Abstract
Airway responsiveness is known to be partly explained by geometric and anatomic factors. This cross-sectional investigation was undertaken to determine whether FEF25–75/FVC as a surrogate measure of airway size relative to lung size is associated with airway responsiveness to methacholine. Posteroanterior chest radiographs and spirometry were performed on a group of 929 middle aged and older men from an ongoing longitudinal study, the Normative Aging Study, who returned for their regularly scheduled examination between 1984 and 1989. Subjects had a mean age of 60.5 ± 7.7 yr. FEV1, FEF25–75, and FVC were taken from spirometric results and FEF25–75/FVC ratios were obtained. Main bronchus (MB) and tracheal (TR) diameters and lung area (LA) were obtained from chest radiographs, and ratios of MB/LA and TR/LA were calculated for each subject and compared with FEF25–75/ FVC as measures of airway size relative to lung size. In a multiple linear regression model adjusting for age, height, initial FEV1, smoking, eosinophil count, and IgE level, FEF25–75/FVC was significantly related to the degree of methacholine airway responsiveness as measured by Log10 dose response slope ( β = − 0.37, p < 0.001). Controlling for the same variables, both MB/LA ( β = − 149.07, p < 0.001) and TR/LA ( β = − 125.87, p < 0.001) were significant predictors of the degree of bronchial responsiveness in separate regression models; however, their effects were greatly attenuated when FEF25–75/ FVC was present in the same model. Similar results were obtained after excluding subjects with FEV1/ FVC ⩽ 0.70 and subjects who had any smoking history. We conclude that FEF25–75/FVC as a surrogate measure of airway size relative to lung size is significantly associated with airway responsiveness.
Because airway hyperresponsiveness has been shown to be a risk factor for asthma and lung function decline, interest has focused on factors that may influence its occurrence. Various factors such as viral respiratory illnesses (1), occupational exposures (2), atopy, and cigarette smoking (3) have recently been shown to be associated with nonspecific bronchial hyperresponsiveness.
Anatomic factors that impact on airway size are also known to affect airway narrowing (4, 5). However, there is great variability of airway size between persons, and airway size is at best only loosely correlated with lung size (6, 7), leading investigators to speculate about disproportionate but physiologically normal growth between the airways and the lung parenchyma. In 1974, Green and colleagues (8) proposed that inherent differences in airway size could account for much of the intersubject variability of maximal expiratory flow-volume curves and coined the term “dysanaptic growth.” Mead (9) showed that dysanapsis manifests itself by an inverse relationship between the ratio of maximal flow at 50% of vital capacity divided by vital capacity times static recoil pressure of the lung at 50% of VC (V˙max50/VC × Pst[L]50) and VC. More recently, Tager and colleagues (10) investigated the ratio of forced expiratory flow in the midportion of FVC divided by FVC (henceforth, FEF25–75/FVC) in subjects 7 to 29 yr of age as a surrogate measure of dysanapsis in relation to nonspecific bronchial hyperresponsiveness as assessed by cold air. In linear regression models, they found that this ratio was negatively associated with bronchial hyperresponsiveness, particularly in subjects who had reported a recent respiratory illness.
In this analysis, we investigated the relationship between surrogate measures of airway size relative to lung size (“relative airway size”) and airway responsiveness as measured by methacholine challenge testing in a cohort of men in the Normative Aging Study. Three measures of relative airway size were investigated: FEF25–75/FVC and ratios of tracheal width and bronchial width divided by lung area as measured from routine posteroanterior radiographs. Airway responsiveness was treated as both a continuous outcome (dose-response slope) and as a dichotomous outcome (PD20).
The Normative Aging Study is a longitudinal study of aging established by the Veterans Administration in 1961 (11). The initial cohort of study subjects consisted of 2,280 community-dwelling men from the Greater Boston area who were 21 to 80 yr of age at the time of entry into the study between 1961 and 1969. Volunteers were health screened at entry and were excluded if they had any antecedent chronic medical condition, including hypertension, diabetes mellitus, heart disease, cancer, cirrhosis, peptic ulcer disease, gout, asthma (except asthma limited to childhood), chronic bronchitis, or chronic sinusitis. Volunteers were not required to be veterans of the U.S. Armed Forces.
