that a fall in lung volume accounted for a portion of the increase in Rlp, that is, the rise in lung resistance observed in patients with nocturnal asthma was due to both sleep and an associated fall in lung volume. However, we also observed that upon awakening, lower airway resistance rose to a value that was higher than the lower airway resistance before sleep onset (6). That resistance was now at a higher level but at the same lung volume suggested to us that airway resistance was no longer mechanically coupled to lung volume.

The objective of the current study was to assess the contributions of the following factors to apparent loss of the volume-resistance relationship in patients with nocturnal asthma: (1) asthma per se, (2) posture, and (3) time. In particular, we hypothesized that there might be an uncoupling of airway parenchymal interdependence at later time points as the lung becomes inflamed. To accomplish this goal, we studied the ability of negative extrathoracic pressure, as delivered through a thoracic cuirass, to change Rlp during awake and sleep states. In this manner we were able to evaluate the degree of interdependence of the parenchyma and airways as assessed with volume-resistance relationships.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Subjects

Five male subjects 27 to 48 yr of age were recruited to participate in the current study. All subjects had well-documented histories of asthma that met the criteria of the American Thoracic Society (8). In addition, all patients had documented decrements of  15% in peak expiratory flow rate on at least three of four nights. All subjects had symptomatic control of their asthma using inhaled beta-adrenergic agonist and were receiving therapeutic twice-daily doses of a theophylline preparation. These protocols were reviewed by an Institutional Clinical Research Committee as well as by the Institutional Review Board, and a formal consent was obtained from all subjects before participation.

Lung Volume

Vtg was measured in both the upright (9) and the horizontal posture (7) with whole-body plethysmographs where volume is directly measured from an attached Krogh-type spirometer. Temperature within the plethysmograph was maintained by recirculating the air in the plethysmograph over an air-to-air heat exchanger (Midland Ross, Cambridge, MA). Vtg was determined by using Boyle's Law with the esophageal balloon (10) as previously described (7). Briefly, all subjects had a 10-cm latex balloon inserted into the lower esophagus and positioned using the criteria of Baydur and colleagues (11) to estimate pleural pressure (Pes). Pes and plethysmograph volume transducers had similar amplitudes and were phase-matched at frequencies as great as 15 Hz. Each subject wore a tightly sealed face mask (Spiderwort Design, Boulder, CO), which covered the nose and the mouth and was attached to a one-way valve. An inflatable latex balloon inside the inspiratory port of the one-way valve was inflated to periodically occlude the airway. The relationship between Pes and plethysmograph volume during each single occluded inspiratory effort was recorded using a storage oscilloscope (R5111A; Tektronix, Englewood, CO) in order to determine Vtg. To validate the system, each subject, while awake, had five measurements of Vtg made within the horizontal plethysmograph. This was performed while the subject was wearing the cuirass system. These values were compared with three to four supine measurements of FRC made with a closed system He dilution method (7).

Lower Pulmonary Resistance

Lower pulmonary resistance (Rlp) was measured as described previously (6, 7). Briefly, supraglottic pressure (Psg) was measured with a microtip-disposable catheter (MPC-500; Millar Instruments, Galveston, TX) positioned directly above the glottis. Esophageal pressure was electronically subtracted from Psg to obtain transpulmonary pressure. Both the esophageal balloon and the microtip catheter were initially calibrated in an airtight pressure chamber using a water manometer and referenced to ambient air pressure. Airflow was obtained with a pneumotachograph (No. 1; Fleisch, Lausanne, Switzerland), which was attached to the exterior of the plethysmograph (Figure 1).

View larger version (33K): [in this window] [in a new window]   Figure 1.   Schematic of horizontal volume-displacement body plethysmograph. Lower pulmonary resistance (Rlp) was derived from the pressure drop from the subglottic millar catheter (Psg) to the esophageal balloon (Pes) and airflow . Lung volume was increased by applying continuous negative pressure (CNP) to the thorax of the subject with a poncho cuirass. Rla = (Psg  Pes)/; TGV = K · Patm · (Vol · Pes).

