INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

ASTHMA IS CHARACTERIZED by airway lability and inflammation. The former phenomenon is probably related to that of bronchial hyperresponsiveness, defined as increased obstructive response to various chemical and physical stimuli (22). Although hyperresponsiveness and airway inflammation are thought to be etiologically linked, the mechanism for such a link is elusive. It has become apparent to many investigators in this field that better understanding of the nature of airway hyperresponsiveness is warranted before the link to inflammation can be unveiled.

In 1995, Skloot et al. (25) showed that, when a methacholine inhalation challenge is carried out under conditions where deep inspirations are prohibited, the response to increasing methacholine concentrations is similar in asthmatic and healthy subjects. This work confirmed the previously reported effects of lung inflation in regulating bronchomotor tone (5, 8, 11, 21, 23). Furthermore, Skloot and co-workers (25) demonstrated that the effect of deep inspiration was absent or diminished in asthma, offering support to a hypothesis, originally suggested by Fish et al. (8), that airway hyperresponsiveness in asthma may merely represent reduced ability of a deep inspiration to relax the airways.

In the study of Skloot et al. (25), during the discussion of the potential mechanisms through which lung inflation benefits the airways, they referred to the work of Malmberg et al. (17) who, in an effort to develop an abbreviated methacholine bronchoprovocation for epidemiological studies, came across the observation that a deep inspiration before the administration of methacholine protected the airways from bronchoconstriction. In this work, it also appeared that the protective effect of deep inspiration was more pronounced than the effect obtained if deep inspiration took place after the administration of methacholine.

A bronchoprotective effect of lung inflation implies that the increase in lung volume triggers a process that renders airway smooth muscle resistant to contractile stimuli. In contrast, the presence of a bronchodilating effect alone does not allow us to differentiate between a process that alters the state of airway smooth muscle and a mere stretch relaxation effect. Consequently, understanding the process behind lung inflation-induced bronchoprotection may lead to the elucidation of the mechanism of airway hyperresponsiveness in asthma, provided that healthy subjects and asthmatics are profoundly different in this respect.

This paper describes a series of studies that we performed with the goals to 1) unequivocally establish the presence of a bronchoprotective effect of lung inflation and provide a methodology for quantitative assessment and 2) compare healthy subjects to asthmatics.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Subjects

Study Design

Phase 1: Screening. The screening evaluation included a respiratory questionnaire, skin prick testing to a panel of common aeroallergens (mixed grasses, mixed trees, ragweed, Alternaria, Cladosporium, D. farinae, D. pteronyssinus, cat, dog, mixed cockroaches), and a methacholine challenge. The methacholine challenge was conducted routinely, i.e., after baseline maximal spirometry diluent was inhaled and new spirometric measurements were performed 3 min later, followed by bronchoprovocation with methacholine. The initial concentration of methacholine was 0.025 mg/ml and was increased by half-log increments (0.025, 0.075, 0.25, 0.75, etc.) until a 20% decrease in FEV₁ from the postdiluent values was obtained or until the maximum concentration (75 mg/ml) was delivered. The PC₂₀ of methacholine causing this fall was calculated by interpolation of the dose-response curve. Each dose of methacholine was inhaled with five deep breaths from functional residual capacity (FRC) to total lung capacity (TLC). Methacholine was delivered by a DeVilbiss 646 nebulizer (DeVilbiss, Somerset, PA) connected to a 30-psi compressed air source through a Rosenthal-French dosimeter that regulated the timing of each actuation to 0.6 s. At each dose level, a minimum of three forced spirometric maneuvers from TLC were performed, and the maneuver with the highest FEV₁ was used for analysis.

