American Journal of Respiratory and Critical Care Medicine

Measurements of lung resistance and elastance (Rl and El) from 0.1 to 8 Hz reflect both the mean level and pattern of lung constriction. The goal of this study was to establish a relation between a deep inspiration (DI) and the heterogeneity of constriction in healthy versus asthmatic subjects. Constriction pattern was assessed from measurements of the Rl and El from 0.1 to 8 Hz in seven healthy subjects and in 12 asthmatics. These data were acquired before and after a DI and before and after a standard methacholine challenge versus a modified challenge in which a DI is prohibited. Generally, avoidance of a DI increased responsiveness. In healthy subjects and in those with mild-to-moderate baseline asthma a bronchial challenge, especially during self-inhibited DI, produced a heterogenous pattern of constriction inclusive of randomly distributed airway closures or near closures. Nevertheless, such subjects were able to reopen their airways via a DI. In contrast, in subjects with severe baseline asthma, there is a more extreme heterogeneous constriction pattern with random airway closures even at baseline. Further, there is no residual bronchodilatory effect of a DI either before or after bronchial challenge. We conjecture that inflammation and wall-remodeling facilitate a dangerous degree of heterogeneous constriction inclusive of airway closures or near closures, and contribute to the prevention of a DI from having a residual bronchodilatory effect.

Current dogma is that a spontaneous asthmatic attack occurs when smooth muscle constricts causing excessive airway narrowing that cannot be mitigated by tidal breathing or deep breaths (1-4). Skloot and colleagues (3) recently focused on the mechanical modulation of smooth muscle tension as central for avoidance of an asthmatic condition. They found that when deep inspirations (DIs) were prohibited, healthy subjects displayed an enhancement in their constriction hyperresponsiveness to methacholine. They concluded that asthmatics have a diminished capacity to modulate smooth muscle tension when exposed to a bronchial provocation and that this modulation is crucial to preventing sustained asthmatic conditions. Subsequent studies provide supportive data that tidal modulations mitigate bronchial responsiveness (2, 5, 6). Could reductions in modulation by preventing deep breaths be mimicking the role of airway inflammation and remodeling in asthma? Recent studies (7-9) have indicated that while prohibiting DI amplifies reactivity, it does not do so in a manner akin to that of asthmatics.

Virtually all studies that have reexamined the bronchodilatory or bronchoprotective effects of a DI have relied on a spirometric index of lung function, which represents the aggregate impact of many structural mechanisms on the ability to generate airflow. Moreover, spirometric measurements are inherently awkward for examining the impact of a DI. Generally, an asthma paradigm derived exclusively from a single index associated with the degree of airway narrowing does not speak to the mechanisms nor structural pathology associated with a particular constriction state.

It is the dynamic resistance and elastance (Rl and El) at typical breathing rates that present the mechanical load against which sufficient air must be inspired. Several mechanisms can potentially alter Rl and El. However, the relationship between changes in lung structure and the degree to which specific mechanisms invoke changes in Rl and El is poorly understood. The purpose of our study was to establish the role that the pattern of airway constriction plays in establishing asthmatic conditions. In particular, we hypothesize that severe asthmatic constriction represents a condition of extreme heterogeneous constriction that includes the occurrence of closed or extremely narrowed airways randomly distributed throughout the bronchial tree, which cannot be reopened and/or dilated after a DI. Our hypothesis arose from our recent studies with morphometrically consistent models of the human airways (10-12). To test our hypothesis, we measured the Rl and El from 0.1 to 8 Hz in healthy subjects and in patients with mild, moderate, or severe asthma before and after a DI both before and after standard and modified methacholine challenges. The modified challenge represents that performed by Skloot and colleagues (3) in which DIs and maximum flow-volume maneuvers are prohibited during the challenge.

Constriction Pattern and the Frequency Dependence of Rl and El

We first review how an airway constriction pattern should be reflected in measurements of Rl and El for frequencies surrounding breathing (e.g., 0.1 to 8 Hz). Otis and colleagues (13) and Mead (14) were the first to show that parallel resistance-elastance time-constant heterogeneities and airway wall shunting caused by a large homogeneous constriction could both increase the frequency dependence in dynamic Rl and El. However, it has not been clear how these mechanisms would alter the Rl and El from 0.1 to 8 Hz when occurring in a realistic human airway trees. We have recently extended the work of Horsfield and colleagues (15) and Wiggs and coworkers (16, 17) to implement a human airway tree model inclusive of distinct wall morphometry for healthy subjects and asthmatics (10-12). Our model predicted that measurements of the frequency dependence of Rl and El from 0.1 to 5 Hz can distinctly depend on the pattern of constriction. In Figure 1, we contrast the Rl and El predicted for the baseline healthy lung at FRC to those from two constriction patterns: one with a large mean diameter reduction (50%) applied homogeneously to all peripheral airways (< 2 mm) and another with a small mean diameter reduction (20%) that is highly heterogeneous, so that some airways constrict a lot more than 20%, whereas others constrict much less than 20%. Note that in both cases the tissue properties were typical healthy values (18) and unaltered during constriction.