Since entry, volunteers have reported for periodic examinations, each consisting of a uniform medical history and physical examination, along with blood and urine tests, spirometry, chest radiography, and electrocardiography. Beginning in 1984, subjects have also been studied with a detailed respiratory symptom and smoking questionnaire, methacholine challenge test, allergy skin tests, and serum IgE measurements. Participation in this study has been approved by the Human Studies Subcommittee of the Research and Development Committee, Department of Veterans Affairs Outpatient Clinic (Boston, Massachusetts). Written, informed consent was obtained from all subjects.
The present analysis is based on a subgroup of the Normative Aging Study consisting of participants who returned for their regularly scheduled examinations between February 1984 and May 1989. A total of 1,406 men were seen during this interval. Of these, 291 had neither methacholine challenge testing results nor chest radiographs, 133 had chest radiographs but no methacholine challenge testing results, and 53 had methacholine challenge testing results but no chest radiograph. This left data for 929 men that had information on both methacholine challenge testing and chest radiography available for the current analysis.
Information on smoking habits was obtained from standard questionnaires based on the American Thoracic Society DLD-78 questionnaire (12). Current smokers were defined as those men who smoked at least one cigarette per day during the previous year and were still smoking at least 1 mo before the examination. Former smokers were defined as those men who previously smoked at least one cigarette per day for at least 1 yr, but who had stopped smoking more than 1 mo before their examination. Never smokers were defined as those men who had never smoked cigarettes or smoked less than a total of 20 packs during their lifetime. The number of pack-years smoked was calculated from the information in the questionnaires.
Spirometry and methacholine challenge testing were performed as previously reported (13, 14). FEV1, FVC, and FEF25–75 were obtained from spirometry, and the ratio FEF25–75/FVC was calculated for each subject. Subjects underwent a methacholine challenge protocol adapted from that of Chatham and colleagues (15). Saline and methacholine solutions were aerosolized using a DeVilbiss 646 nebulizer attached to a DeVilbiss air compressor (DeVilbiss; Somerset, PA). All inhalations were 6-s inspiratory maneuvers from residual volume to total lung capacity, followed by 2 s of breathholding. Incremental doses of methacholine were inhaled at 5-min intervals according to the following schedule: five inhalations of 0 mg/ml (phenol-buffered saline alone), one inhalation of 1 mg/ml, one inhalation of 5 mg/ml, four inhalations of 5 mg/ml, one inhalation of 25 mg/ml, and four inhalations of 25 mg/ ml. Previous determination of nebulizer output by weight (13) indicated that the methacholine inhalation schedule corresponded to the following cumulative doses of methacholine in micromoles: 0, 0.330, 1.98, 8.58, 16.8, and 49.8. Spirometry was performed 30, 90, and 180 s after each inhalation level. If the first two spirograms at each level were consistent (FEV1 within 5%), then the higher measurement of these two was chosen for analysis. Otherwise, the higher FEV1 measurement from the most consistent pair of acceptable spirograms was used. The test was terminated when a 20% decline in FEV1 from the postsaline value occurred, or at the end of the dose schedule if such a decline did not occur.
Phlebotomy was performed during the subjects' scheduled examination. Absolute eosinophil counts were performed by a trained technician using a hemacytometer after staining with an aliquot of blood with the Unopette reagent system (Becton Dickinson, Rutherford, NJ). Total serum IgE concentration (IU/ml) was determined by paper radioimmunosorbent test (Pharmacia Diagnostics, Piscataway, NJ), and the mean of two determinations was used for analysis.
Standard posteroanterior chest radiographs were obtained at the time of the scheduled examination. Trained research assistants obtained the following measurements in a standardized manner: (1) main bronchus (MB) diameter taken as the mean of the measurements of both right and left mainstem bronchi, 1 cm from the carina; (2) mean of tracheal widths (WTR) measured at 2, 4, and 6 cm above the carina; (3) lung area (LA) estimated by taking the product of the height of the right lung (measured from the inferior border of the second rib to the highest point of the right hemidiaphragm) and the internal chest diameter (measured at the widest portion of the chest, using the internal margins of the ribs). From these chest radiographic measurements, two ratios were calculated as measures of relative airway size: the ratio of the main bronchus diameter to the lung area (MB/LA), and the ratio of the tracheal width to the lung area (WTR/LA).