Phase 2: Establishment of a single-dose methacholine challenge. The protocol we designed to demonstrate that deep inspiration acts as a bronchoprotector and to compare healthy subjects to asthmatics in this respect was a modification of that employed by Malmberg et al. (17). The protocol involves single-dose methacholine challenges. Therefore, before entering each subject into the main study, our task was to determine a single dose of methacholine capable of producing an adequate reduction in lung function. We chose 15% as the minimal desired reduction in FEV₁. Both nonasthmatic and asthmatic subjects underwent the following protocol multiple times, each time separated by at least 1 day. At each occasion, a single dose of methacholine was administered, but that dose was increased every time until the dose that produced >15% reduction in FEV₁ was established. In the beginning of every single-dose protocol, all subjects performed three reproducible maximal (full breath) expiratory maneuvers from TLC to residual volume (RV). The maneuver(s) yielding the best FEV₁ and forced vital capacity (FVC) provided us with baseline measurements. After these three maneuvers, deep inspirations were prohibited for 20 min. At the end of this 20-min period, a single dose of methacholine was administered with five tidal inspirations (FRC to end inspiratory volume) using the dosimeter and nebulizer described above. The subjects were instructed to hold their breath for 5 s after each inspiration and to passively exhale thereafter. Three minutes later, three maximal expiratory maneuvers were again performed, and the outcome was compared with the baseline measurements. The concentration of methacholine delivered in the first single-dose bronchoprovocation was 10 mg/ml for nonasthmatics and 0.025 mg/ml for asthmatics. The subsequent single-dose methacholine provocations were performed with doubling dose incremental steps in nonasthmatics (20 and 40 mg/ml) and half-log incremental steps in asthmatics (0.075, 0.25, and 0.75 mg/ml).

Phase 3: Determination of lung inflation-induced bronchoprotection and comparison of asthmatics to nonasthmatics. Phase 3 included a total of three challenges (Fig. 1). Challenge 1 was the same as the last challenge of phase 2, which identified the single dose of methacholine producing >15% reduction in FEV₁. The same dose was also used in challenges 2 and 3, in which the subjects were instructed to take two and five deep inspirations, respectively, immediately before methacholine inhalation. Each deep inspiration consisted of an inhalation to TLC followed by passive exhalation.

View larger version (17K): [in this window] [in a new window]   Fig. 1.   Single-dose methacholine bronchoprovocation protocol was used to establish and quantify the bronchoprotective effect of deep inspiration and to compare asthmatic to nonasthmatic subjects in this respect. The 3 challenges only differ in the number of deep inspirations preceding the methacholine inhalation (challenge 1 = 0, challenge 2 = 2, and challenge 3 = 5, respectively). The single dose of methacholine capable of producing the targeted 15% reduction in forced expiratory volume in 1 s (FEV1), subsequently used in all 3 challenges, was established by repetition of challenge 1 on different days with increasing single doses of methacholine. Spirometry at the beginning and the end of each challenge was performed conventionally, with maximal expiratory maneuvers. Methacholine was inhaled by tidal breaths.

Phase 4: Control experiments. reproducibility. To test the reproducibility of the single-dose methacholine challenge model, five asthmatic and five nonasthmatic subjects underwent challenge 1 (Fig. 1) on three consecutive days. In addition, the same five nonasthmatic subjects underwent challenge 3 (Fig. 1) on 3 consecutive days to determine the reproducibility of the bronchoprotective effect of deep inspiration.

Statistical Analysis

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Phase 1

Phase 2

View larger version (20K): [in this window] [in a new window]   Fig. 2.   Pulmonary responses of 12 nonasthmatic subjects (each symbol represents one subject) to increasing single doses of methacholine, each delivered on a different visit under the protocol depicted in Fig. 1 (challenge 1). Y-axis values represent FEV1 in the first postmethacholine maximal expiratory maneuver, expressed as a percentage of baseline. Two of the 12 nonasthmatic subjects did not reach the targeted 15% reduction in FEV1 with 40 mg/ml of methacholine and were excluded from the ensuing phases of the study.