Protocols

Sleep study. All subjects were studied overnight in the sleep laboratory on two consecutive nights (Figure 2). Each subject was prepared for "lights out" and had fallen asleep by about midnight. The first night served as an acclimatization night where the subject slept in the box with mask, ear oximeter, and cuirass attached. Arterial O₂ saturation was monitored using an ear oximeter (Biox III; Ohmeda, Boulder, CO). On the study night the subject was additionally instrumented with the esophageal balloons, Millar catheter, and sleep staging electrodes. Sleep was staged per standard criteria. All monitored parameters were recorded using a 15-channel chart recorder (Model 78D; Grass Instruments, Quincy, MA). Before "lights out," Rlp were measured at baseline, and at 0.5 and 1.0 L above baseline Vtg. Next, a medication (Halcion, 0.5 mg) was given just prior to "lights out" in order to facilitate sleep. Once stable Stage 2 sleep had been established ("early sleep"), three to five single inspiratory occlusions were performed to determine Vtg. Rlp was recorded during the corresponding time frame (noted by TGV, RL on time line in Figure 2). Continuous negative transthoracic pressure (CNP) was then applied for 10 min and the Vtg was brought back and maintained at baseline levels with CNP while the subject was still asleep. After 10 min of negative pressure, Vtg and Rlp were again measured. Negative pressure was then turned off and the subject continued sleeping for an additional 3.5 h before repeating the same measurements made in early sleep. These measurements are designated as "late sleep." If for any reason, the subject awoke during the measurements, the process was stopped and the subject was allowed to reestablish stable sleep before measuring Vtg and Rlp.

View larger version (17K): [in this window] [in a new window]   Figure 2.   Sequence of the experimental protocols for the sleep study and awake study.

Awake study. On the basis of the results of the sleep phase of the study a second phase was conducted. The same five subjects reported to laboratory on a second occasion (Figure 2). An esophageal balloon was placed and the subject was placed in the cuirass and then studied in an upright plethysmograph of similar design. Five baseline measurements of Vtg and then lung resistance (RL) were determined. The negative pressure source (CNP) was then activated and lung volume elevated first to ~ 0.5 L and then ~ 1.0 L above baseline FRC. The subject was then moved to the horizontal body plethysmograph where the microtip catheter was inserted to measure Psg and Rlp. Measurements of Vtg and Rlp followed by changes in lung volume via CNP were repeated in the same sequence as above.

Data Analysis

Pulmonary resistance was calculated on a breath-by-breath basis, as determined from transpulmonary pressure (the difference between airway opening pressure at the mouthpiece and the esophageal balloon) for the upright awake studies or lower airway transpulmonary pressure (as the difference between the Psg and the esophageal balloon). Rlp was calculated on a breath-by-breath basis, utilizing simultaneous Psg, Pes volume, and flow data by a pulmonary mechanics computer using the Neergard-Wirtz technique (6, 7, 12). Resistance of the apparatus was determined (1.25 cm H₂O/L/s at 0.2 L/s) and subtracted from all measurements of Rlp. Resistance was measured in all situations at the same inspiratory flow rate of 0.250 L/s. Data represent the mean of the resistance from at least 10 breaths.

Data were analyzed for statistical significance by first obtaining a least-squares regression line that was fit for each subject to determine the volume-resistance relationships for each condition. An analysis of covariance was used to test for differences in slope. As the resistance data appeared to be positively skewed, log transformation was performed to obtain a normal distribution. Differences in resistance, volume, or compliance were assessed with Fisher's protected least significance differences. Alpha was set at 0.10 as a one-tailed test was used because of a priori predictions based on indirect evidence from our previous studies (6, 7) for the changes in resistance and volume.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Anthropometric and pulmonary function data for each of the study subjects are found in Table 1. The FEV₁ was 70.4 ± 6.8% predicted, and the PC₂₀ for these subjects was 0.09 ± 0.03 mg/ ml. Vtg averaged 3.31 ± 0.19 L, which was 89.2 ± 7.7% predicted. Vtg fell 0.40 L when the supine posture was assumed (upright: 3.31 ± 0.19 L to supine: 2.91 ± 0.19 L) (p < 0.05). When the FEV₁ was remeasured the morning after the second night study, it was 1.71 ± 0.39 L (p < 0.005), which is an average fall of 42.3 ± 9.4% from presleep values. PC₂₀ at bedtime was 0.092 mg/ml and fell at awakening to 0.018 mg/ml (p = 0.063). Rlp at the presleep point was 10.0 ± 2.0 cm H₂O/L/s which then rose to 16.2 ± 2.8 cm H₂O/L/s (p < 0.01) at the early sleep time point and further rose to 29.8 ± 9.6/L/s at the late sleep time point (p < 0.001). Thus, each subject had documented nocturnal worsening of their lower pulmonary function at the time of study.

