INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Exercise-induced bronchospasm (EIB) is defined as a condition in which vigorous physical activity triggers acute airway narrowing in individuals with increased airway responsiveness. EIB is seen more frequently in children and young adults, probably because of their high level of physical activity. The prevalence of exercise-induced symptoms in asthmatic individuals has been reported to vary from 40% to 90% (1).

Many factors determine the appearance and severity of EIB: the type of exercise and environmental conditions with which it occurs are the most commonly cited. All types of physical exertion can produce airway obstruction in patients with severe airway hyperresponsiveness (AHR), and fluctuations in temperature and humidity affect its intensity (5, 6).

It has been suggested that the expression of EIB depends on the underlying airway responsiveness (5, 7). Recently, it was suggested that exercise challenge could be used as a standardized epidemiologic tool for investigating the prevalence and mechanisms of asthma (8). Several groups have studied the effects of exposure to antigens, air pollutants, and respiratory virus infection on EIB (7, 9). Exposure to these agents can increase bronchial responsiveness and cause airway obstruction following minimal exertion (5). There have also been some studies of the effects of asthma severity on airway responsiveness to exercise (8, 12). However, in most of these studies, children with low baseline values of FEV₁ and/or severe forms of asthma were excluded.

The purpose of the present study was to investigate the effects of asthma severity on the prevalence and intensity of EIB in children. Our hypothesis was that airway responsiveness to exercise would be quantitatively and qualitatively different in children with different grades of asthma.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The study protocol was approved by the institutional ethics committee of the School of Medicine of the University of São Paulo. Informed consent was obtained from the children and their parents. One hundred and sixty four children (ages 7 to 17 yr) with a physician's diagnosis of asthma were recruited by local media advertisement to participate in the study. All children included in the study reported wheezing in the previous 12 mo, and none was receiving inhaled corticosteroids or had received oral corticosteroids in the previous 2 wk. The children were classified as having intermittent (n = 63, mean age: 11 ± 2 yr), mild persistent (n = 30, mean age: 11 ± 2 yr), moderate persistent (n = 40, mean age: 11 ± 2 yr), and severe persistent asthma (n = 31, mean age: 11 ± 2 yr) according to the Global Initiative for Asthma (16) classification. This classification was based on the answers obtained from a questionnaire administered by an interviewer and on lung function tests. None of the subjects had suffered from clinically apparent upper respiratory tract infections or asthma exacerbations in the previous 2 wk. In order not to confound the interpretation of the effects of asthma severity on EIB, children receiving inhaled corticosteroids were deliberately not included in the study. After the study, all the subjects were included in our Asthma Treatment Program (17).

The study took place in an air-conditioned room (temperature around 21° C and relative humidity around 45%) in order to standardize factors known to influence bronchial responsiveness to exercise. All exercise challenges were performed during summer, within a 4-wk period from February to March 1996.

After baseline lung function was measured, children exercised for 6 min on a cycle ergometer at 80% of their maximum heart rate (HR) as measured with a Polar Accurex HR monitor (Polar Electro Inc., Port Washington, NY). HR was recorded at 1-min intervals. Nose clips were worn to ensure mouth-breathing. Following the exercise challenge, measurements of lung function were made at 5, 10, and 20 min. Forced expiratory maneuvers were repeated until two measurements of FEV₁ with a maximal difference of 100 ml were obtained, with the larger value used in data analysis. The tests were done with a Koko spirometer and software (Pds Instrumentation, Inc., Louisville, CO), and the predicted values of FEV₁ were based on the data of Polgar and Promadhat (18). Children withheld using -agonists and sodium cromoglycate for 12 h. Exercise response was recorded as the greatest decrease in FEV₁ following exercise expressed as a percentage of the baseline FEV₁ (decrease in FEV₁/baseline FEV₁ × 100). A positive response to exercise was defined as a decrease in FEV₁ of 10% or greater (EIB-positive), since this value represents twice the coefficient of variation (CV) of the FEV₁ measure. FEV₁ was used to record changes in lung function after exercise because it has less intrasubject variability than peak expiratory flow (PEF).

Statistical Analysis

Values of FEV₁ observed in the four groups of children were compared through analyis of variance (ANOVA) on ranks, followed by Dunn's test. To compare the percentage of children with EIB among the four experimental groups, we used a logistic regression (19). A value of p < 0.05 was considered statistically significant.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The exercise challenge was easy to perform and all children were able to complete the test without any problem, even children with severe persistent asthma. Most of the children reached the required HR within the first minutes of cycling.

The prevalence of EIB was calculated separately for each group of asthma-severity-based subject group. The results are shown in Table 1. The prevalence of EIB in children with moderate and severe persistent asthma was significantly greater than in children with intermittent asthma (p < 0.001, by logistic regression). Table 1 also shows the odds ratios (ORs) and 95% confidence intervals (CIs) for EIB in children, with the persistent asthma groups compared with the group with intermittent asthma.