Phase 3

View larger version (15K): [in this window] [in a new window]   Fig. 3.   Lung inflation-induced bronchoprotection and comparison of asthmatics to nonasthmatics. The data represent means ± SE of the percent reductions in FEV1 (A) and forced vital capacity (FVC; B) from baseline when 0, 2, or 5 deep inspirations preceded the single-dose methacholine inhalation. In the absence of deep inspirations, the reduction in lung function was similar in the 2 groups. Two and five deep inspirations before inhalation of methacholine protected nonasthmatics from methacholine-induced bronchoconstriction in a dose-dependent fashion (FEV1, P < 0.0001; FVC, P = 0.0008) but had no effect in asthmatics (FEV1, P = 0.38; FVC, P = 0.32). The differences in the percent reductions of FEV1 and FVC between the 2 groups were statistically significant for both challenges 2 and 3 (P < 0.002 in all cases). n, No. of subjects.

Phase 4

Reproducibility. In the nonasthmatics, during the three trials in which no deep inspirations were allowed before inhalation of methacholine, the standard deviation (SD) of the reduction in lung function represented, in all cases, <30% of the average reduction [average coefficient of variation (CV) = 25%]. Between-subject variability was low. In the trials in which five deep inspirations preceded the administration of methacholine, the average CV was higher (78%). This was not surprising as the mean reduction in FEV₁ was low because of bronchoprotection. In asthmatics, the average CV of methacholine-induced reduction in FEV₁ was 33%. Although this was somewhat higher than in the nonasthmatics, the difference in the CVs between the groups was not statistically significant (P = 0.56).

Diluent inhalation challenge. No reduction in spirometric values was obtained with saline inhalation, in the absence of deep inspirations, in nonasthmatics or asthmatics. Mean FEV₁ values at baseline and after the inhalation of the diluent were 3.70 ± 0.33 and 3.68 ± 0.34 for nonasthmatics and 3.22 ± 0.60 and 3.24 ± 0.70 for asthmatics, respectively. These data indicate that merely breathing at low lung volume does not induce bronchoconstriction.

Deep inspirations in the absence of methacholine. In asthmatics, five deep inspirations alone caused no reduction in FEV₁ or FVC. If anything, a small improvement from baseline in both FEV₁ and FVC was noted (Table 2). This trend was also observed in the group of healthy subjects, but it was statistically significant only in the asthmatic group and only for FVC (P = 0.04). Nevertheless, no significant differences were found between the two groups in the deep inspiration-induced spirometric changes from baseline. Midmaximal expiratory flow (MMEF) did not follow the improvement trends observed with FEV₁ and FVC. In fact, the values of this outcome appeared to decline, after the five deep inspirations, to an equal degree in both asthmatics and nonasthmatics; however, the changes in MMEF were not statistically significant in either subject group (P = 0.13 and P = 0.35 in asthmatics and nonasthmatics, respectively).

Effect of the length of the deep inspiration avoidance period. Having established the concentration of methacholine at which a >15% reduction in FEV₁ was developed, after five of the nonasthmatic subjects abstained from deep breaths for 20 min, we examined whether the same effect would still be obtained if the deep inspiration avoidance period was diminished to 10 and 5 min. The results of this experiment are presented in Fig. 4. Two subjects showed similar reductions (>15%) in FEV₁ on all three challenges. In the remaining three subjects, only the 20-min deep breath prohibition period resulted in the anticipated reduction in lung function. These data are in support of those of Moore et al. (19), who suggested that time away from TLC is important to the bronchospastic effect of methacholine in healthy subjects and also indicate that the 20-min deep breath avoidance period allows the induction of bronchospasm in most healthy subjects under this simple, single-dose bronchoprovocation protocol. However, we did not test whether longer periods would have a more pronounced effect.