At baseline, a healthy lung behaves fairly homogeneously mechanically. Between 0.1 and 2 Hz, there is a frequency-dependent decrease in Rl and a small increase in El because of tissue viscoelasticity (18, 19). At higher frequencies, airway inertance becomes dominant such that the calculation of a dynamic El from the imaginary part of lung impedance results in a negative value.

With the large homogeneous constriction, the Rl increases roughly uniformly over the whole frequency range. The El is unaffected at the breathing rates, but increases considerably beginning at 1 Hz and above. This shape occurred because the severe and widespread peripheral constriction invokes central airway wall shunting (14). We know this because if we were to make all the airways rigid in the model, the frequency-dependent increase in El is abolished. A frequency-dependent increase in El associated with airway wall shunting requires a rather substantial constriction throughout the entire periphery (> 40%).

Compared with the homogeneous case, the shapes of the Rl and the El response with heterogeneous constriction are distinct. With this heterogeneous case, there was a lower mean constriction, but a small percentage of peripheral airways became highly constricted or closed. The increase in Raw (i.e., the Rl at 5 to 8 Hz) is smaller, but the increase in Rl and El near typical breathing rates (< 0.5 Hz) are much larger. Two new mechanisms have come into play. First, this form of extreme heterogeneity in parallel pathway time-constants (13) has caused substantial increases in the frequency dependence of dynamic Rl and El below 2 Hz, so that even by 0.2 Hz, both are quite elevated from baseline. Second, there are now random closures throughout the peripheral airways, and as a result, there is less tissue participating in ventilation causing an increase in the apparent static tissue elastance and El at 0.1 Hz.

In summary, we predict that extreme heterogenous constriction in which there is random closure or near closure will produce a dangerous elevation in Rl and El at typical breathing rates and can do so even without necessarily creating a large change in Raw. Moreover, in principle, this form of heterogeneous constriction would greatly compromise the efficacy of ventilation distribution during normal breathing (10). The task now is to identify whether experimental data support these predictions.

Experimental Methods

In seven healthy subjects and in 12 asthmatics, we measured Rl and El from 0.1 to 8 Hz before and after a DI at baseline and during standard and modified methacholine (Mch) challenges. The standard and modified challenges were performed after at least 2 d of rest, but not more than 5 d of rest. After each bronchial challenge, at the maximum dose of Mch, the measurements were repeated before and after a DI. Finally, data were taken after inhalation of albuterol. Subjects ranged from 18 to 35 yr of age and their baseline pulmonary function is summarized in Table 1. With one exception (Subject 1), all of the asthmatics were hyperreactive. All asthmatics were receiving daily doses of albuterol. Four of the asthmatics were also receiving additional medications.


Subject No.Age (yr)SexHeight (cm)Weight (kg)FEV1(% Pred )FEV1/FVC (%)PC20Mch (mg/ml )Medications
Asthma Subjects
  119F165 57 8682> 25Albuterol
  223F168 63 4246N/AAlbuterol
  321F155 51 95961.14Albuterol
  421F165 57 90800.14Albuterol
  518F160 63 79630.008Albuterol, Salmetrol,  Fluticasone
  619F160 50 57570.003Albuterol, Leukotriene  inhibitor, Salmetrol
  721M170 75106861Albuterol
  819M188 91 91812.02Albuterol
  935M180 84 85730.78Albuterol, Salmetrol,  Fluticasone
 1020F157 73 76870.006Albuterol
 1119F162 52 93910.3Albuterol
 1218M178 70 73550.07Albuterol, Fluticasone
Mean21168 65 8175
SE 5 10 13 1816
Healthy Subjects
 1323M173 6810388> 25
 1421M18811810388> 25
 1523M180 7311086> 25
 1621M180 8410980> 25
 1721F165 6311393> 25
 1841M175 79103885.18
 1921M183 77 9982> 25
Mean24178 8010686
SE 7  8 18  5 4

To obtain Rl and El data we used the Optimal Ventilation Waveform (OVW) approach described by us in detail previously (19, 20). Briefly, the OVW provides a multisinusoidal flow signal to the subject in which the amplitudes and the phases of the sinusoids are optimized to maximize tidal volume while minimizing peak airway pressures. The resulting flow simultaneously ventilates the subject. Two OVWs were designed with periods of 12.8 or 25.6 s, respectively. Although both OVW contained seven frequencies between 0.15625 and 8 Hz, they differ as to the primary ventilation rate sensed by the subject (21). The one chosen was that which was most comfortable for a particular subject. The sampled flow and transpulmonary pressure (Ptp) signals were then subjected to Fourier analysis to estimate the Rl and El at each frequency in the OVW waveform (20).


