Background
Structural alterations in asthma include inflammation, increased airway smooth muscle mass, and increased airway wall thickening [
1]. These are not easily assessed in patients, so clinicians rely on functional measurements such as spirometry and tests of airway hyperresponsiveness to assess the presence and control of asthma. Another characteristic of asthma is higher airway resistance at maximal inspiration compared to nonasthmatics. Jensen and co-workers [
2] used the minimum resistance achieved at maximum inspiration (Rmin) as representing the maximum airway dilation achievable (averaged over the entire lung) by a subject. They showed that Rmin was abnormally high (i.e., less ability to dilate the airway tree) in asthmatic versus nonasthmatic subjects [
2]. Salome and co-workers confirmed the reduced ability of asthmatics to dilate after deep inspiration and also showed that the magnitude of dilation was negatively correlated with re-narrowing in nonasthmatics [
3]. Black and co-workers [
4] showed that respiratory system resistance (R
rs) measured noninvasively by forced oscillation at maximal inspiration represented the same Rmin as in the Jensen study because the chest wall does not contribute to R
rs at maximum inspiration. These studies attributed the reduced dilation seen in asthmatics at maximum inspiration to increased stiffness of airway smooth muscle (ASM), reflecting structural characteristics such as hypertrophy and a more contractile state of ASM that may be associated with airway hyperresponsiveness, which is a defining characteristic of asthma.
The goal of this study was to test the hypothesis that the minimal airway resistance achievable during a maximal inspiration (Rmin) is abnormally elevated in subjects with airway hyperresponsiveness. To test this hypothesis, we measured Rrs in nonasthmatic and asthmatic adults during tidal breathing and at maximal inspiration at baseline, following albuterol-induced bronchodilation, and following methacholine-induced bronchoconstriction. Because airway resistance is related to height, we examined the relationship of Rmin to height in nonasthmatic volunteers so that Rmin could be analyzed as a percent of the predicted value. In addition, we compared Rmin to spirometric indices in terms of their relationship to methacholine airway responsiveness. If Rmin measured by forced oscillation accurately reflects airway hyperresponsiveness and structural abnormalities associated with airflow limitation, it may provide a valuable clinical test to help assess the presence and control of asthma that is easier to perform than spirometry.
Methods
Subjects
Participants were recruited by advertisement. Asthmatic participants (n = 35) had a clinical diagnosis of asthma and were taking inhaled bronchodilator. Nonasthmatic subjects (n = 34) denied any history of respiratory symptoms or diagnoses. Participants in both groups were required to have less than 10 pack-years of tobacco smoking. In a substudy to determine the height dependence of resistance measurements, we recruited 84 additional nonasthmatic participants who denied smoking, occupational exposure to smoke or dust, respiratory symptoms, and any respiratory disease history. All subjects provided informed consent, and this research was conducted in compliance with the Helsinki Declaration. This study was approved by Boston University Medical Center IRB, Protocol H-25546, and Boston University Charles River Campus IRB, File 1765E.
Experimental Protocol
Asthmatic subjects withheld short- and long-acting bronchodilators 6 and 24 hours, respectively, prior to study visits. All subjects attended two test days at least 24 hours apart. On day 1, the forced oscillation system described below was used to measure end-inspiratory Rrs during tidal breathing and Rmin at maximum inspiration. Subjects took six tidal breaths followed by a slow maximum inspiration followed by a passive exhalation and six more tidal breaths. The procedure was repeated. Subjects then performed spirometry. After baseline studies, subjects inhaled two inhalations of albuterol metered-dose inhaler 90 μg/inhalation via spacer. Forced oscillation and spirometry measurements were repeated after 10 minutes.
On day 2, baseline measurements of R
min and spirometry were obtained, followed by methacholine challenge. Methacholine (Provocholine
® , Methapharm, Canada) was administered in the following concentrations: 0.098, 0.195, 0.391, 0.781, 1.563, 3.215, 6.25, 12.5, and 25 mg/ml. Other than this concentration schedule, testing was performed in accordance with current ATS recommendations using the 5-breath dosimeter protocol [
5] using equipment described below. At the conclusion of the challenge (i.e. when a 20% decline in FEV1 occurred or after the final dose of 25 mg/ml, whichever came first), R
min was measured again, and then 2 inhalations of albuterol were administered. Spirometry and R
min measurements were repeated 10 minutes after albuterol administration.
In the sample of nonasthmatics studied to establish the relationship of Rmin to height, only Rmin and height were measured.
