INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Gastroesophageal reflux is common in adult asthma patients and is a potential trigger of asthma (1). In a questionnaire survey, Field and coworkers reported that 77% of asthma patients experienced heartburn, 55% complained of regurgitation, and in the week prior to completing the questionnaire, 41% of the asthma patients reported reflux-associated respiratory symptoms (2). Twenty-four hour esophageal pH testing accurately diagnoses gastroesophageal reflux with a sensitivity and specificity of approximately 90% (3). The frequency of false-negative results of 24-h esophageal pH testing ranges between 10% and 25% (4). Evaluating 104 consecutive asthma patients, Sontag and coworkers observed that 82% of asthma patients had abnormal amounts of acid reflux on 24-h esophageal pH testing (5). Identifying gastroesophageal reflux in asthma patients is important because aggressive treatment of gastroesophageal reflux may result in improvement of respiratory symptoms in selected patients (6). A double-blind, placebo-controlled, multicentered trial evaluating asthma outcome with aggressive medical therapy using a proton pump inhibitor has not been reported to date.

Patients with asthma aggravated by gastroesophageal reflux may not have classic reflux symptoms of heartburn or regurgitation, leaving the clinician unaware that gastroesophageal reflux may be a trigger for the asthma. Ambulatory 24-h esophageal pH testing plays a key diagnostic role in asthma patients without reflux symptoms. Using this test, Irwin and coworkers studied a group of difficult-to-control asthma patients and found that gastroesophageal reflux was "clinically silent" in 24% (9). They observed that vigorous treatment of gastroesophageal reflux was helpful in converting difficult-to- control asthma patients into ones who were no longer difficult to control (9). More recently, in a retrospective review of 199 asthma patients undergoing esophageal pH testing, we found that 35 (18%) did not have reflux symptoms, of which 10 (29%) had abnormal esophageal acid contact times. Twenty-four hour esophageal pH testing was well tolerated without untoward effects in these 199 asthma patients (10). The prevalence and severity of gastroesophageal reflux in asthma patients without reflux symptoms has not been carefully studied in a prospective manner. Thus, the aims of this study are to prospectively determine the prevalence and severity of gastroesophageal reflux in stable asthma patients of all severities without reflux symptoms using 24-h esophageal pH testing and to compare this group with asthma patients with reflux symptoms undergoing similar testing.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Subjects

This prospective cohort study was approved by the Human Use Committee at the University of Alabama at Birmingham on March 9, 1995. Subjects were recruited from the outpatient pulmonary clinic at the University of Alabama at Birmingham. Potential subjects were screened over the telephone for the presence of reflux symptoms including heartburn, regurgitation, water brash, dysphagia, and epigastric pain. If they denied symptoms, they were informed about the study and called back after 7 d to see if they were still interested in enrolling, and requestioned as to the presence of esophageal symptoms. Consecutive asthma patients without reflux symptoms who met the entrance criteria and had asthma stability for at least 2 wk, and gave informed consent participated. Stable asthma was defined as stable asthma symptoms and no change in asthma medications during the 2 wk before study entry. The subject population met the American Thoracic Society's definition of asthma, including: a 200-ml and a 12% improvement in FEV₁ with bronchodilators, or a 20% decrease in FEV₁ after methacholine challenge, performed in accordance with the guidelines of the Lung Health Study; subjects were nonsmokers and had no symptoms consistent with chronic bronchitis nor other forms of chronic lung disease (11). Asthma patients without reflux symptoms had rare heartburn or regurgitation (once a month or less), no dysphagia, no history of esophageal, gastric surgery, or scleroderma, and no previous treatment with antireflux medications including H₂ antagonists, proton pump inhibitors, prokinetic agents, or regular use of antacids.

To assess the severity of gastroesophageal reflux in asthma patients with asymptomatic gastroesophageal reflux (Sx +pH), comparisons were made with asthma patients with symptomatic gastroesophageal reflux (+Sx +pH). This symptomatic gastroesophageal reflux group is a previously described cohort group which participated in a reflux treatment trial in which subjects had asthma criteria as previously defined, reflux symptoms including the presence of heartburn and/or regurgitation at least twice monthly, had abnormal 24-h esophageal pH tests, and were not on antireflux medication such as antacids, H₂ antagonists, proton pump inhibitors, or prokinetic agents (7).