A standard Mch challenge was performed first as follows. Prior to placement of the esophageal balloon, baseline values of FEV1 were obtained. Then we measured the quasi-static pressure-volume deflation curve by having the subject inspire to TLC and exhale slowly against a repetitively closed shutter at the airway opening down to residual volume. Transpulmonary pressure and flow were measured during the exhalation. Subjects relaxed against each closed shutter occurrence, and the shutter remained closed for several seconds each time. Volume was determined by digital integration of flow. This was followed by a 10- to 20-min training period, allowing the subjects to relax their respiratory muscles and to allow the OVW system to “breathe” for them. During measurements, Ptp was continuously monitored on an oscilloscope. This signal becomes highly erratic if breathing efforts are made by the subject during application of the OVW. Each subject relaxed to FRC, placed his mouth around the OVW mouthpiece, and firmly supported the cheeks. The motor was started such that the first OVW inspiration began from FRC. The OVWs were presented for 40 to 60 s. Several OVW measurements were made during this maneuver.

The subjects then were instructed to perform a DI to TLC and relax back to FRC. We immediately repeated the OVW to estimate Rl and El from 0.1 to 8 Hz. This provided the DI response at baseline. After this, subjects were administered increasing concentrations of aerosolized Mch solution generated by a nebulizer system (Pulmonary Data Service, Inc.). Three minutes after each Mch administration, FEV1 was measured via a maximal expiratory flow maneuver followed by an OVW measurement. The Mch administrations were continued until a dose of 25 mg/ml was reached or until a FEV1 of 80% baseline value was reached. From these data we calculated the PC20 according to ATS guidelines (22). After the maximum dose, subjects were instructed to take a DI to TLC and relax back to FRC, and the OVW measurements were repeated. Finally, the OVW measurements were made after aerosolized administration of albuterol.

On a separate day, the subjects performed a modified Mch challenge for which they were instructed to avoid a DI at all times. After each dose of Mch, the OVW measurements were obtained up to the PC20 dose established during the standard challenge or a maximum of 25 mg/ml. The OVW measurements were made again after albuterol.

Data Analysis

Quasi-static elastance (Est) was calculated for each subject from pressure-volume curves using the pressure range corresponding to OVW ventilation. For each of the three subject groups, we evaluated the dynamic data in the context of the constriction patterns implied by the frequency dependent features of Rl and El of Figure 1. Specifically, we compared four key features of the data: Rl at the lowest frequency (Rlow) and the highest frequency (Rhigh); and El at the lowest and the highest frequency (Elow and Ehigh). These low frequency features should be sensitive to heterogeneities (Rlow) and closure (Elow), whereas the high frequency features should be sensitive to the mean constriction level (e.g., Rhigh represents mean airway resistance) and airway wall shunting (Ehigh). These features were determined at baseline, baseline-post-DI, postchallenge, postchallenge-post-DI, and after albuterol on both the standard and the modified challenge days. Paired t tests were performed to establish significant differences from baseline to post-DI, baseline to postchallenge, and postchallenge to postchallenge-post-DI.

Classification of Subjects

Subjects demographic, spirometric, and PC20 data are shown in Table 1. On the basis of these data it is difficult to unambiguously classify the asthmatics subjects. Some (e.g., Subjects 4, 10, and 11) had FEV1/FVC > 80% but were highly reactive (PC20 < 1.0), whereas others had FEV1/FVC near or less than 80% but were less reactive (e.g., Subjects 1 and 9). Subjects 2, 5, 6, and 10 presented at baseline with the most acute hyperreactivity (PC20 < 0.01, except for Subject 2 who was so severely constricted at baseline that a challenge could not be performed) as well as distinctly elevated Rl and El data compared with other asthmatics (Figure 2). On the basis of the impedance and PC20 data, we chose to refer to these four subjects as presenting with “severe” baseline asthmatic constriction and all other asthmatics as mild to moderate. (Note that although our criteria for labeling the patients with severe asthma were quantitative, they are not meant to infer a clinical classification, i.e., they were not fully consistent with the NAEPP and GINA guidelines.)

Baseline Data and DI Response

The data from all subjects are shown in Figure 2. Most data from subjects with mild-to-moderate asthma at baseline were qualitatively similar to the healthy data, although, in general, they showed evidence of slightly elevated Rl at all frequencies and an El that was elevated primarily at higher frequencies (> 2 Hz). The data from the four subjects labeled as severe asthmatics were quite distinct. The Rl was highly elevated, particularly at low frequencies with a large frequency-dependent drop from 0.1 to 2 Hz. The El was nearly double that of other subjects even at 0.1 Hz, and, unlike subjects with mild-to-moderate asthma, the frequency dependent increase in El began immediately and continued through our 8-Hz range.

The key spectral features for all three groups at baseline are compared in Figure 3, whereas Figure 4 tracks the averages within each group throughout the protocol. Also shown in Figure 3 are the Est values. Specific examples for the three types of subject data (healthy, mild-to-moderate asthmatic, and severe asthmatic) before and after a DI at baseline are provided in Figure 5A.

From Figure 3, it is evident that although Elow (i.e., El at 0.15 Hz) progressively rises with more severe asthma, Est is essentially identical for all three types of subjects. Thus, asthma does not cause a stiffening of lung tissue, and the rise in dynamic El even by 0.1 Hz is due to an airway phenomenon.

