INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

In 1995 we reported that in the absence of deep inspirations (DIs), healthy subjects develop obstructive pulmonary responses after inhaling methacholine, with many of them developing these responses at methacholine concentrations within the asthmatic range (< 8 mg/ml) (1). In contrast, under a conventional provocation protocol that includes DIs, these individuals have a minimal or no response to methacholine when they inhale at least 10-fold higher concentrations of this spasmogen (up to 75 mg/ml). This observation was in concordance with previous suggestive work by several investigators (2, 3), and emphasized the powerful effect of lung inflation on the airways. The importance of this phenomenon is further emphasized by its absence or impairment in asthma (1, 2), indicating that it may have major pathophysiologic significance in this disease, especially with respect to airway hyperresponsiveness. Thus, better understanding of the lung inflation effect on healthy airways may allow elucidation of fundamental mechanisms of airway obstruction in asthma.

In our previously reported experiments (1), following the induction of significant airway obstruction with stepwise greater concentrations of methacholine in healthy and asthmatic subjects in the absence of deep breaths, we asked all volunteers to perform three consecutive DIs in the form of maximal spirometric maneuvers, with the expectation that these DIs would rapidly reverse the bronchoconstriction of the healthy subjects, and that this effect would clearly differentiate healthy from asthmatic subjects. The surprising finding was that although DIs progressively improved lung function in both groups, healthy subjects did not differ from asthmatic subjects in the rate and magnitude of improvement in their lung function. In other words, the bronchodilating ability of lung inflation was the same in the two groups. Taken together with the rest of our findings, this observation suggested that the difference between the two groups in methacholine responsiveness did not merely result from an impaired bronchodilatory ability of lung inflation in asthma, but from some other effect of DI that may have a more profound role.

In 1993, Malmberg and colleagues, in an effort to develop abbreviated airway challenges for epidemiologic purposes, performed single-dose methacholine provocations and found that the obstructive effect was highly dependent on the time interval between the administration of the spasmogen and the preceding spirometry: the longer the interval, the stronger the effect of methacholine (4). Since the spirometry before methacholine involved DIs, our interpretation of Malmberg and colleagues' observation was that lung inflation has a bronchoprotective effect and that this effect may be responsible for the striking difference in airway responsiveness to methacholine between provocations with and without DI in healthy subjects. We recently tested this hypothesis, and established that DIs performed before a methacholine challenge have a remarkable protective effect on the airways of healthy subjects. We further found that this effect was absent in patients with asthma (5).

On the basis of the foregoing observations, it was reasonable to hypothesize that the bronchoprotective effect of DI is stronger than its bronchodilatory effect. Also, with the likely possibility that both effects result from the stretch that airway smooth muscle undergoes during DI, it was reasonable to expect that the two effects would be correlated. The present study was designed to test these hypotheses.

We used a modified, single-dose methacholine bronchoprovocation in which the inhalation of the spasmogen is preceded by a DI-free 20-min period. This, according to the findings of Malmberg and colleagues (4), eliminates the effect of the DIs that are taken in the course of baseline spirometry. Depending on the lung inflation effect that we wanted to test, we asked subjects after the 20-min DI-free quiet breathing period to perform DI maneuvers just before the administration of methacholine (to assess bronchoprotection) or immediately after the lung function evaluation done after methacholine administration (to assess bronchodilation). In our previous work, the mean reduction in FEV₁ in healthy subjects at the end of a multiple-dose, DI-free methacholine challenge was 36% (1). A similar level of bronchial obstruction (20% to 40% reduction in FEV₁ from baseline), at which we would compare the bronchoprotective versus the bronchodilating effects of lung inflation, was targeted in the current study. Because we also hypothesized that the magnitude of the effect of lung inflation on airway caliber might change with the degree of induced narrowing, we incorporated additional methacholine provocations to examine DI-induced bronchoprotection and bronchodilation at a milder level of airway obstruction (10% to 20%).