Esophageal Manometry and Ambulatory 24-h Esophageal pH Testing

Standardized methods of esophageal manometry and ambulatory 24-h esophageal pH testing were performed on all subjects. After an overnight fast, esophageal manometry was done in the supine position using a round polyvinyl catheter (diameter 4.5 mm) (Andorfer Medical Specialties, Greendale, WI) continuously perfused with distilled water at a rate of 0.5 ml/min by a low-compliance pneumohydraulic capillary infusion system (Andorfer Medical Specialties). The location and mean resting pressure at midexpiration of the lower and upper esophageal sphincters, mean esophageal contraction amplitude in the distal esophagus, and percentage of peristaltic contractions in response to ten 5-ml swallows of water were obtained and measured by previously described techniques (14).

Immediately after esophageal manometry, a 2.5-mm-diameter monocrystalline catheter with two antimony pH electrodes (Medtronic Upper Airway, Minneapolis, MN) was passed nasally and positioned with the distal electrode 5 cm above the proximal border of the lower esophageal sphincter and the proximal electrode just below the upper esophageal sphincter. The proximal probe was placed within 3 cm of the upper esophageal sphincter using both commercially available and custom-made probes (Medtronic Upper Airway) with interprobe distances of 10, 12, 15, and 18 cm. The electrodes were calibrated at pH 7 and 1, using a buffer solution (Fisher Scientific, Fairlawn, NJ) before and at the completion of each study. A reference electrode was placed on the anterior chest. Both electrodes were connected to a digital recorder which stored pH data every 4 s. Subjects were sent home with instructions to record meal times, time of assuming the supine position for sleep, and time of arising in the morning. Subjects were encouraged to perform normal daily activities and asked to avoid foods and beverages with a pH < 4. They were also instructed to report respiratory symptoms in a diary and to push the event button on the digital recorder.

After at least 18 h of recording, data were downloaded into an IBM AT personal computer and analyzed separately for the proximal and distal esophageal pH electrodes. Based on 110 healthy control subjects using 95th percentile data in our laboratory, abnormal amounts of acid reflux were present in the distal esophagus if the total percent time pH < 4 exceeded 5.8% during the 24-h study period, or upright acid exposure exceeded 8.2%, or supine acid exposure exceeded 3.5% (15). Based on studies in 20 healthy volunteers, abnormal amounts of proximal reflux occurred if the total percent time pH < 4 exceeded 1.1%, or upright acid exposure exceeded 1.7%, or supine acid exposure exceeded 0.6% (16). Subjects were considered to have reflux present if one or more of these six pH parameters was not in the normal range. The 24-h esophageal pH test also allows correlation of respiratory symptoms with esophageal acid events. Subjects were instructed to record respiratory symptoms and esophageal symptoms. Respiratory symptom and esophageal acid correlation were assessed by reviewing patient diaries, digital recording event markers, and esophageal pH tracings. Respiratory symptoms monitored included wheezing, chest tightness, shortness of breath, and cough. Other symptoms monitored included chest pain, heartburn, regurgitation, and nausea. A respiratory or esophageal symptom was associated with a reflux event if the esophageal pH was < 4 simultaneously with the symptom or within 5 min before its onset. There are minimal data evaluating the symptom-reflux correlation on a temporal basis (17). Investigators have used temporal relationships as long as 10 min and as restrictive as 2 min (18, 19). In our pH laboratory, 5 min has been used routinely for symptom correlation (10, 20).

Data Collection

All patients completed a standardized asthma and reflux questionnaire before esophageal testing. Asthma-specific questions included age of onset, exacerbating triggers, atopy, seasonal variation, nocturnal symptoms, and a family history of asthma. Asthma symptoms, medication usage, and health care utilization determined asthma severity in accordance with the National Asthma Education Program Expert Panel Report (21). Gastroesophageal reflux questions included the presence and frequency of heartburn, regurgitation, chest pain, dysphagia, hoarseness, and sore throat.