After a DI, the healthy subject in Figure 5A showed virtually no change in El at any frequency, no change in Rl at higher frequencies, and a small decrease in Rl below 1 Hz. Of the 14 measurements at baseline (seven healthy subjects, each on 2 d), eight showed this pattern and six showed no change in Rl even at lower frequencies after a DI. The decrease in Rlow after a DI over the whole sample of healthy subjects was not significant (Figure 4). Compared with healthy subjects, subjects with mild-to-moderate asthma present at baseline with elevated Rl at all frequencies (e.g., the mean Rlow and Rhigh were 4.0 ± 1.6 and 2.8 ± 0.1 compared with 3.4 ± 1.3 and 1.8 ± 0.4, respectively, for the healthy (Figure 3)), slightly but significantly elevated El < 1 Hz (e.g., Elow 7.4 ± 1.9 versus 6.1 ± 1.4 for the healthy), and then a more noticeable frequency-dependent rise in El primarily above 2 Hz (e.g., Ehigh of −16.4 ± 18.2 for subjects with mild-to-moderate asthma versus −31 ± 13.5 for the healthy). In most of the subjects with mild-to-moderate asthma, there was little evidence of a change in the data post-DI. Indeed, over the entire sample, none of the four key spectral features changed in a statistically significant fashion after a DI at baseline (Figure 4).

The baseline pattern of the patients with mild-to-moderate asthma (e.g., Figure 5A) is consistent with a pattern of widespread milder baseline constriction sufficient in some cases to invoke some airway wall shunting (i.e., the dynamic El increase at higher frequencies). Because El at 0.1 Hz is only slightly elevated compared with that in healthy subjects and there is little if any abnormal frequency-dependent increase in El from 0.1–2 Hz, one could infer that there is some heterogeneity causing Elow > Est by a greater degree than in healthy subjects, but there was not much if any heterogeneous closure in the lung periphery. Moreover, since few if any airways were highly narrowed prior to the DI, after a DI the El at 0.1 Hz is unaffected (i.e., there is communication to the same amount of lung tissue).

Data from the patient with severe asthma is quite distinct in several ways (Figure 5A). At baseline, the Rhigh represents primarily Raw alone (2) and for the subject shown, the Rhigh is more than 300% larger than that of the healthy subject's. Second, between 0.1 and 2 Hz, there was now a very large frequency-dependent drop in Rl and a rapid rise in El. Third, the Elow was nearly double that of healthy lungs. Fourth, between 2 and 5 Hz there was another distinct rapid rise in El. Referring to the simulated data of Figure 1, one could presume that at baseline the patient with severe asthma presents with substantial constriction that is sufficiently widespread throughout the lung to induce airway wall shunting but yet still highly heterogeneous in that there are random airway closures or near closures creating elevated El at 0.1 Hz and a large frequency-dependent increase in dynamic El between 0.1 and 2 Hz. Despite presenting with severe airway constriction, there is very little if any bronchodilatory impact after a DI. The Rl data remain unchanged, whereas the El data are slightly lower, but remain extremely elevated compared with mild-to-moderate asthma. Thus, at baseline, a DI was unable to alleviate the constriction level or homogenize the pattern of constriction.

Standard versus Modified Challenge

Examples of the modified responses pre- and post-DI for the same three subject types from Figure 5A are shown in Figure 5B. In healthy subjects the Rl and El response to a modified challenge was amplified over that of a standard challenge (Figure 4). However, while preventing a DI amplified the response, it did not do so to the level of even subjects with mild-to-moderate asthma during their standard challenge (Figure 4), i.e., typically modified responses for healthy subjects at 25 mg/ml remained much lower than the asthmatic response at their lower PC20 dose during a standard challenge. The discrepancy was even larger when comparing the healthy modified responses to the asthmatic modified responses.

Healthy subjects. For healthy subjects and a standard challenge, there is a noticeable response at the maximum dose of 25 mg/ml. The Rl tended to rise evenly at all frequencies. Among the groups (Figure 4), there was a small increase in Elow (from 6.4 ± 2 to 7.9 ± 2) and a larger increase in Ehigh (from −34 ± 9 to 6.6 ± 47). The modified challenge produced a more amplified response at all frequencies, but primarily at lower frequencies. The Rl and El data postmodified challenge now suggest an enhanced degree of heterogeneous constriction involving airway closures or severe narrowing randomly throughout the peripheral airway tree (e.g., as in Figure 1). The increase in Rlow (p < 0.01), Elow (p < 0.1), and Ehigh (p < 0.001) were significantly greater during the modified challenge (Figure 4). In fact, over the whole group, the increase in Elow and Rlow were significant for the modified challenge (p < 0.004 and p < 0.01, respectively), but not statistically significant for the standard challenge (Figure 4). Also, there is now a greater frequency dependence of El and Rl, from 0.1 to 2 Hz.