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Subjects

We studied 19 healthy, nonsmoking adults, recruited among the employees of Johns Hopkins University. All subjects were screened through an asthma questionnaire, allergy skin testing, and routine methacholine inhalation challenge (Phase 1, see the following discussion). These subjects had never experienced symptoms consistent with the diagnosis of asthma; all were skin-test negative, with the exception of one individual who had mildly positive skin prick tests to Dermatophagoides pteronyssinus and D. farinae. Under the routine provocation protocol, all subjects received the highest dose of methacholine (75 mg/ml) which produced less than a 15% reduction in FEV₁. Coffee or tea were not allowed in the morning before a test. Subjects were tested at least 4 wk after their most recent upper respiratory infection.

The study was approved by the Johns Hopkins Bayview Medical Center Institutional Review Board, and all subjects gave written, informed consent prior to participation.

Study Design

Phase 1. On the first visit, all subjects underwent a screening evaluation that included allergy skin testing and routine methacholine challenge (6). Skin testing was done with a panel of 10 common aeroallergens, using both the epicutaneous and the intradermal (1:100 dilution) methods. The routine methacholine challenge was performed as follows. First, we performed baseline spirometry in triplicate. Sterile diluent (phosphate-buffered saline solution) was then administered by inhalation with five deep breaths, from FRC to TLC. Three minutes later, spirometry was repeated in triplicate and the first concentration of methacholine (0.025 mg/ml) was inhaled, followed again by spirometry after 3 min. The sequence of methacholine inhalation followed by spirometry was repeated with increasing concentrations of methacholine (0.075, 0.25, 0.75, 2.5, 7.5, 25, and 75 mg/ml) until a 20% decrease in FEV₁ from the value after diluent given alone was obtained, or the highest concentration was delivered. After each step, the best FEV₁ among three acceptable maximal spirometric maneuvers was recorded. Methacholine was delivered through a deVilbiss 646 nebulizer (de Vilbiss Co., Somerset, PA) attached to a Rosenthal dosimeter (Laboratory for Applied Immunology, Inc., Fairfax, VA), which was activated by inhalation for 0.6 s at a time and was driven by compressed air at 30 psi.

Phase 2. This part of the study was directed at determining the two single concentrations of methacholine that induced a 10% to 20% (Dose 1) and a 20% to 40% (Dose 2) reduction in FEV₁ from baseline with prohibition of DIs, and also at measuring the bronchodilatory effect of DI. Our previously described single-dose methacholine challenge was again used (5). Three maximal forced expiratory maneuvers were performed at baseline, and the best FEV₁ and FVC were retained for analysis. Subjects were then instructed not to take DIs for a 20-min period. We did not monitor the subjects' breathing patterns, but none of the subjects reported any difficulty in being compliant with this protocol. After this period, a single concentration of methacholine was administered in five tidal breaths from the dosimeter, and the subject was asked to refrain from taking deep breaths for another 3 min, after which lung function measurement was again repeated with a single, maximal (involving a DI) forced expiratory maneuver (Figure 1a). The difference between the postmethacholine FEV₁ and the baseline FEV₁ was expressed as a percent change from baseline. The single-dose challenge described previously was given multiple times, on different days, at the same time of the day. On each occasion, the single dose of methacholine was increased until the expected level of reduction in FEV₁ was attained; at that time, the bronchodilatory ability of DIs was tested (see the subsequent discussion). The concentrations of methacholine used in this protocol were 10, 20, 40, and 75 mg/ml. One subject underwent a provocation with 5 mg/ml of methacholine. Two to four visits per subject were necessary to complete this phase of the study. To measure the bronchodilatory effect of DI, the protocols of the single bronchoprovocation challenges, in which the two targeted reductions in FEV₁ were obtained, were immediately extended by asking the subjects to take four DIs after the postmethacholine spirometry. An additional maximal spirometry was then performed (Figures 1a and 1b). We deliberately chose four DIs because the fifth DI was included in the first maximal spirometric maneuver, which took place after the inhalation of methacholine. To evaluate bronchodilation, the change in lung function from baseline recorded with the first postmethacholine spirometry (Figure 1a, pre-DI) was compared with that recorded with the second spirometry (Figure 1b, post-DI).

