INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Gastroesophageal reflux disease (GERD) is a common condition, affecting as much as 20% of the adult population. Case reports and studies of patients with chronic upper airway symptoms describe a potential relationship between GERD and upper airway conditions. The upper airway symptoms described in these subjects are primarily cough, hoarseness, and throat complaints (1). The hypothetical mechanisms linking upper respiratory symptoms (URS) with GERD include direct irritation of the larynx, pharynx, and posterior nasal mucosa by gastric fluid and neurogenic inflammation triggered by esophageal irritation. An association of GERD with nasal symptoms has been suggested but no data are available to confirm an increased prevalence of nasal symptoms in patients with GERD (3, 4). URS in patients with GERD may contribute to the association of GERD and asthma (5).

The relationship of URS and GERD is based upon investigations of subjects with persistent URS. In the majority of these cases, GERD was asymptomatic or the symptoms were not sufficiently severe to warrant an evaluation of GERD before the evaluation of URS (1). The prevalence of URS in subjects with GERD has not been investigated.

This study reports the prevalence of URS in subjects undergoing a 24-h, dual-channel pH probe study for suspected GERD (6). URS were assessed with a questionnaire validated in preliminary studies (9).

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

A total of 231 consecutive subjects referred to the Joy McCann Culverhouse Center for Swallowing Disorders at the University of South Florida for suspected GERD were interviewed and completed the questionnaire. Seventy-four of these subjects (32%) met the inclusion criteria. Inclusion criteria were chronic heartburn as the presenting symptom and suspected GERD not treated at the time of the evaluation. Exclusion criteria were chronic obstructive pulmonary disease, restrictive pulmonary disease, pulmonary malignancies, history of cigarette smoking, pregnancy, or laryngeal stenosis. Subjects who underwent pH probe studies for reasons other than suspected GERD were also excluded. A subject with myasthenia gravis was excluded. Subjects with proximal reflux according to the Dobhan criteria but negative for distal reflux according to the DeMeester criteria were excluded (6).

The subjects underwent 24-h proximal (upper) and distal (lower) esophageal pH studies with a dual-channel device to assess GERD. The probes were connected to a Synectics data storage device (EsopHogram Ver. 5.70C2; GastroSoft, Inc., Irving, TX). Gastroesophageal acid reflux was evaluated by the following parameters: (1) number of reflux episodes per 24 h, (2) total time of esophageal pH < 4 in 24 h, (3) total time of pH < 4 when supine, (4) total time of pH < 4 when upright, (5) number of reflux episodes with duration longer than 5 min, and (6) duration of longest reflux episode. The use of composite scores based on multiple parameters has a sensitivity of 96%, specificity of 100%, and accuracy of 98% in the diagnosis of GERD (10).

All selected subjects completed a URS questionnaire. This questionnaire evaluated symptoms during the 4 wk preceding esophageal pH monitoring. Subjects were not told that the purpose of this study was to assess the relationship between URS and GERD. Questions on URS were dispersed among questions on demographic data, past history, comorbid conditions, and GERD symptoms and medications. Collection of the questionnaire answers was completed the same day or within 4 wk of the esophageal pH study. Symptom occurrences were scored as 0, 1, or 2, for absent, infrequent (up to 4 d/mo), or frequent (> 4 d/mo), respectively. URS were evaluated with questions on seven symptom groups providing a score range of 0-14. The symptoms assessed were: hoarse voice, throat clearing; globus (lump in throat; sensation of foreign body stuck in throat, tightening of throat); cough; morning sore throat, dryness, bitter taste; nasal congestion, postnasal drip, clear anterior rhinorrhea, sneezing, nasal itching; sinus headache and pressure or sinusitis; and earache, fullness or otitis media.

Symptoms were grouped as nasal, laryngeal, sinusal, pharyngeal, and aural. Nasal symptoms included congestion, postnasal drip, clear rhinorrhea, sneezing, or nasal itching. Laryngeal symptoms included cough, hoarseness, throat clearing, voice changes, or globus. Pharyngeal symptoms included sore throat, bitter taste, or mouth dryness in the morning upon waking up. Sinus symptoms included headache, facial pressure, mucopurulent nasal discharge, or abnormal sinus computed tomographic (CT) scan. Aural symptoms included earache, ear discharge, or documented otitis media. These definitions are arbitrary. Postnasal drip, rhinorrhea, globus, throat clearing, or cough could have been attributed to more than one anatomic site. Each symptom was recorded only once under one symptom group.

