INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Airway hyperresponsiveness (AHR) describes the tendency of the bronchi to narrow too much and too easily in response to provocative stimuli (1). The relationship between AHR and respiratory symptoms is not strong; about 50% of subjects with AHR report no respiratory symptoms (1, 2). AHR seems to follow a normal distribution among the general population, and it has been reported that about 20% of subjects with no evidence of asthma or other respiratory disease show mild AHR, although this varies considerably from one study to another (2, 3). The physiopathology and significance of asymptomatic AHR are, however, still uncertain.

Several previous studies have identified a familial history of asthma and atopy (4) and a personal history of allergy (5) as risk factors in the development of AHR. These factors may play a role in promoting the development of airway inflammation, generally considered to be a major mechanism in the pathogenesis of AHR and symptomatic asthma (8). On the other hand, both AHR and atopy have been associated with high total serum IgE levels (12).

Although asymptomatic AHR can be found in nonatopic subjects, it seems to occur more frequently in the presence of atopy. In this regard, many patients with allergic rhinitis have increased airway responsiveness and, as observed in asthma, their airway response may increase upon natural exposure to allergens to which they are sensitized (15).

Previous investigations suggested that individuals with AHR who have absolutely no thoracic symptoms may be in a latent phase of asthma that may become clinically active over the course of time (19). However, to date, both the physiopathology of asymptomatic AHR and the clinical outcome of subjects with this physiologic abnormality remain unclear.

The present study therefore examined the clinical, immunologic, and physiologic characteristics of adult subjects with asymptomatic AHR in comparison with subjects with symptomatic asthma and normoresponsive controls, and focused specifically on changes in these features over a period of 3 yr. We investigated whether asymptomatic AHR is a preliminary stage of asthma, and whether certain conditions or subject characteristics could contribute to the development of asthma in the population with AHR.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Subjects

Ninety subjects aged 14 to 65 yr (mean ± SEM: 31.9 ± 1.4 yr) were recruited after a study of the prevalence of AHR and atopy in families of asthmatic subjects as compared with control families. Within this cohort, a group of subjects denied any past symptoms compatible with asthma (including recurrent cough), but showed AHR to methacholine. Of these, the first consecutive 30 subjects (14 women and 16 men) who agreed to take part in the present study were enrolled, regardless of their atopic status. Ten of these subjects were from asthmatic families and 20 from control families (mean age: 31.6 ± 2.8 yr).

A second group of subjects, with symptoms of asthma (asthmatic group; n = 30; 26 from asthmatic families and four from control families), was matched for the concentration of methacholine provoking a 20% decrease in FEV₁ (PC₂₀), age (mean age: 30.3 ± 2.7 yr), and gender with subjects having asymptomatic AHR. The asthmatic subjects had stable asthma at the time of the study, and required only intermittent inhaled bronchodilator to control their symptoms. In the prior month, they had had no respiratory infection nor any exacerbation of asthma.

A third group, of normoreactive subjects, denied any past symptoms of asthma and had normal airway responsiveness (n = 30; nine from asthmatic families and 21 from control families); these subjects were matched for age (mean age: 32.1 ± 3.0 yr) and gender with the asymptomatic AHR subjects.

All subjects signed informed consent forms, and the study was approved by the Laval Hospital Ethics Committee.

Definitions

Asthma was defined according to the criteria suggested by the American Thoracic Society (ATS) (23). Asymptomatic AHR was defined as a PC₂₀ < 8 mg/ml in the absence of symptoms suggestive of asthma in subjects who never required any asthma medication (24). To avoid including subjects with "borderline" AHR in the normoreactive control group, subjects had to have a PC₂₀ > 20 mg/ml in the absence of any respiratory symptoms. Atopy was defined as the presence of at least one positive response (wheal diameter  3 mm at 10 min) to skin-prick tests with a battery of 26 common airborne allergens (25). Atopic index was calculated as the number of aeroallergens to which the patient showed a positive response, and atopic score as the mean wheal diameter of all allergic responses to allergy skin tests.