View larger version (19K): [in this window] [in a new window]   Fig. 4.   Effect of the length of the deep inspiration-prohibition period in 5 nonasthmatic subjects. Data are derived from 3 separate single-dose methacholine challenges and represent the individual percent reductions in FEV1 from baseline when deep inspirations were prohibited for 5, 10, and 20 min, respectively. In 3 of the 5 subjects, only the 20-min deep inspiration prohibition period was adequate to allow the induction of a >15% reduction in FEV1.

Methacholine aerosol deposition pattern in the presence or absence of deep inspirations. The data from this protocol are presented in Table 3 and show that no significant differences in regional aerosol distribution (I/O, P = 0.49; A/B, P = 0.52), distribution homogeneity (skew, P = 0.2), or deposition fraction (P = 0.2) were found between the radioaerosol inhalations that were preceded by deep inspirations and those that were not.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

In 1995, Skloot et al. (25) demonstrated that, in the absence of deep inspirations, methacholine inhalation by nonasthmatic individuals led to severe bronchoconstriction at concentrations that were conventionally regarded as within the asthmatic range (<8 mg/ml). In that study, they found that the airway responsiveness of nonasthmatic subjects, when assessed by partial (from end-tidal inspiratory to RV) forced expiratory maneuvers, was not much different from that of asthmatics. Those observations indicated that lung inflation (through deep inspiration) had a potent beneficial effect on human airways and that this effect was markedly diminished or absent in asthma. Skloot et al. (25) proposed that lung inflation induced bronchodilation and speculated that, in asthma, the phenomenon of airway hyperresponsiveness to a direct spasmogen, such as methacholine, was merely a manifestation of the lack of the bronchodilatory ability of deep breaths. On the basis of the striking effect of deep inspirations on airways reponsiveness, it was expected that, after the induction of severe bronchoconstriction in the absence of deep inspirations, a sequence of deep breaths would rapidly return healthy airways to their baseline, nonconstricted state. The findings of that study disproved this hypothesis. Although the first deep breath taken after the end of the methacholine provocation produced a stronger bronchodilator response in the nonasthmatics, the overall ability of three deep inspirations to induce bronchodilation in this setting was not different between asthmatic and nonasthmatic subjects (25). Given that other investigators have documented the bronchodilatory ability of deep inspiration in experimental designs in which a full spirometric maneuver follows a partial maneuver (26), the explanation Skloot and colleagues offered for their finding was that the potency of this bronchodilatory effect was significantly diminished in the face of preestablished significant smooth muscle contraction (25). This explanation, however, also raised some questions as to whether the striking difference between asthmatic and nonasthmatic subject, with respect to the beneficial effect of deep inspiration, could be solely attributed to a bronchodilator effect, which appeared quite susceptible to increased smooth muscle tone. This uncertainty led to the generation of another hypothesis, that deep inspirations not only have a bronchodilatory but also, most importantly, a bronchoprotective, effect on the airways.

The significance of demonstrating that deep inspiration acts as a bronchoprotector lies in the fact that it can lead to new testable hypotheses regarding the mechanism through which lung inflation operates. It also allows us to speculate that stretching of the airway, which should occur as a result of lung inflation, activates a biochemical or biomechanical process that changes the resting state of the airway smooth muscle or operates in functional antagonism against bronchoconstrictive stimuli. If deep inspiration acted as a bronchodilator alone, it would have been difficult to support such a hypothesis; instead, we would have most probably attributed the bronchodilation to mechanical breaking of actinomyosin bridges in airway smooth muscle. The second major finding of this investigation is related to the magnitude of the bronchoprotective effect of deep inspiration in healthy airways. Bronchoprotection of this level rivals the best pharmacological agents currently available for the prevention of bronchospasm. From this perspective, understanding of the mechanism of lung inflation-induced bronchoprotection has the potential of opening new pathways in the development of more potent therapies for this disease.

To determine whether deep breaths confer bronchoprotection, we developed and tested the validity and reproducibility of a single-dose methacholine provocation protocol, a modification of the protocol previously employed by Malmberg and colleagues (17). Using this protocol, we confirmed the work of these investigators and demonstrated that, after abstaining from deep inspirations for 20 min, nonasthmatic subjects can develop remarkable airway obstruction, some with even a relatively low single dose of methacholine.