View larger version (13K): [in this window] [in a new window]   Figure 1.   Single-dose methacholine challenge protocols used to determine the dose of methacholine inducing the targeted reductions in FEV1 and to assess the bronchoprotective (a and c) and the bronchodilatory (a and b) effects of DIs. For details refer to discussion of study phases 2 and 3 in METHODS. DI = deep inspiration; bsl = baseline.

Phase 3. This phase of the study was directed at determining the bronchoprotective effect of DI. The two concentrations of methacholine (Dose 1 and Dose 2) identified in Phase 2 as those causing 10% to 20% and 20% to 40% reductions in FEV₁ from baseline, respectively, were used in a challenge in which five DI maneuvers were performed at the end of the 20-min DI-free period before the inhalation of methacholine (Figure 1c). To measure the bronchoprotective effect, we compared the methacholine-induced change in FEV₁ from baseline with that of the Phase 2 challenge, which was done with the same single dose of methacholine, but in which DIs were not included (Figure 1a).

The choice of five DIs as the maneuver for testing the effects of lung inflation was based on our previous work with this model (5), in which we have demonstrated that the bronchoprotective effect of DI is a function of the number of DIs, and that five of these maneuvers provides almost complete protection from a subsequent inhalation of methacholine.

Data Analysis

All outcomes in the study were derived from spirometry. Spirometric values, especially in healthy populations, are normally distributed, and we therefore applied parametric statistics. In our primary analysis, we assessed the effects of DI on methacholine-induced bronchoconstriction with two-tailed paired t tests. For bronchodilation, the values of FEV₁ or FVC from the first spirometry after the administration of methacholine were compared with those obtained after the ensuing series of DIs (Figures 1a and 1b). For the bronchoprotective effect, we compared the spirometric values obtained after the administration of methacholine in the protocol in which no DIs were involved (Figure 1a) with those in the protocol in which five DIs preceded methacholine administration (Figure 1c). Paired t tests were also used to compare the spirometric values obtained after the implementation of DIs with the baseline values in both protocols assessing bronchodilation and bronchoprotection. In order to examine whether the bronchodilating and the bronchoprotective effects of DI were complete, we analyzed the absolute or the percent reductions in FEV₁ and FVC from baseline, after the respective DI maneuvers, using one-sample t tests.

In our primary analysis, we also calculated the following indices, in order to compare the bronchodilatory with the bronchoprotective effects of DIs.

1. We defined a bronchodilation index (BDI) as the ratio of the fractional change in FEV₁ from baseline obtained after the first maximal spirometric maneuver after methacholine to the change obtained after the second maneuver (the maneuver that followed four DIs) in the challenge protocol detailed in Figures 1a and 1b. The formula for BDI was:

(1)

where FEV_1bsl was the baseline value (best of three efforts) in the protocol shown in Figures 1a and 1b) FEV_1pre-DI was the value from the first spirometric maneuver performed 3 min after methacholine administration, and FEV_1post-DI was the value from the single spirometry that followed the DI maneuvers.

2. We defined a bronchoprotection index (BPI) as the ratio of the fractional change in FEV₁ from baseline induced by the challenge without any premethacholine DIs (Figure 1a) to the fractional change induced by the challenge that included five premethacholine DIs (Figure 1c). The formula for BPI was:

(2)

The numerator of this formula (first term in parentheses) derives from the protocol depicted in Figure 1a, and the denominator (second term in parentheses) derives from the protocol depicted in Figure 1c. FEV_1bsl was the respective baseline value (best of three efforts) from each protocol; FEV_1pre-DI was the value from the first spirometric maneuver performed 3 min after methacholine administration in the protocol without DIs; and FEV_1post-MCh was the value from the first spirometric maneuver performed 3 min after methacholine administration in the protocol in which five DIs preceded the administration of methacholine.