Despite the amplified response during a modified challenge, a DI was dilatory to the same degree as it was in a standard challenge (i.e., the spectra all decreased nearly to the same levels regardless of which challenge preceded the DI; e.g. Figure 5B). Also, the Rl below 2 Hz decreased far more than the Rl at 8 Hz. Note that the DI did not reduce the Rl or the Ehigh entirely back to the baseline prechallenged levels (Figure 4). The Elow at 0.1 Hz decreased more than half way back to the prechallenged level (e.g., see insert of Figure 5B). The reduction in Elow indicates a combination of substantially reduced heterogeneity and that many formally closed airways became open, producing communication to more tissue.

In summary, inhibiting the DI in healthy subjects amplifies the response (but not to the levels of asthmatics) primarily by permitting a greater degree of a heterogeneous constriction in which some airways close or nearly close. Nevertheless, the healthy subject can stretch these airways open via a DI, and they remain sufficiently open immediately after the expiration so as to homogenize some of the pattern of constriction. In one subject we repeated the OVW 1 min post-DI and found (not shown) that the Rl and El spectra returned to the pre-DI level. This indicates that the post-DI dilatation is transient and that the random airway closure or near closure will reoccur without bronchodilator intervention.

Subjects with mild-to-moderate asthma. These subjects are hyperreactive during a standard challenge. The response during a modified challenge tended to be amplified compared with the standard challenge, but to a less significant extent than in healthy subjects (Figure 4). In both kinds of challenges, the frequency-dependent increase in El begins immediately. For the subject in Figure 5B, note that the modified response in Rl and El after a dose of only 10 mg/ml substantially exceeds that of the healthy response to 25 mg/ml (Figure 5B). There is a substantial dilatory impact of a DI reducing the Rl and El across all frequencies for both types of challenges (Figures 4 and 5). The post-DI reduction in ELow at 0.1 Hz suggests that some formally closed airways or highly narrowed airways are now open. Unlike that in healthy subjects, in neither challenge did the DI reduce the Rl and El back to baseline. Moreover, compared with the DI poststandard challenge, the DI postmodified challenge is less effective at ablating the impact of the constriction; i.e., the Rl and El post-DI are higher than those found after a DI during a standard challenge (Figure 4), particularly the Rlow, Elow, and Ehigh.

These data suggest that compared with healthy subjects: (1) during a standard or a modified challenge, the response is more heterogeneous and inclusive of more closed airways or nearly closed at a lower dose; (2) comparing the magnitude of the response, especially during a modified challenge, to previous simulations with morphometric models that had challenged only the (< 2 mm) peripheral airways (e.g., Figure 1), it is likely that the constriction and closures are occurring in airways even larger than 2 mm in diameter; and (3) prohibiting a DI during the challenge makes it more difficult to ablate the extreme constriction induced by invoking a DI postchallenge. Nevertheless, administering a bronchodilator does completely ablate it (Figure 4).

Subjects with severe asthma. Because of their baseline conditions, we could perform standard challenges in only three of the four patients with severe asthma and a modified challenge in only one. The patients with severe asthma responded greatly at the lowest dose of 0.01 mg/ml resulting in an amplification of all features of the already abnormal baseline condition (Figures 4 and 5B). When performing a DI postchallenge, there is either no significant dilatory impact or, in the case of this subject, on the standard challenge day evidence of further bronchoconstriction. Finally, while albuterol significantly reduced the Rl and El, perhaps even below baseline conditions, the values postalbuterol were still well above the levels of typical healthy values (Figure 4). Thus, the albuterol could remove the impact of the 0.01 mg/ml dose, but not all the residual constrictive conditions that existed at baseline. This is consistent with a challenge permitting the occurrence of increasing amounts of closed and/or extremely narrowed airways higher in the airway tree that are not dilating during a DI. Thus, the inhaled albuterol is likely not to be delivered to a substantial fraction of constricted smooth muscle and its impact is less.

Many studies have addressed how volume history influences the mechanical state of constriction and reactivity in asthma (3, 6, 8, 23-27). These studies have relied on forced-expiratory indices or values of Rl and/or Raw at a single frequency. They have shown that the response to a DI after induced constriction in asymptomatic asthmatics is different from that in spontaneous asthma. Also, the more severe the baseline spontaneous constriction is, the less effective was a DI at producing bronchodilation (26, 28). Although our data are consistent with these studies, they also provide deeper insight on the relationship between constriction condition and the severity of asthma.

Using partial flow-volume loops Skloot and colleagues (3) and Moore and coworkers (6) indicated that with the modified protocol, healthy subjects' airway hyperresponsiveness was closer to that of asthmatics. They concluded that asthmatics have a diminished capacity to modulate smooth muscle length. Fredberg and colleagues (28, 29) speculated that the lack of muscle length modulation would facilitate the tendency of the muscle to “latch” and its subsequent ability to break out of a latch state via a deep inspiration. Conversely, dynamic modulation may act as a potent regulator of steady-state airway constriction (2, 3, 28, 30). The implication is that in healthy subjects with no preexisting inflammation, one must prevent airway smooth muscle stretching in order to mimic this hyperreactive behavior (i.e., prohibit DIs).

