Prevalence and clinical manifestations of gastro-oesophageal reflux-associated chronic cough in the Japanese population

Despite the established evidence that gastro-oesophageal reflux (GOR) causes 10 to 40% of chronic cough [1], and the prevalence of GOR-associated chronic cough (GOR-CC) is increasing in the Western populations [2], this condition has been rarely reported in Japan and its clinical manifestation is not well characterized. Only one case among 37 patients with chronic dry cough was diagnosed as having GOR-CC in our previous study carried out in the mid '90s [3]. Our present study is concerned with the rising number of cases of GOR-CC in Japan and of concomitant reflux laryngitis which is another major extra-oesophageal manifestation of GOR.

We studied 112 consecutive adult patients with chronic cough (≥ 8 weeks) who newly visited the asthma and cough clinic of Kyoto University Hospital from June 2002 to December 2003. Diagnostic investigations included questionnaire, physical examination, blood tests, chest and sinus radiographs, pulmonary function, airway responsiveness and cough sensitivity tests, and sputum induction. Diagnosis of GOR-CC was on the basis of response to 8 week-course of a proton-pump inhibitor (PPI, lansoprazole™) and/or positive results of 24 h ambulatory esophageal pH monitoring (pH Digitrapper MarkII Gold 6,200, Synetics Medical Comp., Sweden) [4]. Laryngeal involvement was based on symptoms or objective diagnosis by specialists; laryngeal irritation, globus sensation, or signs of laryngeal inflammation. Diagnosis of cough variant asthma (CVA) was based on the following criteria; an isolated chronic cough without dyspnea or wheezing not audible on auscultation, airway hyperresponsive to methacholine and symptomatic improvement of coughing with the use of inhaled beta-₂ agonists, sustained release theophylline or both, no past history of asthma, or upper respiratory tract infection within the past 8 weeks [5]. When patients did not undergo methacholine test due to failure of informed consent or presented normo-responsive result but responded to bronchodilator therapy, they were diagnosed as having probable CVA[6]. Diagnosis of sinobronchial syndrome was made on a positive result of sinus images and improvement of cough as well as the symptom related to chronic sinusitis with antibiotics [7, 8]. Diagnosis of atopic cough was made according to the criteria proposed by Japanese Cough Research Society [6, 9]. If examinations and intensive therapeutic trials for GOR-CC, CVA, sinobronchial syndrome, and atopic cough including inhaled corticosteroids and anti-reflux treatment were failed, the chronic cough was considered unexplained (idiopathic). All the chronic cough patients showed normal chest radiographs.

Causes of chronic cough of the 112 patients were as follows; 38 CVA, 24 probable CVA (12 patients did not undergo airway responsiveness test and 12 presented normo-responsive results but responded to bronchodilator therapy), 17 atopic cough, 9 sinobronchial syndrome, 8 GOR-CC, 7 postinfectious chronic cough, 2 other miscellaneous conditions, 4 unexplained cough. Twelve patients were lost before diagnosis was made. Nine patients had multiple conditions. There were 15 ex-smokers, and 2 current smokers. No patients had ACE inhibitor associated chronic cough. The 8 patients (7.1%) with GOR-CC were diagnosed on the basis of response to the PPI (n = 7) and/or positive results of 24 h ambulatory esophageal pH monitoring (n = 4). One patient who complained of chronic cough and heartburn but did not respond to the PPI showed esophageal pH more than 7 in 66.9% of the 24 h monitoring period, and was diagnosed as having GOR-CC due to alkaline regurgitation.

Demographics of the 8 patients were presented in Table 1. Mean body mass index (BMI) was classified as normal at 23.9 (range, 19.4–28.0) kg/m², which was not significantly different from that of the general population in the present study (23.3 kg/m², 16.5–36.3 kg/m²). Frequent association of reflux laryngitis (5 of 8) was observed. Two patients complained of temporal association of coughing and heartburn. Five patients were complicated with other causes of chronic cough; 3 with CVA/probable CVA on inhaled corticosteroids or an anti-leukotriene receptor antagonist, 1 with sinobronchial syndrome on low dose of macrolide, 1 with atopic cough on an anti-histamine receptor antagonist. Airway responsiveness was tested in 4 patients among whom 2 with complication of CVA showed hyperresponsiveness. Cough sensitivity was examined in 2 patients; one complicated with atopic cough had hypersensitivity, another with CVA did not. All were non-smokers and produced minimal amount of sputum or none.

In 7 patients, their coughing with or without laryngeal symptoms was alleviated and ceased within a few days after initiation of PPI. However, in 3 of 4 patients who had stopped treatment with resolution of cough, coughing recurred 1 week to 8 months later. Coughing, laryngitis and heart burn observed in one patient with alkaline regurgitation subsided without any treatment.

In the present study, GOR-CC was a less frequent cause of chronic cough than in the studies of the Western populations [1, 2], although there was a small increase when comparing with our previous study [3]. The low prevalence of GOR-CC in the present study might be biased by the studied population in a university hospital, but may reflect low prevalence of GOR in the general population in Japan where 21 to 27% of the general population are overweight (BMI ≧ 25 kg/m²) [10] and endoscopically diagnosed and/or symptomatic GOR is reported to be 16% and 18% in population based studies [11, 12]. This is in contrast with the finding in Western countries where 34 to 78% of the general population are overweight [10] and 21 to 59% of the general population have symptomatic GOR [13].

In one patient who complained of cough and heartburn, and presented laryngitis and frequent alkaline regurgitation, the cough had ceased spontaneously as well as heartburn and laryngitis. Although there are no diagnostic criteria for alkaline regurgitation, the patient's cough was clinically considered caused by GOR-CC due to alkaline regurgitation.

The present study has shown frequent association of reflux laryngitis and GOR-CC. To date, frequency of this association has not been clarified [2]. In a study of extra-oesophageal manifestations in GOR, laryngeal manifestations are observed in 10.4% of patients with GOR and are significantly related to higher age, longer GOR duration and obesity [14], although these features were not the cases for the 5 patients with average age of 47.4 (28–59) years and BMI of 22.4 (19.2–28) kg/m² in the present study. Except the frequent incidence of concomitant reflux laryngitis and standard range of BMI, characteristics of the present patients including complication with other causes of chronic cough were consistent with those in the previous studies of GOR-CC in the Western populations [1, 4, 15, 16]. Since chronic cough per se can be a trigger of GOR possibly through increased transdiaphragmatic pressure and transient lower oesophageal sphincter relaxation [17], GOR-CC should be considered when cough remains despite the institution of specific treatment to other causes of cough.

We conclude that GOR-CC is a less frequent cause of chronic cough in Japan than in Western countries. The presence of reflux laryngitis may be an important clue to suspicion of GOR-CC.

The ethics committee of our institution approved the study protocol and a written informed consent was obtained from each participant.

This study was supported by Takeda Pharmaceutical Company ltd.