Analysis

Chi-square analysis or the Fisher exact tests were performed to determine if specific asthma demographic variables were associated with abnormal esophageal acid contact times in asthma patients without reflux symptoms. Asthma patients with asymptomatic gastroesophageal reflux (Sx +pH) were compared with 30 previously characterized age-matched asthma patients with symptomatic gastroesophageal reflux (+Sx +pH), to evaluate the severity of gastroesophageal reflux in asthma patients with asymptomatic gastroesophageal reflux (7). Esophageal manometry and pH values from these two groups were compared using the Mann-Whitney rank sum test. Demographic variables, asthma severity, and asthma medications used in asthma patients with asymptomatic gastroesophageal reflux (Sx +pH) were compared with asthma patients with symptomatic gastroesophageal reflux (+Sx +pH) using chi-square analysis or the Fisher exact tests. Data are expressed as mean ± SD (22).

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Demographic and Esophageal pH Values of Asthma Patients Without Reflux Symptoms

Of 220 asthma patients screened, 80 (36%) denied reflux symptoms of whom 58 met entrance criteria with 26 agreeing to participate. Asthma patients without reflux symptoms had a mean age of 43 ± 16 yr, 18 (69%) were women, and 20 (77%) were white. At the time of testing, nine (35%) had mild asthma, 16 (61%) had moderate asthma, and one (4%) had severe asthma as defined by the National Asthma Education Program (21). The esophageal pH probe was well tolerated in all subjects, with none reporting worsening of respiratory symptoms during testing. Asthma medication use included inhaled ₂-agonists on an "as-needed" basis only in five (19%), regular use of inhaled ₂-agonists in 19 (73%), inhaled corticosteroid use in 15 (58%), inhaled nedocromil in three (12%) subjects, inhaled anticholinergics in two (8%), and oral theophylline in three (12%) subjects. None of the subjects used oral ₂-agonists, oral corticosteroids, or leukotriene antagonists. The mean number of medications used per day by each subject was 1.9 ± 1.3. Pulmonary function data show a mean FEV₁ of 85 (± 14) percent predicted, FEV₁/FVC ratio of 68% (± 10), FEF_25-75% of 51 (± 23) percent predicted, and peak expiratory flow rate of 89% (± 20) percent predicted. Methacholine challenge testing was performed in one patient who had a provocative dose causing a 20% fall in FEV₁ (PD₂₀) using normal saline alone.

Table 1 reviews esophageal manometry and pH variables of the 26 asthma patients without reflux symptoms (Sx), consisting of the subset of 10 asthma patients without reflux symptoms with negative 24-h esophageal pH tests (Sx pH) and the subset of 16 asthma patients without reflux symptoms with positive 24-h esophageal pH tests (Sx +pH or asymptomatic gastroesophageal reflux); and the cohort of 30 asthma patients with reflux symptoms with positive 24-h esophageal pH tests (+Sx +pH or symptomatic gastroesophageal reflux). Note that all mean esophageal acid contact times in the Sx group are in the abnormal range. Sixteen (62%) asthma patients without reflux symptoms had at least one abnormal esophageal pH parameter. This 62% prevalence has a wide confidence interval (42% to 80%). In asthma patients with asymptomatic gastroesophageal reflux, 14 (88%) had abnormal esophageal acid contact times at the distal probe during the total period, with five (28%) having supine reflux. Proximal reflux was frequent in the asthma patients with asymptomatic gastroesophageal reflux with 12 (75%) having abnormal total amounts of esophageal acid and five (28%) having supine proximal reflux. Eleven (69%) asthma patients with asymptomatic gastroesophageal reflux had abnormal esophageal acid contact times at both the distal and proximal esophageal pH probes, whereas one (6%) had abnormal values at the proximal probe only. Data at the distal and proximal esophageal pH probes of the asymptomatic gastroesophageal reflux group (Sx +pH) show that 15 of 16 (94%) subjects had more than one abnormal esophageal pH parameter. Two subjects had two, three subjects had three, eight subjects had four, one subject had five, and one subject had six out of six abnormal esophageal pH parameters.