Awake Study: Asthma and Posture

The volume-resistance relationships are presented in Figure 3 and are compared with data from normal subjects as originally published by Vincent and colleagues (3) and Butler and colleagues (2). In our subjects with nocturnal asthma (the data marked current study in Figure 3), Rlp is elevated, but as lung volume is increased resistance still falls significantly (p < 0.05). However, the slope of the volume-resistance relationship in our asthmatics is less steep at approximately 2 cm H₂O/ L/s per liter of volume change versus the normal subject data of Vincent and colleagues or Butler and colleagues, which is less than 1 cm H₂O/L/s per liter of volume change. On the other hand, our data are in agreement with the volume-resistance relationship obtained by Butler and colleagues in asthmatic subjects.

View larger version (17K): [in this window] [in a new window]   Figure 3.   Volume-resistance relationships for the nocturnal asthmatics of the current study in the awake and upright posture. Data from normal subjects using similar techniques was taken from Figure 4 in Vincent and colleagues (3) and from Table III in Butler and colleagues (2) for comparison. The comparative data for asthmatics is from Table III in Butler and colleagues (2).   Lung volume was changed in the current study by applying CNP through the cuirass to sequentially increase lung volume by approximately 0.5 L and then approximately 1.0 L above the naturally assumed FRC. Data points for the current study are the mean ± SEM. TGV = Vtg.

View larger version (11K): [in this window] [in a new window]   Figure 4.   The effect of posture on the volume-resistance relationships for awake, nocturnal asthmatics. When the awake subjects assume the supine posture, the relationship moves parallel and right without an apparent change in slope. Data points are mean ± SEM for the five subjects. TGV = Vtg.

When our subjects assumed the supine posture (Figure 4), there was a 0.40 ± 0.12 L fall in lung volume (p < 0.05) and a further increase in Rlp of 2.6 cm H₂O/L/s, from 6.06 ± 0.76 to 8.70 ± 0.94 cm H₂O/L/s (p = 0.061). However, Rlp still fell when lung volume was increased in all subjects where the slope was again about 2 cm H₂O/L/s per liter volume change, which is similar to that in the upright posture (Figure 4). When expressed as a log resistance/lung volume, the slopes were essentially similar (upright, 0.33 ± 0.08 versus supine, 0.36 ± 0.10; p > 0.10).

Sleep Study: Sleep and Time

Early sleep. On the sleep study night the patients were studied awake and, on average, 33 ± 9 min after obtaining Stage 2 sleep (Figure 5). In the awake stage prior to sleep onset, the volume-resistance values were: Rlp, 10.2 ± 2.0 cm H₂O/L/s and Vtg, 2.94 ± 0.5 L. These values were not significantly different from the awake points obtained on second study day and earlier in the day, which were Rlp, 8.67 ± 0.94 cm H₂O/L/s with Vtg, 2.91 ± 0.19 (p > 0.10). As previously observed (7), with sleep onset, volume fell and resistance rose, with an average rise in resistance of 16.2 ± 2.8 (p = 0.007) associated with a Vtg fall to 2.46 ± 0.13 L (p < 0.05).

View larger version (12K): [in this window] [in a new window]   Figure 5.   The effect of sleep on the volume-resistance relationships. The awake data are replotted from Figure 4 for the awake, supine posture (note that both axes have changed). As measured 33 ± 9 min after obtaining Stage 2 sleep, there is a further increase in resistance and a failure of the increase in lung volume with CNP to decrease Rlp. Data points are mean ± SEM for the five subjects. TGV = Vtg.

However, when end-expiratory lung volume was acutely returned to 3.25 ± 0.15 L with CNP, a level that is slightly above the awake values (2.94 ± 0.15 L), resistance did not fall significantly (16.2 ± 2.8 to 14.9 ± 1.6 cm H₂O/L/s) (p > 0.10). When expressed as a log resistance/lung volume the slopes were: awake supine, 0.36 ± 0.10 versus early sleep, 0.00 ± 0.12, which is significantly different (p < 0.017).

Late sleep. The patients were then allowed to return to a spontaneously assumed lung volume (i.e., negative pressure discontinued). They continued to sleep and were restudied an average 162 ± 23 min later (Figure 6). Resistance showed a progressive rise from 16.2 ± 2.8 to 29.8 ± 9.6 (p < 0.001). When lung volume was altered with CNP, volume rose to 3.01 ± 0.22 L, which was nearly identical to the awake Vtg of 2.94 ± 0.15 L (p > 0.10). As observed in the early period of sleep, resistance failed to fall significantly with the acute increase in lung volume (Rlp at FRC, 29.8 ± 9.6; Rlp at FRC + 0.8 L, 26.6 ± 7.4) (p > 0.10). Indeed, in two of the five subjects, resistance actually rose with the increase in lung volume.