Bronchial responsiveness was analyzed both as a continuous variable using an estimate of the overall slope of the dose-response relationship from the methacholine challenge tests (14), and as a dichotomous variable (using the PD20FEV1 values). The dose-response slope (DRS) was defined as the decline in FEV1 from the postsaline value (expressed as a proportion of the postsaline value) after the final dose of methacholine inhaled divided by the final cumulative dose inhaled by the subject. This dose-response slope is expressed in units of the percentage decline in FEV1 per micromole methacholine and represents the slope of a line connecting the origin to the last point of the dose-response plot. As previously reported (14), among asthmatic subjects the dose-response slope is nearly perfectly correlated with the dose causing a 20% decline in FEV1 (PD20FEV1). A dose-response slope of 1% decline FEV1 per micromole corresponds to a PD20FEV1 of 20 μmol. This analytic approach allows treatment of responsiveness as a continuous variable without censoring the data of subjects failing to experience a specified degree of response. Bronchial responsiveness was also analyzed as a dichotomous outcome by taking PD20FEV1 ⩽ 8.58 μmol as a positive response and those with PD20FEV1 > 8.58 μmol, including those who did not exhibit a 20% drop in their FEV1, as no response.
Because of their highly skewed distributions, dose-response slope and serum IgE concentrations were logarithmically transformed (Log10) for all analyses. To permit analysis in the log scale, a small constant (0.3) was added to each value of dose-response slope to eliminate zero and slightly negative values. Univariate relationships were explored using correlation procedures and analysis of variance. A baseline multiple linear regression model was constructed for bronchial responsiveness (L10DRS) as the dependent variable, using the variables previously shown to be significant predictors. The three measures of dysanapsis were then added separately to the baseline model. Similar analyses were performed using logistic regression models to predict bronchial hyperresponsiveness as a dichotomous outcome (PD20FEV1). All analyses were performed using the SAS statistical package (SAS Institute Inc., Cary, NC).
Nine hundred twenty-nine men had complete chest radiograph and methacholine challenge information (Table 1). The mean age was 60.5 ± 7.7 yr (range, 41 to 86 yr). Pulmonary function test results revealed a mean percent predicted FEV1 of 94.4% and mean percent predicted FVC of 96.9%. There was an equal proportion of men with positive and negative skin tests. The excluded group had lower percent predicted FEV1 and FVC values and had a greater proportion of current smokers than subjects who had complete information for this analysis. There was no difference in the FEF25–75/FVC ratio between the excluded subjects and the subjects in this current analysis (Table 1). The three surrogate measures of relative airway size were significantly correlated with each other. As expected, the two ratios derived from radiographs, MB/LA and WTR/LA, were highly correlated with each other (Pearson's r = 0.76, p = 0.0001). There was a weak but significant correlation between FEF25–75/FVC and the two radiographic measures (Pearson's r = 0.31, p = 0.0001 for both correlations).
| Subjects in Current Analysis (n = 929) | Excluded Subjects (n = 477 ) | |||
|---|---|---|---|---|
| Age, yr, mean (SD) | 60.5 (7.7) | 61.4 (8.2) | ||
| Range | 41–86 | 40–89 | ||
| FEV1, % pred, mean (SD) | 94.4% (14.8) | 87.7% (19.7) | ||
| FVC, % pred, mean (SD) | 96.9% (13.9) | 91.1% (17.6) | ||
| Eosinophils, cells/mm3 | ||||
| Median (range) | 176 (0–950) | 194 (0–1,707) | ||
| Skin test | ||||
| Positive, n (%) | 438 (47.1%) | 167 (35.0%) | ||
| Negative, n (%) | 446 (48.0%) | 141 (31.5%) | ||
| Smoking | ||||
| Pack-years, median (range) | 11.6 (0–180) | 20.0 (0–208) | ||
| Status | ||||
| Current, n (%) | 117 (12.6%) | 94 (19.7%) | ||
| Former, n (%) | 507 (54.6%) | 259 (54.3%) | ||
| Never, n (%) | 305 (32.8%) | 124 (26.9%) | ||
| Relative airway size measures | ||||
| FEF25–75/FVC, mean (SD) | 0.76 (0.27) | 0.75 (0.31) | ||
| MB/LA, mean (SD) | 0.0021 (0.0003) | 0.0022 (0.0004)* | ||
| WTR/LA, mean (SD) | 0.0027 (0.0004) | 0.0028 (0.0005)* |
In univariate tests, age (p = 0.04), baseline FEV1 (p = 0.0001), eosinophil count (p = 0.0001), L10IgE (p = 0.0001), and pack-years of smoking (p = 0.0001) were all significantly associated with L10DRS. A baseline linear regression model with L10DRS as the dependent variable was created that included age, FEV1, height, eosinophil count, L10IgE, and pack-years of smoking (Model A in Table 2). Although pack-years of smoking was not significantly associated with L10DRS in the multivariable model, it was retained in all succeeding models. Each measure of relative airway size was added to the baseline model, and results are shown in Table 2. All three measures of relative airway size were significantly associated with bronchial responsiveness as measured by L10DRS, after controlling for the variables in the baseline model (Models B, C, and D in Table 2). The baseline model accounted for approximately 21% of the variability in L10DRS (model r2 = 0.209). The addition of FEF25–75/FVC to the baseline model improved the model r2 to 0.260, whereas adding MB/LA and WTR/LA separately to the baseline model improved the model r2 to a smaller degree (model r2 = 0.224 and 0.225, respectively). Thus, FEF25–75/ FVC explains about 5.1% of the variance in L10DRS, whereas each of the two radiographic ratios explain about 1.5%. Although these were modest increases in the explanatory capacity of the model, the effect of FEF25–75/FVC was comparable to the effect of FEV1, which accounted for about 8.2% of the variance of L10DRS. FEV1 had the largest effect on the model r2, and each of the other significant variables in Model B in Table 2 only accounted for approximately 1% of the variance in L10DRS (from 0.8% for L10IgE to 1.2% for height).