The significance of bronchodilation and bronchoprotection at each dose of methacholine was tested with a one-sample test by calculating 95% confidence intervals (CIs) for BDI and BPI. The two BDI and the two BPI values obtained with Dose 1 and Dose 2 of methacholine were compared with each other through two-tailed paired t tests. Also, at each dose of methacholine, the respective BDI was compared with the BPI, also through two-tailed paired t tests. To address the second hypothesis of the study, that the bronchoprotective and the bronchodilatory effects of DI are correlated, we constructed simple regressions to evaluate the relationships between BDI and BPI during the two methacholine challenges (Dose 1 and Dose 2).

In all analyses, values of p  0.05 were considered statistically significant.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Single-Dose Methacholine Challenges

To obtain the two targeted levels of reduction in lung function, we performed modified single-dose methacholine challenges in 19 subjects. Individual data are shown in Table 1. In 10 subjects (Subjects 1 to 10), both levels of bronchoconstriction were attainable. These subjects were included in the next phases of the study (Phases 2 and 3). In these subjects, whose demographics and baseline lung function characteristics are given in Table 2, the median Dose 1 and Dose 2 of methacholine was 20 mg/ml and 58 mg/ml, respectively. Dose 1 induced a 12% mean reduction in FEV₁ and Dose 2 a 42% reduction compared with baseline FEV₁. In Subjects 11 to 17, only one level of reduction in FEV₁ was achieved at the concentrations of methacholine we used. In Subjects 18 and 19, lung function did not decrease by  10% even with the 75 mg/ml concentration of methacholine.

Bronchodilatory and Bronchoprotective Effects of DI

The effectiveness of DI in reversing methacholine-induced airway obstruction was first evaluated with absolute spirometric outcome values. Baseline values of FEV₁ and FVC were not different between the challenge in which Dose 1 and that in which Dose 2 of methacholine were administered (FEV₁: 3.63 ± 0.20 L [mean ± SEM] with Dose 1 and 3.64 ± 0.21 L with Dose 2; FVC: 4.44 ± 0.21 L with Dose 1 and 4.50 ± 0.21 L with Dose 2).

The FEV₁ and FVC values obtained from the spirometry that followed the postmethacholine DIs were always higher than those obtained from the first spirometry that took place 3 min after methacholine inhalation (pre-DI) (Figures 1a and 1b), with the difference reaching significance in the Dose 2 challenge (p < 0.0001 for FEV₁ and p < 0.0005 for FVC). BDI was significant even at Dose 1 (95% CI: 1.15 to 2.08; mean: 1.62). However, the effect of DI was not sufficient to fully reverse methacholine-induced obstruction after either of the two doses. In the Dose 1 protocol, the bronchoconstriction remaining after the DIs was significant, at least with respect to FEV₁ (difference from baseline for FEV₁: 0.32 ± 0.05 L, p < 0.0001; for FVC: 0.23 ± 0.11 L, p = 0.07). In the Dose 2 protocol, the remaining reductions in FEV₁ and FVC were statistically significant (difference from baseline for FEV₁: 0.51 ± 0.07 L, p < 0.0001; for FVC: 0.46 ± 0.10 L, p = 0.002). The bronchodilation findings are also shown in Figure 2, in which FEV₁ and FVC are expressed as percent reductions from baseline. As expected, analysis of the data in this form yielded statistical results in full agreement with the analysis of the absolute spirometric values.