Initial Clinical Assessment and Pulmonary Function Tests

First visit. Each subject completed a general questionnaire on respiratory health and family history of asthma and/or atopy. Measurements of expiratory flows were done with a Vitalograph PFT II spirometer (Vitalograph Medical Instrumentation, Lenexa, KS) according to ATS recommendations (26). The best of three FEV curves was used to determined FVC and FEV₁. Bronchodilator response was measured as the increase in FEV₁ at 15 min after a 200-µg dose of inhaled salbutamol. Lung volumes were measured through standard whole-body plethysmography (27). FRC, TLC, and RV were determined for each subject. Peak expiratory flow rates (PEFRs) were measured with a mini-Wright peak-flow meter (Armstrong Medical, Scarborough, ON) in the morning and evening over a period of 2 wk. The best of three repeated measurements was noted on the diary card.

Skin-prick tests were done with a battery of 26 inhalant allergens, which were divided into the six main categories of animal danders, dust, house dust mite, tree pollen, grass pollen, and molds. Serum IgE was measured with enzyme immunofluorometry. Blood eosinophils were counted on a Coulter (Hialeah, FL) STKS.

Second visit. Methacholine inhalation tests were done according to the method described by Juniper and colleagues (28). Briefly, aerosols were generated from a Wright nebulizer (Roxon Medi-Tech, Montréal, Québec) with an output of 0.13 ml/min. After the initial control saline inhalation, increasing doubling concentrations of methacholine, from 0.03 to 128 mg/ml, were inhaled by tidal mouth breathing for 2 min at intervals of 5 min. The response, as the percent decrease in FEV₁, was measured at 30 s and 90 s, and then at 2-min intervals if necessary, to record to the lowest value after each inhalation. The test was stopped when FEV₁ had fallen by  20% or when the maximum concentration of methacholine had been inhaled. The results were expressed as the PC₂₀ of methacholine. During this test, chest symptoms were scored on a modified Borg scale from 0 = nothing to 10 = maximum.

Subjects were reexamined in February or March (in order to avoid the pollen season) at 2 and 3 yr after their baseline evaluation, undergoing the same tests as in the baseline evaluation.

Statistical Analysis

Results were expressed as mean ± SEM values for FEV₁, variability of PEFR, atopic score, and eosinophil counts. Total serum IgE level and PC₂₀ were expressed as the geometric mean ± SEM. For continuous variables, a one-way analysis of variance (ANOVA) was used to compare mean values between groups of subjects. In order to identify significant differences among the three groups (subjects with AHR, subjects with asthma, and normoresponders), we compared means using Scheffe's method. The analysis of paired results (at 1 and 3 yr) was done with a repeated measures design. Categorical variables were analyzed with Fisher's exact test to determine the association between different variables. Significance was accepted at the 95% level.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Clinical Parameters

In the asthmatic, asymptomatic AHR, and normoresponsive control groups, the numbers of subjects with a personal history of any atopic condition were, respectively, 24 (80.0%), 21 (70.0%), and 10 (33.3%, p = 0.0005). More subjects in the asymptomatic AHR group (n = 5, p = 0.01; Table 1) reported having seasonal rhinoconjunctivitis than did subjects in the normoresponsive control group (n = 1). For eczema and urticaria, however, the proportions in the two groups were similar (23.3% and 16.6%, and 26.6% and 26.6%, respectively, both p > 0.05). The number of asymptomatic AHR subjects with a history of respiratory infection of possible viral origin within the 5 yr preceding the study, other than common cold, was significantly higher than in the normoresponsive group (26.67% and 0%, respectively, p = 0.005). The number of smokers was similar in all three groups, and the number of subjects exposed to second-hand smoke was also similar (Table 1).

Physiologic Parameters

Baseline. The subjects with asymptomatic AHR did not differ significantly from normoresponders with regard to FEV₁, perception of symptoms, serum IgE level, or eosinophil count (Table 2A). A significantly greater degree and/or severity of atopy (atopic index: 2.2 ± 0.4 versus 0.6 ± 0.2, p = 0.0003; atopic score: 40.3 ± 8.4 versus 8.4 ± 2.2, p < 0.0001) was found in the asymptomatic AHR group than in the normoresponsive control group (Table 2A). In the asymptomatic AHR group, variability of PEFR was greater than that observed in the normoresponsive control group (5.1 ± 0.5 versus 3.4 ± 0.2, p = 0.002), as was the bronchodilator response (6.3 ± 0.6 versus 3.5 ± 0.4, p = 0.0003; Table 2A).