| Model A β (SE) | Model B β (SE) | Model C β (SE) | Model D β (SE) | Model E β (SE) | Model F β (SE) | |||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Age, yr | −0.006 (0.002)† | −0.006 (0.002)† | −0.007 (0.002)† | −0.007 (0.002)† | −0.006 (0.002)† | −0.006 (0.002)† | ||||||
| Height, inches | 0.031 (0.005)† | 0.019 (0.005)† | 0.025 (0.005)† | 0.025 (0.005)† | 0.018 (0.005)† | 0.017 (0.005)† | ||||||
| Smoking, pack-years | 0.0002 (0.0005)‡ | −0.0002 (0.0004)‡ | −0.00006 (0.0005)‡ | −0.0002 (0.0005)‡ | −0.0003 (0.0005)‡ | −0.0004 (0.0005)‡ | ||||||
| Eosinophil count, cells/mm3 | 0.0003 (0.00008)† | 0.0003 (0.00008)† | 0.0003 (0.00008)† | 0.0002 (0.00008)† | 0.0003 (0.00008)† | 0.0002 (0.00008)† | ||||||
| L10IgE | 0.056 (0.018)† | 0.055 (0.017)† | 0.059 (0.018)† | 0.059 (0.018)† | 0.057 (0.018)† | 0.057 (0.018)† | ||||||
| FEV1 | −0.321 (0.025)† | −0.256 (0.025)† | −0.335 (0.025)† | −0.336 (0.025)† | −0.272 (0.026)† | −0.272 (0.027)† | ||||||
| FEF25–75/FVC | −0.364 (0.046)† | −0.327 (0.049)† | −0.323 (0.049)† | |||||||||
| MB/LA | −149.07 (36.13)† | −71.44 (37.21)‡ | ||||||||||
| WTR/LA | −125.87 (28.60)† | −65.14 (29.44)§ | ||||||||||
| Model r2 | 0.209 | 0.260 | 0.224 | 0.225 | 0.261 | 0.261 |
When MB/LA and WTR/LA were added to models containing FEF25–75/FVC (Models E and F in Table 2), the parameter estimates of both radiographic measures were greatly attenuated, whereas the parameter estimate for FEF25–75/FVC remained essentially the same. This result indicates that a large proportion of the explanatory power of each of the two radiographic measures is accounted for by FEF25–75/FVC. Moreover, addition of either radiographic measure to the model containing FEF25–75/FVC improved the model r2 only slightly (model r2 = 0.261 for both models compared with a model r2 = 0.260 for the model with FEF25–75/FVC only).
The negative signs of the parameter estimates of the three measures of relative airway size indicate that when the values of the ratios of FEF25–75/FVC, MB/LA, or WTR/LA are smaller (i.e., smaller airway size relative to parenchymal size), the larger is the value of the DRS (i.e., the greater the percent decline in FEV1 per micromole of methacholine), thus the greater the degree of airway hyperresponsiveness. This inverse relationship between the measures of relative airway size and L10DRS is illustrated in Figure 1, where each of the measures of relative airway size is divided into quartiles.
Fig. 1. Relationship of measures of relative airway size to methacholine airway responsiveness, after adjusting for age, height, initial FEV1, pack-years smoking, eosinophil count, and L10IgE in an analysis of covariance.