Respiratory symptom correlation showed that 6 of 74 (8%) reported cough episodes, 2 of 2 (100%) reported shortness of breath episodes, 0 of 1 (0%) indicated sputum production episodes, and 1 of 1 (100%) had chest pain episodes associated with esophageal acid exposure. Wheezing or chest tightness was not reported during 24-h esophageal pH testing.

There were no differences in age, asthma duration, body mass index, lower esophageal sphincter pressure, percent peristaltic esophageal contractions, or upper esophageal sphincter pressure between the asymptomatic reflux group (Sx +pH) and the group without gastroesophageal reflux (Sx pH) (Table 1). Asthma patients with asymptomatic gastroesophageal reflux had higher amplitude of esophageal contractions than those without gastroesophageal reflux (p < 0.005). As expected, the group with gastroesophageal reflux had significantly higher amounts of esophageal acid exposure at both the proximal and distal esophageal pH probes than those with normal esophageal acid contact times (p < 0.01 in all esophageal pH variables shown in Table 1 except proximal supine acid exposure [p = 0.13] and the number of episodes lasting greater than 5 min at the proximal probe [p = 0.14]).

Although 24-h esophageal pH monitoring is the best test available to assess gastroesophageal reflux, normal values vary from center to center and values near the definition of "abnormal" may not be reproducible (15). Figures 1 and 2 examine individual esophageal acid contact times in the 26 asthma patients without reflux symptoms (Sx) at the distal probe (Figure 1) and the proximal probe (Figure 2). Normal values are represented below the line on each variable. Open circles represent values from asthma patients with normal esophageal acid contact times (Sx pH), and filled circles represent values from asthma patients with abnormal esophageal acid contact times (Sx +pH). There is a large range of values.

View larger version (16K): [in this window] [in a new window]   Figure 1.   Esophageal acid contact times at the distal esophageal probe in 26 asthma patients without reflux symptoms. Esophageal acid contact times are represented as the percent time in which the esophageal pH was less than 4 at the distal esophageal pH probe located 5 cm above the lower esophageal sphincter. Total values represent the entire recording period, upright values represent time when awake, and supine values represent time of bedtime to morning arising. Open circles represent asthma patients considered to have normal esophageal acid contact times, and solid circles represent asthma patients considered to have abnormal esophageal acid contact times. The solid line represents the point at which values under the line are considered to be in the normal range. Note that asthma patients were considered to have abnormal esophageal acid contact times if one of the six esophageal pH values was in the abnormal range.

View larger version (16K): [in this window] [in a new window]   Figure 2.   Esophageal acid contact times at the proximal esophageal probe in 26 asthma patients without reflux symptoms. Esophageal acid contact times are represented as the percent time in which the esophageal pH was less than 4 at the proximal esophageal pH probe located within 3 cm below the upper esophageal sphincter. Open circles represent asthma patients considered to have normal esophageal acid contact times, and solid circles represent asthma patients considered to have abnormal esophageal acid contact times. The solid line represents the point at which values under the line are considered to be in the normal range. Note that asthma patients were considered to have abnormal esophageal acid contact times if one of the six esophageal pH values was in the abnormal range.

Predictors of Gastroesophageal Reflux in Asthma Patients without Reflux Symptoms

Chi-square and the Fisher exact test analyses showed that no demographic variable (including childhood onset asthma, family history of asthma, asthma severity, seasonal variation, history of atopy, nocturnal symptoms, sore throat, hoarseness, nocturnal awakenings associated with dyspnea, wheezing with eating or alcohol use) predicted abnormal esophageal acid contact times.

Demographic Variables of Asthma Patients with Symptomatic Gastroesophageal Reflux

Tables 1 and 2 review demographic and esophageal variables of the 30 asthma patients with symptomatic gastroesophageal reflux (+Sx +pH). Pulmonary function data show a mean FEV₁ of 72 (± 25) percent predicted, FEV₁/FVC of 65 (± 13), FEF_25-75% of 44 (± 35) percent predicted, and peak expiratory flow rate of 80 (± 25) percent predicted. Methacholine challenge testing was not required for the diagnosis of asthma in these 30 patients. Respiratory symptom correlation showed that 1 of 1 (100%) of reported shortness of breath episodes, 15 of 32 (47%) of wheezing, 14 of 36 (39%) of cough episodes, and 4 of 7 (57%) episodes of chest pain were associated with esophageal acid events. Seventy-one of 86 (83%) heartburn symptoms and 24 of 28 (86%) regurgitation episodes were associated with esophageal acid events.