Twenty-two subjects, 15 with abnormal and seven with normal esophageal pH studies, completed the questionnaire twice: the day the esophageal pH monitoring was started and 4 wk later. None of the 15 subjects with abnormal pH studies had been treated for GERD for longer than 10 d by the time they completed the questionnaire for a second time. The accuracy of the answers obtained was estimated at 92%. Standard deviation in the scores obtained previously from subjects with heartburn was 3.84 (9). The boundary on the error of estimation was set at 0.2 unit of the URS questionnaire score range 0-14.

Seventy-four normal subjects without GERD symptoms completed the URS questionnaire. These normal volunteers were University of South Florida employees or their family members. Some of these subjects may have had asymptomatic gastroesophageal reflux. Twenty-two volunteers were asked to complete the questionnaire for a second time. The accuracy of their responses was 87%.

Subjects included in this study were categorized as: (1) normal volunteers without heartburn, (2) subjects with heartburn and normal pH probe studies, (3) subjects with lower reflux, and (4) subjects with combined upper and lower upper reflux.

Subjects with distal esophageal reflux had an abnormally high DeMeester score (normal < 14.72) (6). This is a composite score resulting from added readings of the following parameters recorded by the distal esophageal probe: (1) 50 or more reflux episodes/24 h; (2) esophageal pH < 4 for a period  4.2% of the recorded time; (3) esophageal pH < 4 for a period  1.2% of the recorded time, while in the supine position; (4) esophageal pH < 4 for a period  6.3% of the recorded time, while in the upright position; (5)  4 reflux episodes that lasted longer than 5 min; (6) longest recorded reflux episode lasting  9.2 min. Subjects with proximal esophageal reflux had abnormal readings according to the Dobhan criteria, which incorporate the following parameters recorded by the proximal esophageal probe: (1)  18 reflux episodes/24 h, (2) esophageal pH < 4 for a period  0.9% of the recorded time, (3) esophageal pH < 4 for any percentage of the recorded time while in the supine position, (4) esophageal pH < 4 for a period  1.3% of the recorded time while in the upright position, (5) any reflux episodes that lasted longer than 5 min, (6) longest reflux episode lasting  3.9 min (7, 8). Upper gastroesophageal reflux, defined by the Dobhan criteria, requires the occurrence of lower reflux. Proximal reflux, detected by pH probe measurements, may occur with normal readings for distal reflux. This event may represent imperfections in the definitions of proximal and distal reflux, rather than a true pathophysiologic entity. Esophageal pH monitoring by composite scoring is a sensitive method for detecting GERD, but some subjects with normal pH monitoring results may have GERD (10).

Nonparametric statistics were used to assess the significance of the URS scores of these groups. A value of p < 0.05 was considered significant. The study protocol was approved by the institutional review board of the University of South Florida and informed consent was obtained from all participants.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Demographic features of the study subjects and normal volunteers are shown in Table 1. The two groups were comparable with respect to age and sex composition. Twenty-two of 74 subjects with heartburn had normal pH probe studies, 16 had lower reflux, and 36 upper reflux. Thirty-one subjects completed the questionnaire the day their esophageal pH monitoring started. Composition by sex of the subgroups identified by the esophageal pH studies varied (Table 1). Esophageal pH data are shown in Table 2. The mean URS score (± SD) for the volunteer group was 3.04 ± 2.71. Mean URS scores for the subjects with normal pH studies, lower reflux, and combined upper and lower reflux were 4.57 ± 3.57, 7.31 ± 3.72, and 8.75 ± 4.25, respectively (Table 3). URS scores are elevated in subjects with GERD (Figure 1). There is no significant difference between URS scores from volunteers and subjects with normal pH probe studies, p = 0.71. The difference between URS scores of subjects with lower and upper reflux is also not significant, p = 0.79. Upper respiratory scores significantly correlate with the number of reflux episodes recorded during a 24-h period, r = 0.47, p = 0.0001 (Figure 2).

View larger version (15K): [in this window] [in a new window]   Figure 1.   URS scores in volunteers and subjects who underwent 24-h esophageal pH monitoring.

View larger version (15K): [in this window] [in a new window]   Figure 2.   Correlation of URS scores with the number of distal esophageal reflux episodes during 24-h pH probe studies. Regression line: y = 0.01x + 5.21, r2 = 0.22, r = 0.47, p = 0.0001. Vertical line: upper normal limit of reflux episodes/24 h (50 episodes).