[More] [Minimize]The analyses above were repeated after removing subjects with spirometry results suggesting overt airway obstruction (FEV1/FVC ⩽ 0.70). One hundred five subjects had a low FEV1/FVC ratio; results of linear regression models on the remaining 824 subjects with FEV1/FVC ratio > 0.70 are shown in Table 3. Although their parameter estimates were slightly smaller than in the previous set of models, the association of the three measures of relative airway size to L10DRS remained highly significant (Models A, B, C, and D in Table 3). In addition, in Models E and F (Table 3), the effects of MB/ LA and WTR/LA become nonsignificant when added to models already containing FEF25–75/FVC. In the analysis of this subset of subjects, FEF25–75/FVC accounted for about 3.7% of the variance in L10DRS. As in the previous set of analyses, FEV1 had the largest effect, accounting for 6.8% of the variance in L10DRS, whereas each of the other variables accounted for less than 1% of the variance.
| Model A β (SE) | Model B β (SE) | Model C β (SE) | Model D β (SE) | Model E β (SE) | Model F β (SE) | |||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Age, yr | −0.003 (0.002)‡ | −0.003 (0.002)‡ | −0.004 (0.002)§ | −0.003 (0.002)§ | −0.003 (0.002)‡ | −0.003 (0.002)‡ | ||||||
| Height, inches | 0.024 (0.005)† | 0.018 (0.005)† | 0.021 (0.005)† | 0.021 (0.005)† | 0.017 (0.005)† | 0.016 (0.005)† | ||||||
| Pack-years smoked | 0.0003 (0.0005)‡ | −0.0006 (0.0005)‡ | −0.0004 (0.0005)‡ | −0.0005 (0.0005)‡ | −0.0006 (0.0005)‡ | −0.0006 (0.0005)‡ | ||||||
| Eosinophil count | 0.0003 (0.00008)† | 0.0003 (0.00008)† | 0.0002 (0.00008)† | 0.0002 (0.00008)† | 0.0002 (0.00008)† | 0.0002 (0.00008)† | ||||||
| L10IgE | 0.044 (0.018)§ | 0.046 (0.017)† | 0.047 (0.018)† | 0.047 (0.018)† | 0.048 (0.018)† | 0.048 (0.018)† | ||||||
| FEV1 | −0.236 (0.026)† | −0.207 (0.026)† | −0.252 (0.026)† | −0.253 (0.026)† | −0.221 (0.027)† | −0.220 (0.027)† | ||||||
| FEF25–75/FVC | −0.278 (0.047)† | −0.256 (0.050)† | −0.255 (0.050)† | |||||||||
| MB/LA | −89.18 (34.93)§ | −40.81 (35.70)‡ | ||||||||||
| WTR/LA | −74.71 (27.75)† | −36.20 (28.32)† | ||||||||||
| Model r2 | 0.132 | 0.169 | 0.140 | 0.141 | 0.169 | 0.169 |
Similarly, the analyses were performed on all subjects who were lifelong nonsmokers (n = 312), and the results were unchanged. For this subset, the parameter estimate (β) associated with FEF25–75/FVC was −0.381 (p < 0.01), and the parameter estimates for the other variables were similar to those in Table 2. For this set of analyses, FEF25–75/FVC accounted for about 6% of the variance of L10DRS when compared with FEV1, which accounted for about 9.4%.
We also investigated methacholine airway responsiveness as a dichotomous outcome. For the total cohort, 112 subjects (12.1%) had a positive response to methacholine challenge testing (defined in Data Analysis). In models controlling for age, FEV1, height, eosinophil count, L10IgE, and pack-years of smoking, subjects who had FEF25–75/FVC ratios in the lowest quartile were more than five times more likely to have had a positive response on methacholine challenge testing than were subjects with ratios in the highest quartile (OR = 5.4, 95% CI = 2.4 to 12.1). Similarly, subjects with MB/LA and WTR/LA in the lowest quartile were more likely to have a positive response to methacholine challenge testing than were subjects with ratios in the highest quartile (OR = 2.7, 95% CI = 1.4 to 5.5 and OR = 3.3, 95% CI = 1.6 to 6.5, respectively).
Results of this analysis on 929 community-dwelling men indicate that relative airway size as measured by the FEF25–75/FVC ratio is significantly associated with the presence of methacholine airway hyperresponsiveness (PD20FEV1) and the degree of airway responsiveness (L10DRS). These results remained significant after controlling for various factors known to be related to airway responsiveness. Similar results were obtained after excluding subjects with FEV1/FVC ratios ⩽ 0.70, and after repeating the analyses on the subset of lifetime nonsmokers.