Subjects with asymptomatic AHR had a higher baseline FEV₁ (% predicted: 104.0 ± 2.6 versus 89.8 ± 2.0, p < 0.0001) and weaker bronchodilator response (% change in FEV₁: 6.3 ± 0.6 versus 9.7 ± 1.1, p = 0.008) than did asthmatic subjects (Table 2A). Subjects with asymptomatic AHR perceived methacholine-induced bronchoconstriction more acutely than did asthmatic subjects (perception score = 3.0 ± 0.1 and 1.2 ± 0.1, respectively, p < 0.0001). In subjects with asymptomatic AHR, the eosinophil count was lower than in the asthmatic group (p = 0.004, Table 2A); the atopic index and atopic score were similar in the two groups (Table 2A).

Sixty-one percent of the subjects included in the study were atopic (33.3% of the control, 70.0% of the asymptomatic AHR group, and 80.0% of the asthmatic group [p = 0.0005]).

Relationship Between AHR and Gender

In this study, 64.4% of the subjects were women. Gender comparisons for the different parameters showed no significant differences except for atopic index and atopic score, both of which were slightly higher in men than in women, with respective values of 2.8 ± 0.5 and 1.9 ± 0.3 (p = 0.09) for atopic index and 49.1 ± 11.2 and 28.7 ± 5.2 (p = 0.06) for atopic score. In subjects with AHR (n = 56, asymptomatic AHR and asthmatic groups), we noted that the PC₂₀ of methacholine was slightly lower in women than in men, although this difference was not statistically significant (3.3 ± 1.2 mg/ml for women and 3.9 ± 1.1 mg/ml for men, p = 0.07).

Relationship Between AHR and Atopy, and Between AHR and Familial History of Asthma

In comparison with nonatopic asthmatic subjects, those with asthma and atopy had a higher degree of airway responsiveness, with respective PC₂₀ values for methacholine of 2.7 ± 1.1 mg/ml and 4.9 ± 1.1 mg/ml, p = 0.01 (Table 3). They also had lower eosinophil counts, and higher serum IgE levels, with respective values of 0.4 ± 0.04 × 10⁹/L and 0.1 ± 0.03 × 10⁹/L (p = 0.01) and 157.8 ± 1.2 µg/L and 7.3 ± 1.3 µg/L (p < 0.0001; Table 2B).

Among subjects with asymptomatic or symptomatic AHR (n = 60), 75% were atopic (Table 2B).

When we examined physiologic parameters according to the presence of asthma in first-degree relatives of the normoresponsive and asymptomatic AHR groups, we found that the subjects with at least one first-degree relative with asthma had a higher prevalence and severity of atopy, as well as higher serum IgE levels (Table 2C).

Two- and 3-yr Follow-up

With regard to the variability of the study parameters over the 3-yr total follow-up period, we observed no increase, and even a slight reduction, of bronchodilator response (Table 2A) in the normoreactive control group.

In the asthmatic group, we observed a reduction of FEV₁ (p = 0.01), bronchodilator response (p = 0.01), and blood eosinophil counts (p = 0.0009). In addition, airway responsiveness (p = 0.001), atopic score (p = 0.001), and mean serum IgE level (p < 0.0001) were mildly increased. First-degree relatives of asthmatic subjects had a more marked increase in airway responsiveness at 3 yr than did relatives of asymptomatic AHR subjects (p = 0.01).

In asymptomatic AHR subjects, FEV₁ was significantly decreased (p < 0.0001); this reduction was strongly associated with atopic status. In fact, over the 3-yr follow-up period, the reduction in FEV₁ in asymptomatic AHR subjects was 5.5 ± 0.7% for subjects with atopy, and 2.4 ± 0.9% for nonatopic hyperresponders (p = 0.01). In comparison with hyperresponders who did not develop asthma symptoms, those who developed such symptoms had greater decreases in FEV₁, with values of 3.6 ± 0.6%, for the asymptomatic group and 9.3 ± 1.3% for those who developed symptomatic asthma (p = 0.0007). In addition, in those subjects who developed asthma symptoms, the mean premethacholine FEV₁ at Year 3 became similar to that observed in the asthmatic group (87.0 ± 1.6% predicted versus 88.5 ± 1.8% predicted, p = 0.94).