Mean baseline values of FEV₁ (% predicted) for each asthma-severity group are shown in Figure 1. FEV₁ % values were also calculated separately for children with EIB and without EIB. Children with and without EIB presented similar baseline FEV₁ values (p = 0.634). The groups with intermittent and mild persistent asthma showed similar values of FEV₁. The group with moderate persistent asthma showed baseline values of FEV₁ significantly lower than the groups with intermittent and mild persistent asthma, and greater than that of the group with severe persistent asthma (p < 0.001, ANOVA on ranks, followed by Dunn's test).

View larger version (25K): [in this window] [in a new window]   Figure 1.   Box plots showing baseline values of FEV1 (% predicted) for all children subjected to the protocol of exercise challenge, and also for children who exhibited exercise-induced bronchospasm (EIB+) or not (EIB). The lines represent the medians, the boxes represent percentiles 25 and 75, and the bars represent percentiles 10 and 90 (*p < 0.001 compared with intermittent and mild persistent groups; **p < 0.001 compared with the other three groups).

The time course of change in FEV₁ (% predicted) values after exercise observed in the four groups of patients with EIB is shown in Figure 2. For most of the children with EIB, the lowest value of FEV₁ was reached at 5 min after exercise (Figure 3).

View larger version (18K): [in this window] [in a new window]   Figure 2.   Time course of FEV1 (% predicted) after exercise in children who exhibited EIB (decrease in FEV1 of 10% or greater). Values are means ± SEM.

View larger version (28K): [in this window] [in a new window]   Figure 3.   Time after exercise at which the greatest decrease in FEV1 (% predicted) was observed in children who exhibited EIB. Children are categorized according to asthma severity (intermittent, n = 17; mild persistent, n = 12; moderate persistent, n = 28; severe persistent, n = 18).

The greatest decrease in FEV₁ (% baseline) after exercise challenge for each asthma-severity group is shown in Figure 4. Children with intermittent asthma presented smaller changes in FEV₁ than children with moderate and severe persistent asthma (p < 0.001, ANOVA on ranks, followed by Dunn's test). When we evaluated only EIB-positive children, we observed similar results: children with intermittent asthma presented lower changes in FEV₁ than children in the other three groups. There were no significant differences among the three groups of children with persistent asthma. In fact, we did not observe decreases in %FEV₁ of more than 30% in children with intermittent asthma (Figure 5). Exercise response, recorded as the greatest decrease in FEV₁ after exercise and expressed as a percentage of baseline FEV₁ (decrease in FEV₁/ basal FEV₁), for each child with EIB, was then correlated with baseline FEV₁ (Figure 6). We observed no correlation between initial FEV₁ and the magnitude of decline in FEV₁ (r = 0.103, p = 0.634).

View larger version (31K): [in this window] [in a new window]   Figure 4.   Box plots showing the greatest decrease in FEV1 (% baseline) in response to exercise in all children studied and also in children who exhibited EIB (EIB+) and children who did not (EIB). Children are categorized according to asthma severity (*p < 0.001 compared with moderate and severe persistent groups; **p < 0.001 compared with the other three groups).

View larger version (27K): [in this window] [in a new window]   Figure 5.   Greatest decrease in FEV1 (% baseline) induced by exercise challenge in children with asthma who exhibited EIB. Children are categorized according to asthma severity (intermittent, n = 17; mild persistent, n = 12; moderate persistent, n = 28; severe persistent, n = 18).

View larger version (19K): [in this window] [in a new window]   Figure 6.   Relationship between greatest decline in FEV1 (% baseline) in response to exercise and baseline FEV1 (% predicted) in 164 children with asthma. The dashed line separates children positive and negative for EIB.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Exercise is one of many nonpharmacologic stimuli that can produce airway obstruction in asthmatic patients. Although bronchial responsiveness is usually measured with histamine and methacholine challenges, bronchial response to exercise has also been used to assess bronchial responsiveness. Furthermore, exercise challenge differentiates asthma from other chronic lung diseases, since this test is more specific for asthma than is a histamine or methacholine challenge (20). As already observed by other groups (8, 21), we found the exercise challenge easy to perform and without any side effects, even in children with severe asthma.