View larger version (11K): [in this window] [in a new window]   Figure 6.   The effect of sleep time on the volume-resistance relationships. The data on the left are replotted from Figure 5 (note change in x-axis) for the early sleep time points (33 ± 9 min), and the data on the right are for the late sleep time point (162 ± 23 min). With increasing time of sleep, there is a further fall in Vtg and a rise in Rlp. Again increasing Vtg with CNP failing to lower Rlp. Data points are mean ± SEM. TGV = Vtg.

Respiratory System Compliance

We had observed that as the night proceeded, the amount of CNP pressure required to raise the Vtg TGV progressively increased (Figure 7, top panel). When related to the rise in lung volume affected by CNP, respiratory system compliance (Crs) could be derived and is presented in Figure 7, bottom panel. Crs progressively fell, although nonsignificantly, from 0.090 ± 0.010 cm H₂O/L when the subjects were upright to 0.077 ± 0.020 cm H₂O/L when the supine posture was assumed (p > 0.10). During early sleep, Crs fell to 0.079 ± 0.020 cm H₂O/L (p < 0.05) and fell more at the late time point to 0.035 ± 0.002 cm H₂O/L (p < 0.05). Hence, nocturnal asthma is associated with a 50% fall in Crs.

View larger version (46K): [in this window] [in a new window]   Figure 7.   CNP required to increase lung volume. Pressures required to increase lung volume to awake supine values increased in early and then to a greater extent during late sleep. When this increase was related to the change in Vtg, respiratory system compliance (Crs) was derived. The fall in Crs with the supine posture were nonsignificant, but the falls in Crs at early (p < 0.05) or late sleep (p < 0.05) were significantly different from the awake condition. Data are mean ± SEM.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Nocturnal asthma is defined as a periodic worsening in lung function that has a circadian variation. The clinical significance of nocturnal asthma has been its association with increased asthma fatality (12). Although the underlying cause of nocturnal asthma is unclear, a circadian inflammatory process has been implicated in a potential cause and effect role (13). In the current study, we have further explored the mechanisms of nocturnal asthma by investigating the dependence of lower pulmonary resistance on lung volume as a reflection of airway-parenchymal interdependence. Loss of airway-parenchymal interdependence was postulated to occur on the basis of the results of our previous investigations (7).

We had postulated that the remarkable rise in pulmonary resistance in nocturnal asthma was due to sleep (6) and a failure of the respiratory system to defend airway caliber by allowing the lung volume (FRC) to fall (7). On the other hand, we had also speculated that at later time points during sleep that airway-parenchymal independence would decrease (uncouple) because of the inflammatory processes that are known to occur in the airways (13). This prediction was based on the observation that when lung volume returned spontaneously to presleep levels the lower pulmonary resistance remained high (7). However, in the current study, we found that even during early sleep the airways and parenchyma had already uncoupled, an effect that appears to persist for hours.

In the second phase of this study, we addressed the role of the first two factors by studying the same subjects in the upright and supine postures. Comparison of our data in the upright posture with the data obtained in normal subjects by Vincent and colleagues (3) or Butler and colleagues (2) allows for evaluation of the asthmatic state. As the asthmatic subjects in the current study still show airway-parenchyma coupling, we would conclude that it is unlikely that asthma per se accounted for the nocturnal uncoupling. Moreover, we know from our previous investigations (6, 7) that normal subjects do not show increases in resistance nor major falls in lung volume during sleep. When the asthmatic subjects assumed the supine posture, increasing lung volume still caused Rlp to decrease, thus suggesting that nocturnal uncoupling was not due to some mechanism associated to postural change. However, when the subjects fell asleep the increase in lung volume to presleep levels as achieved with CNP did not cause Rlp to fall, an effect that persisted to the later sleep point. The possibility that atelectasis and peripheral airway closure occurred during sleep and an acute increase in lung volume did not reopen those units was not evaluated in this study; however, this is unlikely as our prior reported study demonstrated that lung volume maintenance throughout sleep did not improve overnight lung function or bronchial hyperresponsiveness (14). Hence, we conclude that the rise in pulmonary resistance observed during sleep in patients with nocturnal asthma is not accounted for by asthma per se, nor by postural change, but by the uncoupling of airways and parenchyma, which is likely due to some aspect of the sleep state.