In the asymptomatic AHR group, the PC₂₀ methacholine was significantly reduced at 3 yr compared with the values for the control and asthmatic groups, with a change in the number of doubling methacholine concentrations of 0.90 ± 0.61 mg/ml, 0.13 ± 0.04 mg/ml, and 0.02 ± 0.002 mg/ml (p < 0.0001; Figure 1A). The reduction of PC₂₀ methacholine from baseline to Year 3 was more marked in subjects with atopy (Figure 1B) and in subjects with first-degree relatives who had asthma (Figure 1C); this held true for all groups. At 3 yr, four of the asymptomatic AHR subjects had developed asthma symptoms (none of the normoresponsive control group had done so, p = 0.11). These symptoms included wheezing and exercise-induced cough in all four subjects, and wheezing and chest tightness upon exposure to relevant allergens in three of the four. These four subjects also developed a > 15% increase in FEV₁ after bronchodilator therapy (mean: 17.2 ± 2.1%), and a greater variability (increase) in their FEV₁ response to bronchodilator than did the 24 other asymptomatic AHR subjects (5.0 ± 2.1% versus 0.2 ± 0.6%, p = 0.003). The asymptomatic AHR subjects who developed asthma symptoms did not differ significantly from the 24 other asymptomatic AHR subjects with regard to changes in the circadian variation of their PEFR over the course of the 3-yr follow-up (0.44 ± 0.3% and 0.55 ± 0.4%, respectively, p = 0.94). Although the four asymptomatic AHR subjects who developed asthma symptoms had a mean daily PEFR fluctuation of less than 15%, they had occasional fluctuations of more than 15%, indicating the development of variable airflow obstruction (according to most current criteria). All were atopic and from families with asthmatic members. No change in these patients' environment was noted, but each was exposed to an animal, as were other subjects with asymptomatic AHR.

View larger version (28K): [in this window] [in a new window]   Figure 1.   Variability of PC20 methacholine over 3 yr in the different groups (A); intragroup variability according to atopic status (B); and intragroup variability according to presence of at least one first-degree relative with asthma (C  ). *p < 0.05 for difference between groups.

Onset of Asthma Symptoms

Analysis of the relationship between the incidence of newly diagnosed asthma and the severity of AHR (Table 3) showed that the more severe the AHR, the more likely the subjects were to develop symptomatic asthma (p = 0.01). About 40% of the subjects with an initial PC₂₀ < 4 mg/ml developed asthma within the 3-yr follow-up period. Furthermore, about 40% of the subjects with asymptomatic AHR and a familial history of asthma developed asthma within the 3-yr follow-up period, whereas none of the subjects without first-degree relatives with asthma did so (p = 0.01; Table 3).

Among those asymptomatic AHR subjects reporting a history of viral respiratory tract infection within the previous 5 yr, 44.4% developed asthma; in contrast, among the subjects of this group not reporting such infection, none developed asthma (p = 0.006, Table 3).

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

This study shows that subjects with AHR to methacholine, but no current or past symptoms of asthma or other respiratory disease, exhibit a diurnal variability in PEFR and prevalence or degree of atopy greater than those of normoresponder subjects but similar to those of asthmatic subjects. Furthermore, over a period of 3 yr, subjects with asymptomatic AHR tended to have a higher incidence of symptomatic asthma than did normoresponders. All subjects who developed symptomatic asthma were atopic, had a history of respiratory viral infection, and had a first-degree relative with a diagnosis of asthma; they also had higher baseline AHR than those who did not develop asthma. Perception of bronchoconstriction-induced symptoms was not deficient in asymptomatic AHR subjects; in fact, it was even greater than in the asthmatic group.

Our observations suggest that asymptomatic AHR may be an intermediate stage between normality and asthma, although the reason why this AHR does not translate into respiratory symptoms is unclear. Subjects with asymptomatic AHR had features suggesting increased variation in airway caliber, as observed previously (15, 29, 30), although this finding differs from the observations of Power and colleagues, who found no differences in spirometric values between subjects with asymptomatic AHR and normoresponders (31). Furthermore, our asymptomatic AHR subjects had a greater bronchodilator response than that of normal subjects.

Why then, if subjects with asymptomatic AHR have such variability in airflow obstruction, do they have no symptoms? In addition to the possibility that the airway caliber changes in our subjects with asymptomatic AHR were not great enough to induce symptoms, we considered the possibility that the subjects' asymptomatic AHR was due to a reduction in their ability to recognize airway constriction. We did note, however, that the mean perception scores of the asymptomatic AHR subjects and normoresponders were similar, and were higher than those of the asthmatic group. This observation suggests that in general, subjects with asymptomatic AHR have no significant impairment in their ability to recognize bronchoconstriction-related symptoms. This agrees with the data reported by Pin and coworkers (32) and Gibson and colleagues (29). Conversely, Brand and associates (33) reported that subjects with asymptomatic AHR were less likely to report symptoms during a methacholine challenge. These contradictory results may be explained by several factors. For example, in Brand's study, the challenge was stopped when FEV₁ fell by 10% or more, a fact that accounts for the smaller decreases in their subjects' expiratory flows. Another explanation for inconsistent findings in studies of subjects with asymptomatic AHR could be that the definitions and methods used for symptom evaluation are not standardized.