In population samples, measurements of nonspecific bronchial responsiveness to a variety of stimuli display a unimodal, approximately log-normal distribution, which is skewed toward higher levels of responsiveness (14, 16, 17), and there is overlap between the ranges of responsiveness observed in asthmatics and nonasthmatics, with some asymptomatic persons displaying high degrees of responsiveness (16-18). The pathophysiologic significance of hyperresponsiveness in apparently normal persons is uncertain, but such hyperresponsiveness may potentially be of importance in the pathogenesis of asthma and chronic obstructive pulmonary disease. Indeed, bronchial hyperresponsiveness has been shown in several recent studies to precede the symptoms and clinical recognition of asthma in both children and adults (19-21).
It has been suggested that certain patterns of lung structure predispose to certain diseases. One study found that subjects with asbestos-related pulmonary fibrosis had tracheas that were, on average, wider and shorter, and chest diameters that were narrower than those of control subjects (22). Larger tracheal diameters have also been found to predispose grain handlers to faster than normal rates of lung function decline (23). In contrast to the case for occupational parenchymal lung diseases, smaller airways may predispose non-dust-exposed subjects to have respiratory symptoms such as wheezing and to develop airway hyperresponsiveness. Narrower airways in infancy and early childhood appear to be a risk factor for childhood wheezing illnesses (24, 25). Martinez and colleagues (25) showed that children who had wheezing episodes in the first 6 yr of life had diminished lung function evident shortly after birth and before any lower respiratory tract illnesses had occurred. Likewise, it is established that airway size relative to lung size is larger in girls than in boys before puberty, and it is generally assumed that the sex-related differences in airway size compared with lung size partly explain the different prevalence of respiratory disease in boys and girls (26).
Green and colleagues (8) noted that there was great inter-individual variability in maximal expiratory flow-volume curves that was not significantly diminished even after correction for lung size or static lung recoil. They concluded that airway size could account for much of this variability. They proposed the term “dysanaptic” to describe the disproportionate growth between the airway and the lung parenchyma and hypothesized that this airway-parenchymal dysanapsis may predispose to the development of different types of lung diseases. Mead (9) later showed that dysanapsis manifests itself by an inverse relationship between the ratio of maximal flow at 50% of vital capacity divided by vital capacity times static recoil pressure of the lung at 50% of VC (V˙max50/VC × Pst[L]50) and VC. Mead also noted that women and boys had airways that were smaller relative to lung size than those of men. More recently, Tager and colleagues (10) illustrated that the FEF25–75/FVC ratio was a reasonable estimate for Mead's measure since the correlation between FEF25–75/FVC and FVC in subjects 7 to 29 yr of age showed the same relationship observed in Mead's data. Using the FEF25–75/FVC ratio as a surrogate measure of dysanapsis, they found that this ratio was negatively associated with the presence of bronchial hyperresponsiveness assessed by cold air challenge, particularly in subjects who had reported a recent respiratory illness.
In contrast to the studies cited above that examined children and young nonsmoking adults (8-10), our subjects were much older, and thus, it is not possible to conclude with certainty that the variation in FEF25–75/FVC seen in our subjects is the result of dysanaptic growth of the airways and lung parenchyma alone, particularly in this cross-sectional analysis. In this age group, we recognized that considerable variation in lung elastic recoil and/or acquired abnormalities in airway function may have contributed to the variation in FEF25–75/ FVC, especially in those with a history of smoking. Because we did not have measurements of lung mechanics, we repeated the analyses on the subset of participants who were lifetime nonsmokers and obtained almost identical results. Although this does not fully address this limitation, our results are consistent with previous findings from younger populations. Furthermore, there is evidence that the relationship between airway size and lung size is established early in life and may persist into adulthood. Martin and colleagues (27) followed a cohort of 47 healthy nonsmoking 6- to 27-yr-old subjects and showed that a subject with relatively small (or large) airways (inferred from maximal expiratory flow measurements) for his/her parenchymal size (inferred from lung volumes) maintained this airway-parenchymal relationship during the 18 yr of follow-up.
Aside from the dysanaptic growth between airways and the lung parenchyma, other factors may play a role in the relationship between a low FEF25–75/FVC ratio and airway reactivity. The degree of airway narrowing in response to stimuli depends on the magnitude of airway smooth muscle shortening and on geometric factors such as the proportion of muscle in the circumference and wall thickness (4, 28). Airway wall thickening caused by inflammation can profoundly affect the airway narrowing caused by smooth muscle shortening (28) even in airways that were initially of larger caliber. Presumably, these effects of inflammation will lead to greater constrictive responses in airways of initially relatively smaller caliber.