With the use of this model, our data fully supported our hypotheses that deep inspirations before the administration of methacholine had a striking protective effect in nonasthmatic subjects and the magnitude of this effect was a function of the number of deep inspirations preceding the methacholine administration. In contrast, the response of asthmatics to methacholine was not affected by the presence or absence of deep inspirations. All asthmatic subjects that participated in our study had mild asthma with a mean percent predicted baseline FEV₁ of 85 ± 8.5% and, with the exception of one subject who used salmeterol, only as-needed usage of inhaled -agonists; however, all asthmatics had high airway responsiveness (methacholine PC₂₀ < 1 mg/ml). Our findings indicate that the lack of the bronchoprotective effect of deep inspiration may be such a cardinal feature of asthma that it is evident even at the lowest level of asthma severity. Also, this observation suggests a strong association between the lack of bronchoprotection and the magnitude of the conventionally determined airway hyperresponsiveness.

During the course of the 20-min quiet breathing period, subjects were only visually monitored to ensure avoidance of deep breaths. It could be argued that "small" deep breaths may have been missed and that the two groups may have been different in this respect; for example, subjects with asthma, as a result of their pathological process, may have been taking deeper breaths or may have had an increased frequency of small deep breaths. However, such small deep breaths are unlikely to have any effect on lung function. Brusasco and colleagues (3) recently demonstrated that doubling the tidal volume of healthy subjects did not affect airway responsiveness to methacholine. Furthermore, because of their almost normal baseline lung function and the very mild nature of their respiratory illness, the subjects with asthma that participated in the present studies had no reason to take deeper breaths or to breathe with an increased frequency of small deep breaths. Finally, because our work demonstrates that deep inspirations are completely ineffective as bronchoprotectors in these individuals, any potential difference in breathing pattern between the two groups during the 20-min quiet breathing period should not be expected to alter the outcome of the single-dose methacholine provocation.

The strong effect of deep inspiration, acting as a bronchoprotector in nonasthmatic subjects, explains the data of Burns and Gibson (4), who failed to show a significant response to methacholine in nonasthmatics using a modification of the protocol used by Skloot et al. (25). The difference was that, in their protocol, each methacholine dose was inhaled with deep inspirations (as opposed to tidal inspirations) and was followed by measurement of partial flows 4 min later. Those deep inspirations, according to our present findings, conferred protection to the airways of the nonasthmatics and did not allow for bronchoconstriction to develop.

In view of previous work (25), it may appear somewhat peculiar that we chose to perform maximal spirometric maneuvers, as opposed to partial ones, to measure the outcome of modified, single-dose methacholine provocations. One can argue that our approach in the present study risked a distorted result because deep inspiration is inherent in the maneuver that is supposed to evaluate the effect of a spasmogen in the absence of deep inspirations. Such a risk would be expected to be greater in the nonasthmatic group, in which the single deep breath, taken during spirometry, could possibly partially reverse the spasmogen-induced obstruction (25). Our decision to employ maximal maneuvers was based on two observations. First, the reliability and reproducibility of maximal maneuvers is higher than that of partial ones. Second, the preliminary work of Skloot et al. (25) showed that, even by incorporating the maximal maneuver, it was still possible to demonstrate significant reductions in FEV₁ in nonasthmatic subjects. In fact, if we consider the cumulative dose of methacholine delivered from the nebulizer during a bronchoprovocation with stepwise, half-log increasing concentrations, such as the one used in the study by Skloot et al. (25), the single concentration of 20 mg/ml of methacholine [at which most of our volunteers reacted in the current study with 15-53% reductions of FEV₁ (Fig. 2)] is equivalent to an ~10 mg/ml top concentration of the stepwise protocol. In their study, Skloot et al. (25) reported a mean reduction in FEV₁ of 36%, with an average top concentration of methacholine of 10.8 mg/ml. Although it did not prevent us from addressing our hypothesis, the incorporation of the full (deep) breath spirometric maneuver in the present study increased the difference between nonasthmatic and asthmatic subjects, with respect to the single dose of methacholine at which the targeted 15% reduction in FEV₁ from baseline was achieved. The single, deep breath produced bronchodilation in the nonasthmatic subjects but not in the asthmatics, in whom the bronchodilatory ability of deep inspiration is probably low (25). We believe that this is why the two groups were quite different in terms of methacholine responsiveness in the present protocol, whereas their responsiveness in the study of Skloot et al. (25) was close.