In our study, we focused on evaluating the pattern of airway constriction during induced and spontaneous asthma conditions. Our previous modeling studies demonstrated that this pattern is an important determinant of the functional impact of asthma on the work of breathing. The pattern arises from the distribution of a bronchoprovocation in concert with the distribution of wall inflammation and thickening (11). To our knowledge this is the first study to probe (1) how a prohibition of DIs during bronchoprovocation influences the pattern of constriction and (2) how a DI alters the level and pattern of constriction during spontaneous compared with induced constriction in healthy versus asthmatic lungs. Our primary findings are that induced airway constriction or constriction in patients with more severe baseline asthma is substantially heterogeneous and inclusive of closed or nearly closed airways, that avoiding a DI increases the heterogeneity of response in normal persons, but not to the levels of asthmatic hyperresponsiveness, and that in healthy subjects a DI is more effective at reversing the mean level and the heterogeneity of induced constriction than in asthmatics.

Implications of Rl and El Data

By measuring Rl and El spectra, we provide data that (1) directly represent the mechanical load against which pressures must be generated for breathing and (2) provide a window on the mean level and pattern of airway constriction. Comparing static with dynamic El (Figure 3) showed that subjects with mild-to-moderate asthma have a small but significant elevation in baseline El even by 0.15 Hz and that the elevation is much larger for those with severe asthma. Nevertheless, these elevations are not due to stiffening of lung tissue, but rather are a consequence of heterogeneous airway constriction because the Est was the same for all groups.

Does preventing a DI in healthy subjects mimic the hyperreactive constriction response of asthmatics? Qualitatively, the answer is yes, in that preventing a DI during a challenge amplified the low frequency response of Rl and El in a manner consistent with excessive heterogeneity inclusive of randomly distributed severe airway narrowing. Asthmatics show similar low frequency changes even without preventing a DI. Nonetheless, there are two substantial differences between healthy and asthmatic subjects. First, the Rl and El response of asthmatics at their PC20 dose during a standard challenge far exceeds that seen in healthy subjects (Figures 4 and 5). Consistent with the findings of Burns and Gibson (8) and Brusasco and colleagues (9), one cannot make a healthy subject achieve the degree of airway hyperresponsiveness seen in asthmatics simply by prohibiting DIs during a challenge. Second, in healthy subjects, the impact of a DI poststandard or postmodified challenge was bronchodilatory and returned the Rl and El spectra nearly completely back to their prechallenged level. In contrast, subjects with mild-to-moderate asthma showed a substantially diminished DI response, especially for the modified challenge, and those with severe asthma showed no response to a DI with either challenge, or even had a paradoxical bronchoconstriction after a DI.

The above leads to two additional questions. (1) Why are asthmatics who present with mild-to-moderate asthmatic conditions at baseline hyperreactive? (2) Why are asthmatics with severe baseline constriction showing no bronchodilatory impact of a DI before or after challenge? Consider first the potential role of airway wall-thickening and smooth muscle biological state. We speculate that preventing a DI might promote the airway smooth muscle to transition toward a state in which more force would be created during bronchoprovocation. Such a state could be consistent with either the latch hypothesis (28, 29) or a plastic remodeling of the muscle filament lattice (31). Regardless of how it gets there, once provoked the muscle now can constrict more and become stiffer once constricted. Our healthy subjects did constrict more when a DI was prevented, but never to the level of an asthmatic. Combined with the fact that healthy subjects have no standing inflammatory or thickened airway walls, a DI should be more effective at transmitting the forces necessary to reopen and dilate constricted airways, and this is what we found. Alternatively, the asthmatics at baseline are more likely to have some elevations in inflammatory cells (32) and/or thicker walls not only within a given airway but throughout a greater portion of the airway tree, i.e., extending higher up in the tree (33). Such conditions might facilitate less fluctuations in the smooth muscle length during tidal breathing or a DI, which would facilitate the smooth muscle being closer to a latch or remodeled state that allows more reactivity when provoked. If so, the response to Mch would be greater throughout the airway tree. Certainly, the asthmatic responses suggested a substantially more constricted airway smooth muscle than that in healthy subjects. It is likely that in asthmatics the muscle is so stiff that a DI is simply less effective at generating the forces to open and dilate these airways. Extending this scenario, asthmatics presenting with severe spontaneous constriction likely have substantial wall thickening surrounded by highly stiff muscles. Hence, a DI becomes essentially completely ineffective. Indeed, Johns and colleagues (34) have recently provided evidence that asthmatic airways are less distensible than are healthy airways.