There is accumulating evidence that airway inflammation plays a central role in asthma. The relationship between asthma symptoms, airway inflammation, and AHR is still a debated topic. Some studies have addressed the relationship between clinical expression of asthma and bronchial hyperresponsiveness (BHR) evaluated by methacholine or histamine challenges in either children or adults (22). It was shown that the severity of BHR tends to increase with asthma severity (22, 23), but there are many patients who do not follow this pattern (24). The overall correlation between the degree of airway responsiveness and asthma severity observed in a group of patients may not apply to individual patients. Josephs and colleagues (25) showed that individual measurements of the provocative dose of methacholine that reduces FEV₁ by 20% (PD₂₀) were not consistently related to concurrent asthma severity, since in several patients exacerbations of asthma occurred in the absence of BHR (PD₂₀ > 12.8 µmol). They concluded that the relationship between nonspecific BHR and asthma is complex, and that this functional abnormality is only one mechanism contributing to the clinical expression of the disease. Pattermore and coworkers (24) reported that the median PD₂₀ for responders to histamine was lower among children with more frequent wheezing, and that although severity of BHR tended to increase with frequency of wheezing, asthma of all grades of severity was found for any given frequency of wheezing. Pattermore and coworkers also found a trend for more children with more frequent wheezing to respond to histamine. On the other hand, Martin-Munoz and associates (27) showed that the intensity of airway responsiveness to methacholine has no predictive value for the severity of pure extrinsic childhood asthma.

Some studies found a significant correlation between responsiveness to methacholine and exercise-induced bronchial responsiveness (21, 28, 29), but others found no correlation (30). Haby and associates (21) found a correlation coefficient between the two tests of 0.65 (p = 0.001). Backer (29) also reported a significant but weak relationship of methacholine responsiveness and exercise-induced bronchial responsiveness. Exercise and histamine or methacholine challenges do not always identify the same individuals (28, 29, 32). Haby and associates (21) found children who responded to one challenge and not to the other, and suggested that the two tests identify different abnormalities of the airways. Mediators other than histamine (such as leukotrienes) are important in bronchial responsiveness to exercise, and non-mast-cell mechanisms may also be involved. Benckhuijsen and coworkers (33) reported that a month's stay in a hypoallergenic environment caused a reduction in BHR to exercise and not to methacholine, suggesting that exercise challenge might be a better indicator of asthmatic airway inflammation.

In our study we found a prevalence of 45.7% for exercise-induced bronchospasm among asthmatic children. This figure is in agreement with the previously reported prevalence of 40 to 90% (1). West and colleagues (12), using dry air to improve the sensitivity of the exercise challenge, reported a prevalence of 57%. When we studied each asthma-severity group separately, we observed that children with more severe asthma responded more frequently to exercise, although we could find no response throughout all grades of asthma severity.

Our observation that asthma severity had an effect on the incidence of exercise-induced bronchospasm confirmed previous observations (14). However, previous studies resulted in different conclusions about the relationship of asthma severity and the intensity of the response to exercise: West and colleagues (12) reported a significant correlation between the percentage change in FEV₁ after exercise and a severity index for children reporting wheezing in the preceding 12 mo. Haby and associates (8) also reported a significant relationship between frequency of wheezing attacks and the percentage decrease in FEV₁ induced by exercise challenge. Ponsonby and coworkers (13) reported that the percentage decrease in FEV₁ induced by exercise increased with the frequency of use of asthma medication in the preceding 12 mo. However, in all these studies, children with baseline values of FEV₁ below 75% predicted were excluded from the exercise test. Linna (15), investigating children with mild or moderate asthma, observed a negative correlation between baseline flow values and the response to exercise. In contrast, Kattan and associates (14) studied children with asthma of a wide spectrum of severity, ranging from an occasional mild attack to severe asthma, and concluded that the intensity of the response to exercise was not related to the severity of asthma. We studied more children with severe asthma and/or low values of FEV₁ than did previous studies. We observed a difference in the intensity of the response to exercise only between children with intermittent asthma and those in the other three asthma- severity groups but not among the three groups with persistent asthma. Since an effect of asthma severity on the intensity of EIB was observed only in studies in which children with more severe asthma were excluded, it is possible that differences in the intensity of the response to exercise exist only when children with mild asthma are compared with children with more severe asthma, and not among children with persistent forms of asthma.

Because the use of a 10% cutoff (or any other value) to define a positive response to exercise is somewhat arbitrary and can preclude the possibility that the response is not an "all-or- nothing" effect, we also compared the four groups of children in our study independently of their response to exercise, and achieved the same results for the changes in FEV₁ induced by the exercise test: children with moderate and severe persistent asthma presented significantly lower values of FEV₁ than did children with intermittent asthma. Another finding that deserves consideration is that the response to exercise was not influenced by preexisting airway narrowing, and that there was no relationship between baseline FEV₁ values and the magnitude of the decline in FEV₁ after exercise (Figure 6).

We conclude that although the prevalence of EIB is greater in children with more severe asthma, the response to exercise can be absent even in children with severe persistent asthma, and that the intensity of the response to exercise is not consistently related to the clinical severity of asthma. In these respects, the airway response to exercise is similar to responses to histamine or methacholine. In addition, the preexercise FEV₁ does not predict the severity of response or the presence of EIB.