Our results indicate that relative airway size, whether caused by dysanaptic growth or acquired factors, is a significant determinant of methacholine airway responsiveness among middle-aged and older men. Among the three measures of relative airway size, FEF25–75/FVC was shown to be a better predictor than either of the two radiographic measures. Thus, in models where either radiographic measure was included with FEF25–75/FVC, the effect estimates of both MB/LA and WTR/ LA were greatly attenuated. Because measurements of airway size and lung size from radiographs were one- and two-dimensional, respectively, it is possible that estimating tracheal and bronchial areas by such methods as acoustic reflection (7), or planimetric techniques (29), or by estimating areas and volumes from both posteroanterior and lateral chest films (30), a stronger relationship between radiographic measurements and methacholine airway responsiveness could have been found. Furthermore, although the radiographic indices estimate the size of the more central airways, FEF25–75 is more generally accepted as a measure of the smaller and more peripheral airways, and likely accounts for the poor correlation between the radiographic measures and FEF25–75 /FVC.
To address the possibility that our results were due to subjects with overt airway obstruction, we repeated the analyses after removing subjects with evident preexisting airway obstruction. Although this does not ensure that all subjects with airway obstruction were excluded, in particular those with very mild obstruction, the results we obtained were similar and thus not reliant only on those with overt airway obstruction.
In summary, we have shown that FEF25–75/FVC is significantly associated with methacholine airway responsiveness, independent of baseline FEV1, atopy, and smoking. This association remained significant after excluding subjects with low FEV1/FVC ratios and subjects with a history of current or past smoking. Longitudinal analyses will be needed to determine whether a low FEF25–75/FVC ratio predisposes persons to develop subsequent airway responsiveness.
The writers would like to thank Deborah DeMolles for assistance with computer programming.
Supported by the Cooperative Studies Program/Epidemiology Research Information Center of the Department of Veterans Affairs and by Grants no. HL34645-12 and HL07427 from the National Insitutes of Health.
| 1. | Hegele R. G., Hayashi S., Hogg J. C., Pare P. D.Mechanisms of airway narrowing and hyperresponsiveness in viral respiratory tract infections. Am. J. Respir. Crit. Care Med.151199516591664 |
| 2. | Chan-Yeung M., Vedal S., Kus J., MacLean L., Enarson D., Tse K. S.Symptoms, pulmonary function, and bronchial hyperreactivity in western red cedar workers compared with those in office workers. Am. Rev. Respir. Dis.130198410381041 |
| 3. | O'Connor G. T., Sparrow D., Segal M. R., Weiss S. T.Smoking, atopy, and methacholine airway responsiveness among middle-aged and elderly men. The Normative Aging Study. Am. Rev. Respir. Dis.140198915201526 |
| 4. | Moreno R. H., Hogg J. C., Pare P. D.Mechanics of airway narrowing. Am. Rev. Respir. Dis.133198611711180 |
| 5. | Jongejan R. C., De Jongste J. C., Kerrebijn K. F.The increased responsiveness to inhaled methacholine in asthma: combination of causative factors. Am. Rev. Respir. Dis.143199111941196 |
| 6. | Martin T. R., Castile R. G., Fredberg J. J., Wohl M. E. B., Mead J.Airway size is related to sex but not lung size in normal adults. J. Appl. Physiol.63198720412047 |
| 7. | Brooks L. J., Byard P. J., Helms R. C., Fouke J. M., Strohl K. P.Relationship between lung volume and tracheal area as assessed by acoustic reflection. J. Appl. Physiol.64198810511054 |
| 8. | Green M. M., Mead J., Turner J. M.Variability of maximum expiratory flow-volume curves. J. Appl. Physiol.3719746774 |
| 9. | Mead J.Dysanapsis in normal lungs assessed by the relationship between maximal flow, static recoil, and vital capacity. Am. Rev. Respir. Dis.1211980339342 |
| 10. | Tager I. B., Weiss S. T., Munoz A., Welty C., Speizer F. E.Determinants of response to eucapneic hyperventilation with cold air in a population-based study. Am. Rev. Respir. Dis.1341986502508 |