Measurement of Rrs
We measured R
rs as previously described [
4]. Briefly, a 12-in diameter subwoofer delivers an 8 Hz oscillation, with amplitude of ± 1 cmH
20, superimposed on spontaneous breathing. Jensen and co-workers [
2] showed that because soft-tissue is viscoelastic, it has a tissue resistance that decreases hyperbolically with frequency and that by 8 Hz the lung tissue resistance is negligible and the chest-wall tissue resistance is at its minimum. A three-way valve allows the subject to breathe fresh air through a high-inertance tube. Flow at the airway opening is measured by a pneumotachograph (4700 Series, Hans Rudolph, Kansas City, MO) connected to a differential pressure transducer (ATD02AS, SCIREQ, Montreal, QC). Pressure at the airway opening is recorded with a differential pressure transducer (ATD5050, SCIREQ). These pressure and flow signals are transmitted through demodulator circuits and then to a 10 Hz low-pass filter (S/N 980987, SCIREQ). The filtered signals are sampled at 40 Hz and stored digitally by LabView (National Instruments, Austin, TX). Pressure and flow data were separately low- and high-passed filtered using Matlab software (Natick, MA) at a cut-off frequency of 4 Hz. The signals were processed using a recursive least squares algorithm, described previously[
2], to estimate R
rs eight times per second. Minimum airway resistance, R
min, was derived as R
rs at maximum inspiration.
The system used in the substudy of nonasthmatic subjects (n = 84) conducted to determine normative predicted values for Rmin differed only in its differential pressure transducers (Model LCRV, CELESCO, Chatsworth, CA) and had a 1% error from the system used in the main study.
Spirometry and methacholine challenge methods
Spirometry measurements were made with an integrated spirometer-dosimeter system (KoKoDigidoser
® spirometer, Ferraris Respiratory, Louisville, CO) using the DeVilbiss 646 nebulizer. Our measured nebulizer output was 8.7 ± 0.8 uL/breath (mean ± SE), very close to that reported in the literature for this equipment[
6]. Predicted values for spirometric indices were based on published regression equations [
7]. Spirometry was performed in accordance with published standards [
8]. For methacholine challenges, interpolation was used to calculate the provocative concentration causing a 20% drop in FEV
1 (PC
20). We also calculated the methacholine dose-response slope [
9] as a two-point slope of a line connecting the first and last point of the dose-response curve, measured in units of % decline from baseline FEV
1 per mg/mL of methacholine, an approach that permits analysis of methacholine responsiveness as a continuous measure even in subjects not experiencing a 20% decline in FEV
1. For logarithmic transformation of dose-response slope prior to graphic display and correlation analysis, the constant 0.1 was first added to deal with zero or slightly negative values.
Data Analysis
Among subjects that denied asthma, those with a PC20 greater than 25 mg/mL were defined as "nonasthmatic methacholine nonresponders." Among subjects that reported asthma, those with a PC20 ≤25 mg/mL were defined as "asthmatic methacholine responders."
A second-order linear regression analysis, using the bisquare method[
10] to account for undue influence of outliers, was performed to derive a prediction equation for R
min based on height, using data from 84 nonasthmatic substudy participants and 26 of the 34 nonasthmatic participants in the full study who had a PC
20 ≥25 mg/ml. Predicted values calculated with this equation were used to derive R
min % predicted = (measured R
min /predicted R
min) * 100.
Statistical comparisons were made using paired or unpaired t-tests, or a Mann-Whitney Rank Sum test if a test of normality or equal variance failed, with a significance level of 0.05. Correlations were examined using the Pearson correlation coefficient. For subjects that did not experience a 20% or greater decline in FEV1 by the highest concentration of 25 mg/mL, we assigned a PC
20 value of 25 mg/mL so that we could calculate a geometric mean for Table
1 (Subject characteristics).
Table 1
Characteristics* of 34 nonasthmatic and 35 asthmatic participants.
Sex | 22 F/12 M | 22 F/13 M |
Age (yr) | 21 ± 2 | 21 ± 3 |
Height (cm) | 170 ± 10 | 168 ± 10 |
Weight (kg) | 65 ± 13 | 66 ± 12 |
FEV1 (% predicted) | 95 ± 10† | 88 ± 11 |
FEV1/FVC (% predicted) | 100 ± 7† | 90 ± 9 |
PC20 (mg/ml) | 19 ± 2† median: 25 | 1.8 ± 5.2 median: 1.3 |
≥ 25 | 26 | 4 |
16 - 24.9 | 1 | 0 |
8 - 15.9 | 4 | 5 |
< 8 | 3 | 26 |
Receiver operator characteristic (ROC) curves to examine the ability of Rmin % predicted and other parameters to predict airway hyperresponsiveness (defined as a PC20 < 16 mg/mL) were created by plotting sensitivity (true positive rate) versus 1-specificity (true negative rate), for each value of the test. The best threshold for any test is that which maximizes sensitivity while minimizing the false positive rate, represented by the left upper most value on the curve. The area under the curve (AUC) represents a measure of test accuracy (AUC of 1.0 indicates perfect prediction; AUC of 0.50 indicates prediction no better than chance) and was calculated via numerical integration.
Discussion
Our goal was to test the hypothesis that the minimal airway resistance achievable during a maximal inspiration (Rmin) is abnormally elevated in subjects with airway hyperresponsiveness. The breathing maneuver required to measure Rmin by the forced oscillation method is less burdensome and less subject to performance-related errors than is spirometry. We observed that the baseline Rmin, as a percent predicted value based on height, identifies people with airway hyperresponsiveness approximately as well as FEF25-75 and slightly better than FEV1.