The various control experiments incorporated in this body of work validate our methodology and rule out certain potential explanations for the observed bronchoprotection in the healthy state. 1) The reproducibility of both the reduction in lung function induced by the single dose of methacholine and of the bronchoprotective effect of deep inspiration in nonasthmatics was good. The average intra-individual CV of the bronchoconstrictive effect of the single-dose methacholine challenge (25 and 33% in healthy and asthmatic subjects, respectively) was better than the average intra-individual CV of the log PD₂₀ obtained by routine, incremental provocation with methacholine, which is ~50%. 2) By performing a saline inhalation challenge in both asthmatics and nonasthmatics, we showed that a 20-min prohibition of deep inspirations alone did not induce bronchoconstriction. 3) We also showed that, in the mild asthmatics of our study, five deep breaths alone do not induce bronchoconstriction; therefore, the bronchoconstriction we observed in asthmatics, in the presence of deep inspirations, was not influenced by the deep inspirations themselves. On the contrary, both asthmatics and nonasthmatics showed some improvement in FEV₁ and in FVC when five deep inspirations were taken after the 20-min quiet breathing period. This improvement was only significant in the asthmatics and only in the case of FVC (Table 2). Because bronchoprotection by deep inspiration was only found in the nonasthmatics, the above findings indicate that the bronchoprotective effect is not related to any change in airway geometry induced by deep inspiration. 4) The bronchoprotective effect of deep inspiration in healthy individuals was so impressive, that we questioned whether it could involve an artifact related to the administration of the aerosol. For this reason, we designed the radiolabeled aerosol study to examine whether the five deep breaths that preceded the inhalation of methacholine could result in substantial changes in airway deposition pattern. The data we collected failed to detect any change in aerosol distribution induced by the preceding deep breaths. Therefore, it is highly unlikely that the bronchoprotective effect of lung inflation is due to differential distribution of aerosolized methacholine in the lungs.

An interesting observation was the increase over baseline in FVC that occurred when the nonasthmatic subjects performed two or five deep inspirations before administration of methacholine. As this was a serendipitous finding, it is difficult to interpret. If we were to attempt an interpretation, however, it would be important to note that similar increases in FVC (and not in FEV₁ or MMEF) were observed in the asthmatics, as well as in nonasthmatics, when five deep inspiration maneuvers were performed in the absence of methacholine (phase 4, control experiment C; Table 2). In the absence of deep inspirations, some air spaces may become atelectatic. The deep breaths taken after the 20-min quiet breathing period may recruit these spaces, which now become filled with air. In the absence of spasmogen, the air that was recruited into these spaces is expirable as part of the FVC. Because the same deep breaths in healthy subjects prevent any methacholine-induced airway constriction from taking place, the air that was recruited in the previously atelectatic spaces continues to be expirable in the presence of spasmogen, and the FVC continues to be increased compared with baseline. However, in asthmatics, methacholine induces bronchospasm, resulting in early airway closure, air trapping, and reduction in FVC.