If indeed airway closures of near closures are occurring during asthmatic constriction, we should consider the phenomenon of airway reopening during a DI. Generally, the threshold pressure necessary to reopen a closed airway is Pth = Kγ/r, where r is airway radius to which the airway would “pop open” (or the r from which these airways closed during constriction), γ is surface tension, and K is a constant that depends on the local geometry and the airway wall stiffness (35). At baseline, in asthma, the radius would be smaller, and K would be larger because of both increased muscle tension and wall-thickening. Thus, it is possible that during a DI, patients with severe asthma simply cannot generate sufficient transmural pressures to reopen the closed airways. Nevertheless, it is uncertain from our data whether asthmatics have a reduced capacity to dilate their airways or whether dilation occurs, but the reconstriction during the exhalation from TLC is so rapid as to not provide any residual reduction in the measured Rl and El spectra.

There is certainly evidence of airway closure in asthmatics with provocation, especially in the small airways, or with provocation in healthy subjects administered from a lower lung volume (36-39). Indeed, Gunst and colleagues (30, 40) had already reported that closure is possible. This phenomenon has been the subject of increased attention, but not in the context we propose; namely, to produce a dangerous pattern of heterogeneous peripheral constriction. We showed evidence that when DIs are prohibited during induced constriction, airway closures and severe airway narrowing are more likely to occur. Such closure also reduces the volume of lung tissue communicating with the airway opening, so that the effective tissue elastance and resistance, as measured from the mouth, increase. We do not suspect, or have any evidence of, substantial real increases in tissue elastance during asthma. Evidence of acute heterogeneity with closure or near closure occurred in our asthmatics, or when one attempts to suppress dynamic modulation of smooth muscle by preventing a DI (i.e., a modified Mch challenge).


Extreme heterogeneity of constriction represents an important phenotype of the clinical defect during a severe asthmatic condition. Certain forms of heterogeneous constriction can establish excessive elevations in dynamic Rl and El at typical breathing rates. Our data suggest that preventing a DI during bronchoprovocation amplifies the heterogeneity of the constriction response. By implication, inflammation may serve to exacerbate the effects of increased tone by permitting random airway closures or near closures and preventing residual reopening via DI. In fact, if inflammation also inhibits stretching of these airways, such patients are in double jeopardy. First, smooth muscle stimulation now creates random closures or near closures with consequent ineffective ventilation distribution at typical breathing rates (10). Second, DIs by the patient cannot result in a reopening of these airways and consequent improvement of ventilation. In short, the patient becomes locked in this heterogeneous and dysfunctional state. Only bronchodilators acting directly on airway smooth muscle will “unlock” them. Effective treatment protocols should then focus on distributing the bronchodilator to all airway smooth muscle, including those subtended by closed airways.