Comparing Asthma Patients with Asymptomatic Gastroesophageal Reflux with Asthma Patients with Symptomatic Gastroesophageal Reflux

Table 2 reviews demographic characteristics and Table 1 reviews esophageal manometry and acid contact times in 16 asthma patients with asymptomatic gastroesophageal reflux (Sx +pH) and 30 age-matched asthma patients with symptomatic gastroesophageal reflux (+Sx +pH). Asthma patients with asymptomatic gastroesophageal reflux had less severe asthma, and were not using theophylline or oral corticosteroids. Asthma patients with asymptomatic gastroesophageal reflux had higher lower esophageal sphincter pressures and higher amounts of total proximal esophageal acid exposure compared with asthma patients with symptomatic gastroesophageal reflux. There were no significant differences between groups in asthma duration, body mass index, mean amplitude of esophageal contractions, percent of peristaltic contractions, upper esophageal sphincter pressure, or distal esophageal acid exposure.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

This prospective cohort study examines the prevalence of gastroesophageal reflux, defined as abnormal amounts of esophageal acid on 24-h esophageal pH tests, in asthma patients without reflux symptoms. The high prevalence rate of 62% (16 of 26) of asthma patients with "clinically silent" reflux illustrates that gastroesophageal reflux may be commonly associated with asthma, even in the absence of reflux symptoms. Identifying "clinically silent" gastroesophageal reflux in asthma patients requires 24-h esophageal pH testing. Demographic asthma variables including nocturnal asthma symptoms were not helpful in identifying asthma patients with asymptomatic gastroesophageal reflux. However, this study has limited power to identify characteristics associated with reflux among asymptomatic patients because of the small patient population, including whether asthma medications predispose to the development of reflux. Furthermore, this prevalence rate of 62% is associated with wide confidence levels. Also, this is a preliminary report, performed in a university-based clinic setting, and may not represent the general asthma population.

The severity of reflux in asthma patients with asymptomatic gastroesophageal reflux is not less than asthma patients with symptomatic gastroesophageal reflux. Compared with asthma patients with symptomatic gastroesophageal reflux, asthma patients with asymptomatic gastroesophageal reflux had higher lower esophageal sphincter pressures, but similar esophageal acid contact times at the distal esophageal pH probe (5 cm above the manometrically defined lower esophageal sphincter). This finding may be partially explained by two factors. First, asthma patients with symptomatic gastroesophageal reflux were more likely to be on theophylline (p < 0.005) and theophylline may decrease lower esophageal sphincter pressure (23). Second, asthma patients with symptomatic gastroesophageal reflux had more severe asthma as characterized by the National Asthma Education Program than did asthma patients with asymptomatic gastroesophageal reflux. The diaphragmatic crura participates in lower esophageal sphincter pressure generation (24, 25). Hyperinflation associated with bronchospasm may place the diaphragmatic crura at a functional disadvantage because of geometric flattening (26, 27). Although no studies to date have examined asthma patients, they may be more prone to developing functional impairment of the diaphragmatic crura resulting in alterations in lower esophageal sphincter pressure generation.

Asthma patients with asymptomatic reflux had higher amounts of proximal esophageal acid compared with gastric patients with symptomatic reflux. Interestingly, in two previously reported studies, proximal esophageal acid predicted asthma improvement with antireflux therapy (7, 28).