View larger version (35K): [in this window] [in a new window]   Figure 2.   Reductions in FEV1 and FVC from baseline in the course of the protocol for evaluating the bronchodilatory ability of DI. Bars represent mean ± SEM. Hatched bars represent percent reductions in spirometric outcomes obtained immediately after the single-dose methacholine inhalation. Open bars represent percent reductions in spirometric outcomes obtained after the five DIs following methacholine inhalation. *p < 0.0001 for the postmethacholine FEV1 recorded before versus that after deep inspirations. **p < 0.0005, for the post-methacholine FVC recorded before versus that after deep inspirations. p = 0.007 for the percent remaining reduction in FEV1 at Dose 1 compared with that at Dose 2.

The bronchoprotective effect of DI was evaluated by comparing the FEV₁ and FVC values recorded after the inhalation of each of the two single doses of methacholine on the day on which the subjects were instructed not to take DIs before bronchoprovocation and on the day on which DIs were taken (Figures 1a and 1c). In none of the four protocols were baseline values significantly different. At Dose 1, the postmethacholine FEV₁ and FVC were higher on the day when DIs were taken before methacholine administration (FEV₁: 3.34 ± 0.20 L versus 3.20 ± 0.16 L; FVC: 4.31 ± 0.22 L versus 4.18 ± 0.18 L), but the differences were not statistically significant (p = 0.20 and p = 0.39, respectively). The BPI, however, was significant at Dose 1 (95% CI: 1.12 to 2.91; mean: 2.02). At Dose 2, the postmethacholine FEV₁ obtained during the protocol that included DIs was significantly higher than that obtained in the protocol without DIs (FEV₁: 3.32 ± 0.17 L versus 2.07 ± 0.15 L, p < 0.0001). Similar data were obtained for FVC. When the spirometric outcomes were expressed as percent reductions from baseline (Figure 3), the remarkable protective effect of DI after Dose 2 of methacholine was confirmed. Still, the reduction in FEV₁ in the protocol in which DIs preceded methacholine administration was significant (p < 0.005), suggesting that the protection with DIs is not complete. On the other hand, at both doses of methacholine, DIs conferred complete protection when FVC was considered (percent reduction from baseline with DIs for Dose 1: 1.6 ± 2.99%, p = 0.61; and for Dose 2: 2.4 ± 2.18%; p = 0.30).

View larger version (38K): [in this window] [in a new window]   Figure 3.   Reductions in FEV1 and FVC from baseline in the course of the protocol for evaluating the bronchoprotective ability of DI. Bars represent mean ± SEM. Hatched bars represent percent reductions in spirometric outcomes after the single-dose methacholine challenges that were not preceded by DIs. Open bars represent percent reductions after the challenges in which five DIs were taken before methacholine administration. *p < 0.0001 for the postmethacholine FEV1 and FVC recorded in the protocol without DIs versus the postmethacholine FEV1 and FVC obtained in the protocol that included DIs before methacholine inhalation.

It is noteworthy that with both bronchoprotection and bronchodilation, the effect of DIs was more impressive in the Dose 2 than in the Dose 1 protocol. This observation is confirmed in Figure 4, in which the BDI and BPI are plotted for each subject at both doses of methacholine. In Figure 4 it is seen that both indices increase significantly with the increased dose of methacholine (p = 0.0006 for BDI and p = 0.007 for BPI). It was also interesting that in the case of bronchoprotection (Figure 3), the remaining percent reduction of FEV₁ from baseline was remarkably similar after both doses of methacholine (8 ± 2.14% versus 8.4 ± 1.49% for Dose 1 and Dose 2, respectively; p = 0.87). In the bronchodilation protocol, on the other hand (Figure 2), the remaining percent reduction in FEV₁ after DIs at Dose 2 of methacholine was higher than the remaining percent reduction after Dose 1 (Dose 1: 8.76 ± 1.21%; Dose 2: 13.88 ± 1.55%; p = 0.007). The respective FVC data were similar, although the difference between the Dose 1 and Dose 2 bronchodilation protocols was not statistically significant (percent reduction with Dose 1: 4.74 ± 2.00%; with Dose 2: 10.03 ± 2.10%; p = 0.07).