Previous reports suggested a decreased ability of asthmatic airways to dilate in response to a deep inspiration, a deficiency that was accentuated after bronchial challenge [
2,
11] Our measurements in a larger sample of subjects agree with these previous observations. At baseline, R
min, an inverse measure of airway caliber, was significantly higher in asthmatics compared to nonasthmatics. Following inhalation of albuterol, subjects with asthma still had higher R
min than nonasthmatics (Figure
3). In fact, asthmatic subjects had a higher mean R
min after albuterol than the nonasthmatic methacholine nonresponder group before albuterol (not shown), indicating that in subjects with asthma, albuterol cannot always dilate airways to levels achievable in nonasthmatic airways. This suggests that either albuterol does not relax the airway smooth muscle of asthmatics to the same extent as nonasthmatics or that the airway walls have become stiff or narrowed by other mechanisms. In that our data on response to albuterol suggest that asthmatics have an approximately similar decline in R
min in response albuterol as nonasthmatics (reduction in R
min % predicted 17 ± 6.3 SE vs. 11 ± 2.2 SE for asthmatics and nonasthmatics, respectively; p = 0.38), this may favor the explanation of residual differences in the airway wall independent of ASM tone. It must be noted that the dose of albuterol administered in our protocol, i.e. 180 ug (two inhalations), is not a maximally bronchodilating dose. When the stiffer asthmatic airway is constricted by methacholine, the inability to dilate with a deep inspiration is exaggerated compared to nonasthmatic participants, the R
min % predicted increasing in response to methacholine by 85 ± 12 SE vs. 38 ± 5.6 SE (p < 0.001) in asthmatics and nonasthmatics, respectively (Figure
3).
There are several factors that influence airway caliber, including airway smooth muscle tone and stiffness, the passive properties of the airway wall (e.g. airway wall thickening), parenchymal tethering and transmural pressure acting to distend the airway. Several of these can be influenced by airway wall remodeling. Direct measurement of airway distensibility in the intact lung (i.e. the relationship between airway caliber and airway distending pressure) is difficult. Recent work by Brown et al. confirms the ability to indirectly assess airway distensibility non-invasively using forced oscillations [
12,
13]. Specifically, distensibility was quantified as the linear slope of respiratory system conductance (1/R
rs) and volume between 75% and 100% of total lung capacity. This slope was decreased in asthmatics and unaffected by reduction of bronchomotor tone with albuterol. Brown et al. concluded that reduced airway distensibility in asthmatics is consistent with structural changes associated with airway wall remodeling and is not reflective of increased airway smooth muscle tone. This is consistent with the data of our study. Another key determinant of the ability to dilate could be lung elastic recoil pressure; past studies have reported a significant loss of recoil in moderate-to-severe though perhaps not mild asthma[
14‐
16]. We did not measure elastic recoil in our study and can only speculate on its role.
Several limitations of our study must be recognized. The sample size was relatively small (n = 69 for the full protocol and n = 84 for the limited testing to establish predicted values for Rmin), the age range was limited to to 18-29 years, and most subjects were Caucasian race. A larger and more diverse sample would permit better evaluation of the potential relationship of R
min to age and race, as well as subgroup analyses. In addition, the asthmatic subjects had mild to moderate disease, so the full spectrum of asthma was not reflected in our sample, and we were not able to assess the correlation of R
min with clinical status. It is possible that there could be important differences in the physiology of milder versus more severe asthma. Finally, the deep inhalations performed during the dosimeter protocol for methacholine challenge have been reported to result in bronchoprotection and falsely negative challenge results among mild asthmatics, compared to the tidal breathing protocol[
17,
18]. It would be of interest to have data on the relationship of R
min to airway responsiveness assessed by both protocols.
Conclusions
Our study reveals that after adjusting for height, Rmin differs between asthmatics and nonasthmatics, predicts methacholine responsiveness, increases with administration of methacholine, and decreases with albuterol. Compared to spirometry, this test requires less patient effort and is easier for a technician or clinic staff member to administer with technically acceptable results. In conjunction with symptom history, Rmin could provide a clinically useful tool for assessing asthma control and monitoring the response to treatment. Longitudinal studies are needed to assess the utility of Rmin as an indicator of asthma control and response to asthma therapy.
Competing interests
The authors declare that they have no competing interests.
Authors' contributions
NTM contributed to study design, acquisition of data, analysis and interpretation of data, and drafting and revising the manuscript. JK contributed to study design, acquisition of data, analysis and interpretation of data, and drafting and revising the manuscript. ASL contributed to acquisition of data, analysis and interpretation of data, and drafting and revising the manuscript. SNS contributed to study design, acquisition of data, analysis and interpretation of data, and drafting and revising the manuscript. GTO contributed to study design, acquisition of data, analysis and interpretation of data, and drafting and revising the manuscript. KRL contributed to study design, acquisition of data, analysis and interpretation of data, and drafting and revising the manuscript. All authors read and approved the final manuscript.