| 11. | Bell B., Rose C. L., Damon H.The Normative Aging Study: an interdisciplinary and longitudinal study of health and aging. Aging Hum. Dev.31972517 |
| 12. | Ferris, B. G., Jr. 1978. Epidemiology standardization project. Am. Rev. Respir. Dis. 118(6, Pt. 2):1–88. |
| 13. | Sparrow D., O'Connor G., Colton T., Barry C. L., Weiss S. T.The relationship of nonspecific bronchial responsiveness to the occurrence of respiratory symptoms and decreased levels of pulmonary function: the Normative Aging Study. Am. Rev. Respir. Dis.135198712551260 |
| 14. | O'Connor G., Sparrow D., Taylor D., Segal M., Weiss S. T.Analysis of dose-response curves to methacholine: an approach suitable for population studies. Am. Rev. Respir. Dis.136198714121417 |
| 15. | Chatham M., Bleeker E. R., Norman P., Smith P. L., Mason P. A.A screening test for airways reactivity: an abbreviated methacholine inhalation challenge. Chest8219821518 |
| 16. | Weiss S. T., Tager I. B., Weiss J. W., Munoz A., Speizer F. E., Ingram R. H.Airways responsiveness in a population sample of adults and children. Am. Rev. Respir. Dis.1291984898902 |
| 17. | Cockroft D. W., Bersheid B. A., Murdock K. Y.Unimodal distribution of bronchial responsiveness to inhaled histamine in a random human population. Chest831983751754 |
| 18. | Woolcock A. J., Peat J. K., Salome C. M., et al.Prevalence of bronchial hyperresponsiveness and asthma in a rural adult population. Thorax421987361368 |
| 19. | Carey V. J., Weiss S. T., Tager I. B., Leeder S. R., Speizer F. E.Airway responsiveness, wheeze onset, and recurrent asthma episodes in young adolescents: the East Boston Childhood Respiratory Disease Cohort. Am. J. Respir. Crit. Care Med.1531996356361 |
| 20. | Hopp R. J., Townley R. G., Biven R. E., Bewtra A. K., Nair N. M.The presence of airway reactivity before the development of asthma. Am. Rev. Respir. Dis.14119902 |
| 21. | Sparrow D., O'Connor G. T., Basner R. C., Rosner B., Weiss S. T.Predictors of the new onset of wheezing among middle-aged and older men: the Normative Aging Study. Am. Rev. Respir. Dis.1471993367371 |
| 22. | Becklake M. R., Toyota B., Stewart M., Hanson R., Hanley J.Lung structure as a risk factor in adverse pulmonary responses to asbestos exposure. Am. Rev. Respir. Dis.1281983385388 |
| 23. | Vedal S., Enarson D. A., Chan-Yeung M.Airway size and the rate of pulmonary function decline in grain handlers. Am. Rev. Respir. Dis.138198815841588 |
| 24. | Martinez, F. D., W. J. Morgan, A. L. Wright, C. Holberg, L. M. Taussig, and the Group Health Medical Associates. 1991. Initial airway function is a risk factor for recurrent wheezing respiratory illnesses during the first three years of life. Am. Rev. Respir. Dis. 143:312–316; |
| 25. | Martinez F. D., Wright A. L., Taussig L. M., Holberg C. J., Halonen M., Morgan W. J.the Group Health Medical AssociatesAsthma and wheezing in the first six years of life. N. Engl. J. Med.3321995133138 |
| 26. | Merkus P. J. F. M., ten Have-Opbroek A. A. W., Quanjer P. H.Human lung growth: a review. Pediatr. Pulmonol.211996383397 |
| 27. | Martin T. R., Feldman H. A., Fredberg J. J., Castile R. G., Mead J., Wohl M. E. B.Relationship between maximal expiratory flows and lung volumes in growing humans. J. Appl. Physiol.651988822828 |
| 28. | Wiggs B. R., Bosken C., Pare P. D., James A., Hogg J. C.A model of airway narrowing in asthma and in chronic obstructive pulmonary disease. Am. Rev. Respir. Dis.145199212511258 |
| 29. | Dolyniuk M. V., Fahey P. J.Relationship of tracheal size to maximal expiratory airflow and density dependence. J. Appl. Physiol.601986510506 |
| 30. | Collins D. V., Cutillo A. G., Armstrong J. D., Crapo R. O., Kanner R. E., Tocino I., Renzetti A. D.Large airway size, lung size, and maximal expiratory flow in healthy nonsmokers. Am. Rev. Respir. Dis.1341986951955 |
Presented in abstract form at the American Thoracic Society International Conference, New Orleans, LA, May 1996.
Dr. Sparrow is an Associate Research Career Scientist from the VA Medical Research Service.