Laryngeal symptoms for more than 4 d/mo were reported by 7 of 74 (9%) volunteers, 8 of 22 (36%) subjects with normal pH probe studies, 11 of 16 (68%) subjects with lower reflux, and 27 of 36 (75%) of subjects with combined upper and lower reflux. Subjects with abnormal esophageal pH studies have a higher prevalence of laryngeal symptoms lasting longer than 4 d/mo compared with subjects with normal pH studies or normal volunteers, p < 0.001 (Table 3).

Nasal symptoms present for more than 4 d/mo were described by 11 of 74 (14%) volunteers, 7 of 22 (31%) subjects with normal pH studies, 8 of 16 (50%) subjects with lower reflux, and 25 of 36 (69%) subjects with combined upper and lower reflux. Nasal congestion was the most frequently reported symptom. Postnasal drip was the second most common nasal symptom. The prevalence of nasal symptoms among subjects with GERD is significantly greater compared with subjects with normal pH studies or normal volunteers, p < 0.02 and p < 0.001, respectively (Table 3). The prevalence of pharyngeal, sinus, and aural symptoms was not significantly higher among subjects with GERD (Figure 3).

View larger version (31K): [in this window] [in a new window]   Figure 3.   Comparison of symptom prevalence in volunteers, subjects with normal esophageal pH studies, and subjects with GERD.

Forty-four (85%) subjects with GERD reported at least one URS present for more than 4 d/mo, compared with 13 (59%) subjects with normal pH studies. Thirty-five (67%) subjects with GERD reported two or more symptoms from different URS groups for longer than 4 d/mo, compared with seven (32%) subjects with normal pH studies, p < 0.01 (Table 4). Occurrence of laryngeal symptoms along with any other URS for longer than 4 d was reported by 29 (56%) subjects with GERD, compared with 5 (23%) subjects with normal pH studies, p < 0.01. Occurrence of nasal symptoms along with any other URS for longer than 4 d was reported by 31 (59%) subjects with GERD, compared with six (27%) subjects with normal pH studies, p < 0.02. Symptoms from more than one symptom group were reported at similar rates in subjects with upper and lower reflux.

The odds ratio for GERD proven by positive pH monitoring studies is 1.5 for a subject with URS score greater than 4. The odds ratio for demonstrating acid reflux on a pH monitoring study increases to 3.8 or 6.8 if the URS score equals 8 or 11, respectively. Forty-one subjects of the 52 (78.8%) with GERD had a URS score greater than 4. Eleven subjects of the 22 with normal pH studies had a URS score greater than 4 (positive predictive value 78.8%).

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Previous studies demonstrated GERD in patients with chronic sinusitis, laryngitis, and pharyngitis and support the consideration of GERD in patients with upper airway symptoms recalcitrant to treatment (1). These investigators postulated a causal relationship between GERD and upper airway conditions and maintained that persistent URS are manifestations of GERD (1). Support of this hypothesis requires the demonstration of an increased prevalence of URS in a population of patients with GERD. The subjects of the present study were selected on the basis of their chronic heartburn and not upper airway pathology. The diagnosis of GERD in these subjects was supported by esophageal pH monitoring. Collection of the questionnaire answers was completed without the subjects being aware that the association of URS with GERD was being studied. The present study demonstrates that subjects with symptomatic GERD and positive pH studies have significantly higher URS scores than subjects with heartburn and negative pH probe studies. The studied subjects were recruited among patients with heartburn and the normal volunteers were only evaluated by questioning for the presence of GERD-related symptoms. Therefore, the observations of this study may not apply to subjects with asymptomatic reflux.

URS scores are proportional to the severity of GERD expressed in numbers of reflux episodes recorded by 24-h monitoring. The proportional increase of URS scores with reflux episodes supports an association between GERD and URS. Although resistance to acid exposure may differ from patient to patient, esophageal mucosal injury and motility impairment correlate with exposure to acid and severity of reflux patterns (11). The severity of reflux seems to affect the perception of symptoms in the patient with heartburn. Reflux episodes in patients with heartburn lower the esophageal sensory threshold and lead to increased "visceral sensitivity" (12). This effect has been demonstrated in subjects with GERD as well as in subjects with heartburn and normal esophageal pH studies (12). The URS scores in subjects with heartburn and normal esophageal pH in this study were not statistically increased compared with normal volunteers, but the scores suggested a difference, 4.57 ± 3.57 versus 3.04 ± 2.71. A larger population comparison would be necessary to definitely answer this question. This study suggests that the number of reflux episodes in subjects with heartburn is associated with increased occurrence and awareness of URS. A similar association of symptomatic GERD with symptoms originating in the lower respiratory tract has been described among patients with asthma (13).