As originally described by Mead and colleagues (5), interdependence can be thought of as the physical linkage of the elements of the lung, which tend to decrease the effects of lung expansion or reduction on the rest of the lung. So as a region of lung is expanded, there is a loss of recoil in the surrounding tissue, limiting the effect of the expansion. The linkage between the lung parenchyma and airway caliber is thought to occur through the attachments of the elastic fibers to the airway wall and that are felt to be operational during static states such as those achieved when lung volume was changed with CNP. Reduction in structural interdependence could result in closure and collapse of lung units and a requirement for increased pressures for reinflation. Sleep may also lead to increased transudation of fluid, which would be consistent with the falls in Crs we observed. We have documented that the nocturnal inflammatory response is greater at the level of the alveolar tissue versus the more proximal airway (15). Thus, either the loss of interdependence in this area leads to the increased inflammation and/or the inflammatory response and edema formation is producing the loss of interdependence.

More recently it has been recognized that lung volume has a highly significant role in determining the load to smooth muscle contraction (16) and in turn this has an effect on bronchial responsiveness (17). In this later study Ding and colleagues (17) have shown that lung volume changes of just 0.5 L have a profound effect on the magnitude of airway resistance when a decrease in lung volume causes resistance to rise approximately fivefold. The converse is true for lung inflation. Lung volume has also been recently implicated in the pathogenesis of hyperresponsiveness by the observation that inhibition of large volume changes induces a hyperresponsive state in normal subjects similar to that observed in asthmatics (18). As we have previously shown (7), when patients with nocturnal asthma fall asleep, there is a significant fall in lung volume and an associated rise in lower pulmonary resistance. We had concluded in that study that the rise in pulmonary resistance was in large part due to the fall in lung volume. The results of the present study and another previous study where the fall in lung volume was prevented (14) show that this conclusion was wrong.

The impetus for the current study was twofold: first, we wanted to readdress the role of lung volume falls in airflow limitation in nocturnal asthma by another means and, second, to explore an observation from our previous study (7). In the latter we had noticed that upon awakening the subjects with asthma demonstrated a prompt return to a hyperinflated state; however, lower pulmonary resistance failed to fall. We wondered whether the inflammatory events associated with nocturnal asthma (13) may have led to a physical uncoupling of the airways to the parenchyma as had been predicted to occur in the inflamed lung (19).

Our experiments shed light on some of the possible mechanism that might account for a loss of airways/parenchymal interdependence in this setting. First is the possibility that asthma per se reduces lung interdependence; this is clearly not the case. In Figure 3, we have plotted our data from patients with nocturnal asthma together with data taken from normal subjects published by Vincent and colleagues (3) or Butler and coworkers (2). We feel justified in making these comparisons, as the data of Vincent and colleagues were derived using techniques similar to those of the present study. Volume was derived from a volume-displacement plethysmograph in both cases. Flow was determined at the airway opening using a pneumotachograph. Esophageal pressure was determined with identical methods. The only difference is that in their study lower airway opening pressure was derived from a transtracheal catheter, whereas we used a pressure-tip manometer positioned just above the vocal cords. Although beyond the scope of the current investigation, the vocal cords might contribute to the increase observed in pulmonary resistance. However, we do not think the vocal cords account for our observations, as the vocal cords are not normally closed during inspiration, nor is the size of the aperture known to be lung-volume-dependent, the only difference might be a constant value of resistance, which would shift the relationship right by some constant, but would not be expected to change the slope of the resistance/ volume relationship. Nevertheless, the potential contribution of vocal cord closure remains to be further investigated.

The numerical differences between Vincent and colleagues (3) and Butler and colleagues (2) for normal subjects is easily accounted for by the different techniques used to measure resistance since airways conductance with the classic panting technique normally yields lower values of resistance (20). In comparison, the asthmatic subjects show that not only is there no evidence of a loss of interdependence but, if anything, an increased dependence of pulmonary resistance on lung volume. That there is an increase in volume-resistance relationship in asthma is likely true is suggested by the previous data of Butler and colleagues (2). If we recalculate and plot Butler's data on ours (Figure 3), we see that the increased dependence of pulmonary resistance on lung volume is nearly identical between the studies. The higher volumes in the asymptomatic asthmatics of Butler and colleagues is probably due to known errors in measurement of Vtg (10), which would tend to shift Vtg to higher levels, whereas we used Pes to determine Vtg. Why there is an increased dependence of airway resistance on lung volume in asthmatics is not totally clear.