Additionally, our subjects had not previously experienced the symptoms they developed during the challenges, making it unlikely that the previous absence of any report of asthma symptoms was due to the failure to recognize these symptoms as being asthma symptoms.

Our observations show that subjects with asymptomatic AHR are at greater risk for developing asthma than are normoresponders. It would be worthwhile to determine which individuals among those with asymptomatic AHR are at greater risk for developing symptomatic asthma within the succeeding few years. Previous observations in children have led to useful information on the prevalence and determinants of the development of symptomatic asthma in these subjects. Jones (21) reported that among a cohort of children with asymptomatic AHR, eight of 60 subjects developed asthma symptoms by the end of a 5-yr follow-up period. On the other hand, Burrows and colleagues found a tendency for airway responsiveness to decrease with age in a birth cohort of children aged 9 to 15 yr (22). This change in AHR was closely related to the skin-test index, a higher allergy skin-test reactivity being associated with a lesser reduction in airway responsiveness. Differences between these other studies and ours probably reflect the effects of different influences on children than in adults, such as endocrine or other types of factors and changes in airway caliber with growth.

One of the major findings of the present study was that all of the hyperresponsive asymptomatic subjects who developed symptomatic asthma were atopic and had a first-degree relative with asthma. This strongly suggests that genetic influences, in conjunction with atopy, play a role in the development of AHR and symptomatic asthma. We also found that in comparison with normal subjects, those with asymptomatic AHR had higher atopic indexes and atopic scores. In our study, two thirds of the subjects with AHR were atopic, as compared with one third of the control group. This is consistent with previous studies showing that atopic nonasthmatic subjects have a higher prevalence of methacholine-induced AHR than do nonatopic subjects (15). In addition, our results showed an increase in the prevalence of atopy in subjects with a family history of asthma.

Zhong and associates (34) reported that among a group of students without respiratory symptoms, those with more severe AHR developed asthma in the 2 yr following initial study, and that 80% of these subjects had a history of early respiratory illness. Hegele and coworkers (35) reported that viral respiratory tract infections are associated with an acute increase in airway responsiveness in both normal subjects and patients with asthma. Our results are in keeping with those observations. Asymptomatic AHR subjects who developed asthma had more marked AHR, and all of them had reported an episode of respiratory viral infection within the previous 5 yr (Table 3). The mechanisms by which respiratory viral infections can increase airway responsiveness remain unclear. The viral infection may induce airway epithelial damage and/ or airway inflammation, but it may also trigger an airway repair process, with these phenomena resulting in acute or sometimes persistent changes in airway function.

We should also point out that the asymptomatic hyperresponders and asthmatic subjects in our study had a similar degree of AHR, but that asthmatic subjects had a lower FEV₁. Furthermore, those subjects with asymptomatic AHR who developed asthma experienced a reduction in FEV₁ to the level seen in asthma. This suggests that clinical asthma may not be related only to the presence of hyperresponsive airways, but also to the combined effect of airflow obstruction and hyperresponsiveness, both of which are secondary to the underlying airway inflammatory process or its structural consequences.

In conclusion, asymptomatic AHR appears to be an intermediate stage between normality and symptomatic asthma, and even as adults, individuals with AHR seem to be at greater risk of developing symptomatic asthma than the general population. Subjects in the asymptomatic AHR subgroup who are at greater risk of developing symptomatic asthma are atopic and have a first-degree relative with asthma, are currently exposed to indoor allergens such as animal danders, and more frequently report a history of respiratory viral infection in previous years. Preventive measures, such as avoiding allergen exposure, and possibly other measures to reduce airway inflammation, should therefore be emphasized, particularly in this subgroup, in order to minimize the chance of development of chronic symptomatic asthma. On the other hand, the mechanisms by which the asymptomatic phase of AHR becomes overtly symptomatic asthma remain to be documented.