There is no significant difference in the prevalence of URS in subjects with distal (lower only) and proximal (combined upper and lower) esophageal acid reflux. Severity of URS, need for specific treatment, and persistence of symptoms beyond 5 d/mo were not evaluated. These observations suggest that URS are associated with GERD even in the absence of gastric acid contact with the upper airway mucosa. Examination of this hypothesis would require ambulatory upper airway pH monitoring.

Neurogenic inflammation, which is implicated in lower respiratory complications of GERD, may be involved in the pathogenesis of URS in patients with GERD (14). The lower esophagus and the upper airways receive sensory and motor innervation from vagal branches. Nerves can activate immune cells through neuropeptides. Large numbers of mast cells attached to peripheral nerves show an electrical response to substance P. Mast cells become more sensitive to substance P- induced degranulation after repetitive exposure to substance P (15). Vagal impulses from the lower esophagus to the lower respiratory tract cause release of substance P and inflammation (14). Repeated neural impulses generated from acid contact with the esophageal mucosa may similarly prime mast cells and cause degranulation in the upper airway. The effect of neuropeptides on mast cells can be enhanced by the action of inflammatory cells. Activated eosinophils, which play a prominent role in allergic as well as in some cases of nonallergic rhinitis, produce substance P, stimulate neural production of substance P, and activate mast cells directly (16). Vagal interactions with the immune cells of the upper airway have not been studied. An interaction of the vagus and other nerves with effector immune cells would be consistent with the correlation of URS scores with the number of lower esophageal reflux episodes. The proposed neuroimmune relationship would also explain the similar rates of URS in subjects with upper reflux compared with reflux limited to the lower esophagus.

Nasal symptoms are frequent among subjects with GERD in this study. Nasal symptoms are present for more than 4 d/mo in 63% of subjects with pH probe proven GERD. This is consistent with data demonstrating nasal hyperreactivity with histamine provocation in patients with GERD (17). The frequency of nasal symptoms is second to the 73% prevalence of laryngeal symptoms in subjects with GERD.

The difference in the URS scores when subjects with GERD are compared with subjects with a normal esophageal pH study is primarily a result of laryngeal and nasal symptoms. Subjects with GERD tend to report symptoms from more than one symptom group. A single URS was reported by only nine subjects with GERD (17%) compared with six (27%) subjects with normal pH studies. Among subjects with GERD who experienced any URS for more than 4 d, 80% recorded symptoms from two or more symptom groups compared with 53% of subjects with normal pH studies. Laryngeal and nasal symptoms appreciated either alone or in concurrence with other URS are significantly more frequent among subjects with GERD compared with subjects with heartburn and normal pH studies. Furthermore, URS scores obtained by summation of individual URS are higher in subjects with GERD. These observations are in accordance with the sensitivity of the respiratory tract to the effects of symptomatic GERD and the evidence that the lower esophagus originates from the primordial trachea (13, 18).

Bias in reporting symptoms is possible when questionnaires are used, even though subjects in this study were not aware of the studied association of URS with GERD when they completed the questionnaire. Subjects undergoing invasive studies for chronic symptoms may overrate their URS. Subjects with heartburn and negative studies for GERD showed higher URS scores than those obtained from normal volunteers, but the difference was not significant. Seventy-nine percent of subjects presenting with chronic heartburn and reporting a URS score greater than 4 had abnormal esophageal pH studies. URS questionnaire data could be used in the evaluation of patients with chronic heartburn to augment the predictive value of their history.

A therapeutic trial for GERD for 3 to 6 mo may be indicated for patients with chronic URS and suspected GERD (19). This hypothesis requires further investigation. If a therapeutic trial of GERD is initiated for management of URS, this questionnaire could be useful as an instrument to quantify and monitor symptoms. Failure of URS to respond after such empiric therapy may be related to either a non-GERD origin or inadequate suppression of gastric acid secretion. A 24-h intragastric pH study with the patient on acid suppression therapy will make this determination.

In conclusion, URS are frequent among subjects with symptomatic GERD diagnosed by esophageal pH studies. The correlation of URS with reflux episodes supports an association of URS with symptomatic GERD. Most subjects with GERD report symptoms involving more than one upper airway site. Nasal symptoms occur frequently among patients with GERD. Studies are needed to determine the cellular and cytokine features of upper airway pathology in GERD and their response to GERD treatment.