Increased coupling of airways to lung parenchyma as measured by lung resistance/volume relationship could be explained by airway remodeling. Chronic asthma clearly causes marked structural changes in the lung (21). Airways of asthmatics may exhibit increased collapsibility. First, there is a common experience that during lavage procedures in asthmatics that the airways are prone to closure during the withdrawal of the lavage fluid, even with gentle application of pressure. Second, several studies (22, 23) have shown that patients with asthma have increased gas compression, which has been interpreted as increased airway collapsibility. One could easily imagine that such a structurally altered airway would, in fact, be more influenced by lung elastic recoil and, hence, lung volume. We believe that this best explains the increased fall of resistance for a given volume change observed in the asthmatics.

One explanation for loss of airways/parenchyma interdependence observed during sleep in our asthmatics might be a change in posture. It has long been known that when asthmatics assume a horizontal position that asthma worsens and airflow limitation increases (24). One explanation for this postural asthma is blood pooling and/or airway flooding. Weissler and colleagues (25) have shown that the increase in central blood volume upon assuming the supine posture increases between 0.14 and 0.62 L. Similar changes are thought to occur with submaximal paralysis (26). Unknown is whether this increase is altered by the presence of asthma. In large part, this was the impetus for the second phase of the study. As shown in Figure 4, the volume-resistance relationship of our patients is compared after a postural change from the upright to the supine postures. There is, in fact, an increase in pulmonary resistance at all lung volumes, which might be due to blood pooling or other mechanisms. However, there is no loss of the dependence of airway resistance on lung volumes or, in other words, the slope of the lower airway resistance and lung volume relationship is essentially unchanged.

The next possibility is that the sleep state is responsible for the loss of airway/lung interdependence. In Figure 5, we present the data from the lower pulmonary resistance/lung volume relationships in the awake supine state. As the patients fall asleep and are studied about 30 min into the sleep period, there is a rapid rise in lower pulmonary resistance and a 500 ml fall in lung volume similar to that which we have previously reported (7). When lung volume is acutely returned to presleep levels there is no significant reduction in lung resistancethe airways and lungs are mechanically uncoupled. As this effect is both sleep-related and fairly rapid, it suggests that the process could be neural in nature. This is especially attractive since as discussed above, the volume-resistance relationship in normal subjects are known to be influenced by neural input (3, 27).

Autonomic neural abnormalities have long been known to occur in nocturnal asthma. These abnormalities have been functionally linked to the airflow limitation that occurs in such patients. Normal vagal activity has a circadian rhythm (28, 29). Atropine and ipratropium bromide have both been shown to ameliorate the decrements in lung function that occur at night (30). As an example, the study of Morrison and colleagues (32) showed that when atropine was given at a dose that increased heart rate, the fall in PEFR at 4:00 A.M. was prevented. Also, in normal subjects, volume dependency falls after treatment with atropine (3, 27), suggesting volume-resistance dependence is related to vagal influence and tonic airway smooth muscle.

Finally, the sleep state may have caused a reduced number of large volume sighs. Although we did not specifically investigate this possibility, it could be important for the following reason. Fredbreg and colleagues (33) have recently shown in isolated trachea smooth muscle that reduction in tidal stretches leads to increased force generation. Hence, as the patient falls asleep, lung volume falls, leading to less stretch if the sigh rate also fell, leading to airway narrowing because of increased smooth muscle effective force. This explanation is also likely to explain the failure of conductance to correct upon awakening (6). Further investigation is required to explore this potential mechanism.

The ability of the lung to influence airway caliber through the mechanism of interdependence probably represents the most important defense of airway caliber. In the current study we investigated this mechanism by measuring the volume- resistance relationship in a number of situations in patients with nocturnal asthma. We found that soon after entering the sleep state increases in lung volume no longer caused resistance to fall, i.e., the airways were mechanically uncoupled and this persisted throughout several hours of sleep. This airways uncoupling could not be explained by the asthmatic state or postural changes. The inflammatory response at the level of the alveolar tissue (15) with possible additional neural influences provides one reasonable explanation for this process. The loss of airways-parenchyma mechanical coupling may be the underlying reason that accounts for the increased morbidity and mortality of asthmatics at night.