Two major hypotheses can be raised regarding the mechanism by which lung inflation confers protection on the nonasthmatic lung from a spasmogenic stimulus. The first hypothesis is that this phenomenon results from a direct effect of stretch on airway smooth muscle. Airways are capable of complete closure at very high transmural pressures in vitro (12, 20). Gunst et al. (13) found that maximal, and even submaximal, doses of methacholine could produce airway closure in bronchi at constant transmural pressures in excess of 25 cmH₂O. The closure was completely prevented and the response to methacholine was markedly reduced if the bronchi were subjected to volume oscillations during exposure to methacholine. Thus the in vitro work suggests that volume oscillations during exposure to methacholine can prevent closure, but, once closure occurs, large forces must be generated to overcome the closure. These in vitro observations may provide an explanation for deep inspiration in nonasthmatic subjects markedly attenuating the effect of inhaled methacholine. If healthy subjects do not invoke this method of attenuation, as in the case of the 20-min deep breath prohibition period in our protocol, more constriction takes place and persists, even after deep inspirations. Gunst et al. (13) proposed that depression of contractility during length oscillations is a function of the plasticity of the organization of contractile filaments within airway smooth muscle cells, which allows contractile element length to be reset in relation to smooth muscle cell length as a result of changes in the length to which the muscle is stretched (24). Fredberg and colleagues (9, 10) explain the depression of contractility during length oscillations by a different mechanism. During the initial activation of the smooth muscle, there is a period of rapid cross-bridge cycling and rapid velocity of shortening. With prolonged stimulation, there is slower cycling of cross bridges, and the nature of the cross bridges changes to the latch or frozen state. Also, absence of stretch allows the system to move closer to the frozen state. By taking deep breaths before the administration of methacholine, nonasthmatic subjects place their airway smooth muscle in a condition that is most resistant to the development of latch bridges. It is important to point out that several differences exist between the experimental settings of the in vitro studies performed by Gunst et al. (13) and Fredberg et al. (9, 10) and the in vivo settings of our work. These investigators examined airway smooth muscle after it was exposed to methacholine and investigated the effects of tidal oscillations on methacholine responsiveness. In our experiments, the deep inspirations induce presumably much stronger smooth muscle stretch, and they precede the administration of the spasmogen.

The second hypothesis is that airway stretch activates a process that antagonizes a bronchoconstrictive stimulus in a functional way. For example, airway stretch may activate neural pathways that could lead to inhibition of cholinergic tonic activity (18), NANC bronchodilation (6, 7), or even release of a bronchodilator such as nitric oxide from nonneural sources (1).

Whatever the bronchoprotective effect of deep inspiration is in nonasthmatic subjects, the question remains as to why it is absent or markedly attenuated in asthma. As long as elastic recoil is normal, TLC is not altered in asthma, indicating that asthmatics do not have reduced ability to inflate their lungs. Therefore, in asthma, the function of lung inflation may theoretically become defective at three levels: 1) lung inflation may not be capable of stretching the airway; 2) airway stretch may take place, but its effect on airway smooth muscle is blunted because some elements of airway inflammation defunctionalize this effect; and 3) an intrinsic abnormality of the asthmatic smooth muscle prevents airway stretch from functioning in a bronchoprotective manner. In studies using high-resolution computerized tomography to measure airway area at FRC and TLC in the absence or presence of a bronchoconstrictive stimulus, Brown et al. (2) reported that nonasthmatic subjects do not differ from mild asthmatics in their ability to increase airway area with lung inflation. These observations indicate that the second or third possibilities offer more plausible explanations for the lack of a beneficial effect of deep inspiration in asthma.

In conclusion, we report the presence of a previously suggested, but uncharacterized, physiologic respiratory function, that of bronchoprotection against smooth muscle spasmogens, which is conferred by deep breaths. This function is dose dependent and has remarkable potency. The bronchoprotective effect of deep inspiration is absent in asthma and may be a pivotal pathophysiologic abnormality in this disease. Investigation of the cause of this airway dysfunction may offer the long-awaited link between airway inflammation and airway lability in asthma.