View larger version (15K): [in this window] [in a new window]   Figure 4.   Individual data for BDI and BPI after low (Dose 1) and high (Dose 2) doses of methacholine. The open circles in the left panel and the closed rectangles in the right panel represent mean ± SEM. *p < 0.001 for the two doses of methacholine compared with one another; p = 0.02 for BPI versus BDI at methacholine Dose 2.

Bronchodilation versus Bronchoprotection

To compare the bronchodilatory and bronchoprotective effects of DI, we first compared the absolute FEV₁ and FVC values obtained at the end of the bronchodilatory with those obtained at the end of the bronchoprotective maneuvers (Figures 1b and 1c, respectively). Although no difference was found at Dose 1 of methacholine (FEV₁: 3.32 ± 0.19 L versus 3.34 ± 0.20 L, p = 0.74; FVC: 4.21 ± 0.19 L versus 4.31 ± 0.22 L, p = 0.20), the spirometric values obtained at the end of the bronchodilation protocol, when Dose 2 of methacholine was administered, were significantly lower than those after the bronchoprotection protocol (FEV₁: 3.13 ± 0.18 L versus 3.32 ± 0.17 L, p = 0.02; FVC: 4.04 ± 0.19 L versus 4.39 ± 0.15 L, p = 0.01). Similar results were obtained when the data were analyzed as percent remaining reductions in lung function, as shown in Figures 2 and 3. Baseline values in the two Dose 2 bronchoprovocation protocols were not different (FEV₁: p = 0.93; FVC: p = 0.72).

We then compared the BDI and BPI at both the Dose 1 and Dose 2 challenges (Figure 4). The BDI and BPI obtained during the Dose 1 protocols, when a single dose of methacholine inducing only mild bronchial obstruction was inhaled, were not significantly different from one another (BDI: 1.62 ± 0.21; BPI: 2.02 ± 0.40; p = 0.26). With Dose 2 of methacholine, the increase in these two indices was greater for bronchoprotection, with the BPI becoming twofold greater than the BDI (BDI: 3.40 ± 0.43; BPI: 6.98 ± 1.42; p = 0.02).

To evaluate the extent to which the remaining reduction in FEV₁ from baseline after DIs (Figures 1a and 1b), or the percent reduction in FEV₁ from baseline in the protocol in which DIs preceded the inhalation of methacholine (Figure 1c), were associated with the degree of methacholine-induced bronchoconstriction in the absence of DIs, we generated the simple regression plots shown in Figure 5. In Figure 5a it is evident that with increasing levels of bronchoconstriction in the protocol without DIs, the remaining bronchoconstriction after the bronchodilatory DI maneuvers also increased. This suggests that the bronchodilatory effect of DI is limited. In contrast, and as shown in Figure 5b, increased bronchoconstriction in the absence of DIs does not seem to limit the bronchoprotective effect of DIs: the bronchoconstriction in the protocol in which DIs preceded the administration of methacholine remained in the range of 10%, regardless of the bronchoconstriction induced by the same dose of methacholine in the absence of DIs.

View larger version (13K): [in this window] [in a new window]   Figure 5.   Relationships between the degree of bronchial obstruction induced by single-dose methacholine provocations in the presence and absence of DIs. (a) DIs were taken after the administration of methacholine (bronchodilation protocol). (b) DIs were taken before administration of methacholine (bronchoprotection protocol).

The concept that the bronchodilatory effect of DI develops a limitation with increased bronchoconstriction is further demonstrated in Figure 6, in which the BDI and BPI are plotted against each other at Dose 1 and at Dose 2. The relationships indicate that within the range of the methacholine doses we administered, as BPI increases, BDI reaches a plateau. This is clearly more evident with the Dose 2 plot. The correlation between BDI and BPI was not statistically significant for either dose (for Dose 1: r = 0.54, p = 0.10; for Dose 2: r = 0.49, p = 0.15).