Four studies prior to our own investigation have noted that asthma patients may have asymptomatic gastroesophageal reflux (7, 9, 28, 29). Larrain and coworkers prospectively examined adult onset nonallergic asthma patients and found that 21 of 81 (26%) subjects had gastroesophageal reflux without ever experiencing heartburn (29). Our study differs from Larrain and coworkers' study in two ways: the patient population selected (nonallergic asthma without a family history of asthma), and, most importantly, the technique by which gastroesophageal reflux was diagnosed (an esophagram after a 300-ml barium meal which had an 8% false-positive rate and a 56% false-negative rate in their patient population) (29). Irwin and coworkers also noted prospectively that 24% of difficult-to-control asthma patients with gastroesophageal reflux had no reflux symptoms (9). Our study differs from Irwin and coworkers' study in that their patient population included only asthma patients who required more than 10 mg of prednisone every other day for at least three consecutive months per year. In a retrospective study, Harding and coworkers reviewed 24-h esophageal pH test results in 35 asthma patients without reflux symptoms, of whom 10 (29%) had abnormal esophageal acid contact times. These asthma patients were referred to the esophageal pH laboratory, thus selection bias may have been introduced (10). Finally, Schnatz and coworkers, in a retrospective review of asthma patients and patients with chronic cough, found that eight of 35 (23%) with gastroesophageal reflux presented without esophageal symptoms (28). Schnatz and coworkers' study included a mixed population of patients with pulmonary disease (asthma and chronic cough) and examined consecutive patients referred for esophageal evaluation because of the clinical suspicion that gastroesophageal reflux might be causally related to the pulmonary symptoms (28).

It is unlikely that this study overestimated the prevalence of gastroesophageal reflux because false-positive tests are rare (< 5%), and in fact, false-negative results are much more common (range 10 to 25%) (4). Also, all but one (6%) of the asthma patients with asymptomatic gastroesophageal reflux had more than one abnormal esophageal pH parameter with 13 (81%) having three or more abnormal esophageal pH parameters.

Irwin and coworkers showed in difficult-to-control asthma patients that identifying and treating gastroesophageal reflux, regardless of the presence of reflux symptoms, improved asthma control (9). The association between asthma and gastroesophageal reflux is complex. For instance, esophageal acid caused a decrease in peak expiratory flow rates in asthma patients with gastroesophageal reflux. In a guinea pig model, Hamamoto and coworkers found that esophageal acid infusion results in the release of airway substance P and that this release of substance P was coupled with airway edema (30). However, there are conflicting data. Likewise, therapeutic trials using a cross-over design have not shown significant improvement in asthma outcomes (31, 32). For example, a double-blind, placebo-controlled cross-over study of 107 asthma patients using omeprazole 40 mg daily or placebo for 8 wk showed an improvement in nocturnal asthma symptoms during the omeprazole phase; however, there may have been an order effect from the cross-over (33). Furthermore, Boeree and coworkers evaluated 30 asthma patients with gastroesophageal reflux using omeprazole 40 mg twice daily for 3 mo in a randomized, double-blind, placebo-controlled parallel manner trial, and found no difference in spirometry, asthma symptoms scores, medication use, peak expiratory flow rates, or methacholine PD₂₀ (34). Unfortunately, there was difficulty with patient compliance with 13 (43%) of the subjects taking less than 75% of the study drug (34). Recently, Field and Sutherland reviewed all English studies in the MEDLINE database with 326 medically treated asthma patients with gastroesophageal reflux, noting that asthma symptoms improved in 69% of subjects, asthma medications were reduced in 62%, and evening peak expiratory flow rates improved in 26% of patients without pulmonary function test improvement (8). Likewise, antireflux surgery performed on 417 asthma patients improved asthma symptoms, asthma medications use, and pulmonary function in 79%, 88%, and 27%, respectively (35). Clearly, the association between asthma and gastroesophageal reflux needs further investigation. A double-blind, placebo-controlled, multicentered trial evaluating asthma outcomes with aggressive medical antireflux therapy using a proton pump inhibitor has not been reported to date.

This report suggests that gastroesophageal reflux is present in patients with stable asthma, even in the absence of esophageal symptoms with a prevalence rate of 62%, and it shows the value of esophageal pH monitoring because demographic variables did not identify this asthma population. The severity of gastroesophageal reflux in asthma patients with clinically silent reflux is not less severe than in asthma patients with reflux symptoms. In fact, asthma patients with asymptomatic gastroesophageal reflux had higher amounts of proximal esophageal acid exposure. Asthma patients with asymptomatic gastroesophageal reflux should be included in future studies evaluating asthma outcome with aggressive antireflux therapy in asthma patients with gastroesophageal reflux.