1. Ingram Jr RH. Site and mechanism of obstruction and hyperresponsiveness in asthma. Am Rev Respir Dis 1987;136:S62–S64.
2. Shen X, Gunst SJ, Tepper RSEffect of tidal volume and frequency on airway responsiveness in mechanically ventilated rabbits. J Appl Physiol83199712021208
3. Skloot G, Permutt S, Togias AAirway hyperresponsiveness in asthma: a problem of limited smooth muscle relaxation and inspiration. J Clin Invest96199523932403
4. Solway J, Fredberg JJPerhaps airway smooth muscle dysfunction contributes to asthmatic bronchial hyperresponsiveness after all. Am J Respir Cell Mol Biol171997144146
5. Mitzner W, Brown RHPotential mechanism of hyperresponsive airways. Am J Respir Crit Care Med161200016191623
6. Moore B, King LGG, D'yachkova Y, Ahmad HR, Pare PMechanism of methacholine dose-response plateaus in normal subjects. Am J Crit Care Med1581998666669
7. Crimi E, Spanevello A, Neri M, Ind PW, Rossi GA, Brusasco VDissociation between airway inflammation and airway hyperresponsiveness in allergic asthma. Am J Respir Crit Med157199849
8. Burns GP, Gibson GJAirway hyperresponsiveness in asthma. Am J Respir Crit Med1581998203206
9. Brusasco V, Crimi E, Barisone G, Spanevello A, Rodarte J, Pellegrino RAirway responsiveness to methacholine: effects of deep inhalations and airway inflammation. J Appl Physiol871999567573
10. Gillis HL, Lutchen KRHow heterogeneous bronchoconstriction affects ventilation and pressure distributions in human lungs: a morphometric model. Ann Biomed Eng2719991422
11. Gillis HL, Lutchen KRAirway remodeling in asthma amplifies heterogeneities in smooth muscle shortening causing hyperresponsiveness. J Appl Physiol86199920012012
12. Lutchen KR, Gillis HThe relation between airway morphometry and lung resistance and elastance during constriction: a modeling study. J Appl Physiol83199711921201
13. Otis AB, McKerrow CB, Bartlett RA, Mead J, McElroy MB, Selverstone NJ, Radford EPMechanical factors in the distribution of ventilation. J Appl Physiol81956427443
14. Mead JContribution of compliance of airways to frequency dependent behavior of the lungs. J Appl Physiol261969670673
15. Horsfield K, Dart G, Olson DE, Cumming GModels of the human bronchial tree. J Appl Physiol311971207217
16. Lambert RK, Wiggs BR, Kuwano K, Hogg JC, Pare PDFunctional significance of increased airway smooth muscle in asthma and COPD. J Appl Physiol74199327712781
17. Wiggs BR, Bosken C, Pare PD, James A, Hogg JCA model of the airway narrowing in asthma and chronic obstructive pulmonary disease. Am Rev Respir Dis145199212511258
18. Kaczka DW, Ingenito E, Suki B, Lutchen KR. Partitioning of airway and lung tissue resistance in humans: effects of bronchoconstriction. 1997; 82:1531–1541.
19. Kaczka DW, Ingenito EP, Israel E, Lutchen KRAirway and lung tissue mechanics in asthma: effects of albuterol. Am J Respir Crit Care Med1591999169178
20. Lutchen KR, Yang K, Kaczka DW, Suki BOptimal ventilator waveform for estimating low frequency mechanical impedance. J Appl Physiol751993478488
21. Lutchen KR, Suki B. Understanding pulmonary mechanics using the forced oscillation technique: emphasis on breathing frequencies. In: Khoo MCK, editor. Bioengineering approaches to pulmonary physiology. New York: Plenum Press; 1996. p. 227–253.
22. American Thoracic Society. Guidelines for methacholine and exercise challenge testing: 1999. Am J Respir Crit Med 2000;161:309–329.
23. Burns CB, Taylor W, Ingram Jr RH. Effects of deep inspiration in asthma: relative airway and parenchymal hysteresis. J Appl Physiol 1985;59:1590– 1596.
24. Fish JE, Ankin MG, Kelly JF, Peterman VIRegulation of bronchomotor tone by lung inflation in asthmatic and nonasthmatic subjects. J Appl Physiol50198110791086
25. Lim TK, Pride N, Ingram REffects of volume history during spontaneous and acutely induced air-flow obstruction in asthma. Am Rev Respir Dis1351987591596
26. Pelligrino R, Wilson O, Jenouri G, Rodarte JRLung mechanics during induced bronchoconstriction. J Appl Physiol811996964975
27. Pliss LB, Ingenito EP, Ingram Jr RH. Responsiveness, inflammation, and effects of deep breaths on obstruction in mild asthma. J Appl Physiol 1989;66:2298–2304.
28. Fredberg JJ, Jones KA, Nathan M, Raboudi S, Prakash YS, Shore SA, Butler JP, Sieck GCFriction in airway smooth muscle: mechanism, latch, and implications in asthma. J Appl Physiol81199627032712
29. Fredberg JJ, Inouye D, Miller B, Nathan M, Jafari S, Raboudi SH, Butler JP, Shore SAirway smooth muscle, tidal stretch, and dynamically determined contractile state. Am J Respir Crit Care Med156199717521759
30. Gunst SJ, Stropp JQ, Service JMechanical modulation of pressure-volume characteristics of contracted canine airways in vitro. J Appl Physiol68199022232229
31. Chun YS, Pratusevich VR, Ford LESeries-to-parallel transition in the filament lattice of airway smooth muscle. J Appl Physiol892000869876
32. Vignola AM, Chanez P, Campbell AM, Souques F, Lebel B, Enander I, Bousquet JAirway inflammation in mild intermittent and persistent asthma. Am J Respir Crit Care Med1571998403409
33. Kuwano K, Bosken CH, Pare PD, Bai TR, Wiggs BR, Hogg JCSmall airways dimensions in asthma and in chronic obstructive pulmonary disease. Am Rev Respir Dis148199312201225
34. Johns DP, Wilson J, Harding R, Walters EHAirway distensibility in healthy and asthmatic subjects. J Appl Physiol88200014131420
35. Gaver III DP, Samsel RW, Solway J. Effects of surface tension and viscosity on airway reopening. J Appl Physiol 1990;69:74–85.
36. Ding DJ, Martin JG, Macklem PTEffects of lung volume on maximal methacholine induced bronchoconstriction in normal human lungs. J Appl Physiol62198713241330
37. Jophannes C, Beekman AJ, Bel E, Sterk PFRecurrent exacerbations in severe asthma are associated with enhanced airway closure during stable episodes. Am J Respir Crit Care Med161200019021906
38. King G, Eberl S, Salome M, Meikle SR, Woolcock AJAirway closure measured by Technegas bolus and SPECT. Am J Respir Crit Med1551997682688
39. Macklem PTThe physiology of small airways. Am J Respir Crit Care Med1571998S181183
40. Gunst SJ, Warner DO, Wilson T, Hyatt REParenchymal interdependence and airway response to methacholine. J Appl Physiol65198824902497
Correspondence and requests for reprints should be addressed to Kenneth R. Lutchen, Ph.D., Department of Biomedical Engineering, Boston University, 44 Cummington St., Boston, MA 02215.

Supported by: National Institutes of Health 62269 and National Science Foundation-BioEngineering Division.


No related items
American Journal of Respiratory and Critical Care Medicine

Click to see any corrections or updates and to confirm this is the authentic version of record