View larger version (13K): [in this window] [in a new window]   Figure 6.   Correlations between the BDI and BPI at each dose of methacholine. The continuous regression line is derived from the Dose 1 data (open circles; r = 0.54, p = 0.10). The interrupted line is derived from the dose 2 data (closed circles; r = 0.49, p = 0.15).

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

We recently established a methodology by which we can examine the bronchoprotective effect of DI (5). We were able to show that this effect is present and is very powerful in the healthy lung. We also showed that it is absent in patients with asthma. In the present study we strongly confirmed the bronchodilatory effect of DI, and further compared it with the bronchoprotective effect of DI in healthy subjects. More importantly, we confirmed our hypothesis that the bronchoprotective effect of DI is stronger than the bronchodilatory effect. Our findings show that bronchodilation and bronchoprotection with DI are not different at a low level of bronchoconstriction, and that their relative potency is increased with increasing doses of methacholine. However, this relative increase is less pronounced in relation to the bronchodilatory than in relation to the bronchoprotective capacity of DI. In other words, a limitation of the bronchodilatory effect of DI is observed with increasing bronchoconstriction.

The experimental challenge model used in this study has the unique feature of involving single doses of methacholine. This design allows a relatively easy assessment of DI-induced bronchoprotection and bronchodilation, since DIs can be introduced before or after the administration of methacholine. An important element of this model is the 20-min quiet breathing period that precedes the challenge. It is known that healthy humans spontaneously sigh approximately every 6 min (7). The exact role of this function is unknown, but our findings suggest that sighing probably provides constant bronchoprotection against extrinsic or even intrinsic spasmogenic mechanisms. Malmberg and colleagues (4), as well as ourselves (5), have previously shown that with longer periods during which DIs are absent, the effects of methacholine are potentiated in healthy humans. With 20 min of quiet breathing, the majority of healthy individuals will develop substantial bronchoconstriction from methacholine. It is worth noting, however, that although in our original study (1), healthy subjects, in the absence of DIs, developed an average 36% reduction in FEV₁ upon challenge with a stepwise increasing dose of methacholine at a top median dose of 7.5 mg/ml, the median single dose of methacholine in the current study causing a 42% reduction in FEV₁ was 58 mg/ml. Our hypothesis is that this discrepancy is related to a potential additive or even synergistic effect of multiple methacholine administrations, as opposed to a single dose. This hypothesis, however, needs to be tested.

Lung inflation has a stretching effect on the airways, especially the noncartilaginous airways. Although there is good reason to hypothesize that the stretch applied to the airways through a DI may account both for the ability to prevent and to reverse bronchial narrowing, the underlying mechanism through which airway stretching operates is still unknown, and the possibility of a different mechanism behind each of these two effects of DI also needs to be entertained. For example, stretching could cause a breakage in myosin-actin crossbridges, leading to the relaxation of smooth muscle. This mechanism could explain the recovery from induced bronchoconstriction (i.e., the bronchodilatory capacity of DIs); however, when a DI takes place under resting conditions, in which the number of myosin-actin bridges is relatively small and the rate of attachments is close to the rate of detachments induced by breathing (8), the additional disruption of myosin-actin interactions produced by a DI may not have enough of an effect on the airway smooth muscle to explain the dramatic phenomenon of bronchoprotection.

Fredberg and associates (9) have recently advanced an interesting hypothesis that can be used to interpret some of our findings. This hypothesis claims the presence of a dynamic equilibrium within airway smooth muscle that is modulated by the tidal stretches involved in normal breathing. A greater stretch imposed by DI in the resting state may put the muscle fibers into a condition of greater disequilibrium that makes the activation of smooth-muscle contraction more difficult. On the other hand, when contraction, and therefore equilibrium, has been reached, the energy required to disrupt this state is much higher. This is confirmed by in vitro observations by Gunst and colleagues (10) that volume oscillations applied to bronchi can prevent closure, and that once closure has been attained, large forces are required to open airways. Under this hypothesis, DI will be more effective before than after induced bronchoconstriction; this is compatible with our findings.

An additional hypothesis regarding the mechanism through which stretch exerts its effects on the airways is that DI may induce changes in the status of airway smooth muscle through the release of relaxant factors. There is evidence, for example, that bronchodilating autacoids, such as prostaglandins E₂ and I₂, or neurotransmitters (e.g., vasoactive inhibitory polypeptide), are released in response to stretch (11). More recently, a possible role in smooth-muscle relaxation has been suggested for endogenous nitric oxide (NO), as well as for atrial natriuretic peptide (ANP) (12, 13). In the heart, ANP is known to be released with stretching of atrial tissue; a similar mechanism may be operable in the lower airways. Another possibility is that stretch receptors in the airways are activated by DI, leading to central inhibition of parasympathetic tone. Unfortunately, there is little experimental evidence for or against the foregoing mechanisms, and they therefore remain purely speculative.

The relative increase in the bronchodilatory and the bronchoprotective effects of DI in the context of an increased dose of a spasmogenic stimulus (Figures 2-6) is worth discussing. One explanation for this increase may be that the forces of airways-parenchyma interdependence, which are thought to mediate lung inflation-induced airways stretching (14, 15), may become stronger with increased bronchoconstriction, as demonstrated by Moreno and colleagues (16). This mechanism, however, would apply only to the bronchodilating and not to the bronchoprotective effects of DI; the latter are induced before the generation of smooth-muscle spasm by methacholine. Another explanation for the relative increase in bronchodilation and bronchoprotection with a spasmogenic stimulus is that the effect of DI is potent enough to overcome the smooth-muscle spasm induced even by higher doses of the spasmogen (by both doses of methacholine in the present study). If this was true, however, one would have expected that at the lower methacholine dose in our study, DIs would have conferred complete protection from, or would have completely reversed, airways obstruction. Although this is the case with bronchoprotection and FVC, a component of methacholine-induced reduction in FEV₁ is resistant to bronchoprotection; as shown in Figure 3, at both doses of methacholine, an approximately 7% reduction from baseline in FEV₁ is observed despite the five DIs that precede administration of the spasmogen. We believe that this obstructive component reflects large airways constriction, a phenomenon that should not theoretically be as sensitive to the beneficial effects of lung inflation; lung inflation is expected to stretch large airways to a lesser degree than small airways. With bronchodilation, the picture is more complicated. Not only did we observe a resisting obstructive component, as in the case of bronchoprotection, but the remaining obstruction after the five DIs at Dose 2 of methacholine was higher than that at Dose 1 (Figure 2). This observation, associated with the data presented in Figures 4-6, suggests that the bronchodilatory effect of DI is limited as the spasmogenic stimulus increases.

The limitation in the bronchodilatory capacity of DI is probably a consequence of the development of small airways obstruction and closure. Once muscle shortening has taken place, especially in the absence of stretch other than that from tidal oscillations, the muscle becomes resistant to the dilating effect of DI because it is approaching latch-state equilibrium (17, 18). In addition, with small airways closure, surface tension increases and large pressures are required to open the collapsed airways (19).

In conclusion, the results of the present study confirm the hypothesis that the bronchoprotective effect of DI is stronger than the bronchodilatory effect, mainly because of the limitation on the effects of airway stretch imposed by obstructed small airways. We also found a weak relationship between the bronchoprotective and bronchodilating effects of DI, suggesting that the two phenomena may involve two different physiologic mechanisms. Since DI-induced bronchoprotection is such a powerful property of the healthy human lung, yet is limited or absent in asthma, further investigation of its nature may enhance understanding of the pathophysiology of this condition.