There is no doubt that gastro-oesophageal reflux can cause a chronic cough. However, how frequently reflux is the underlying cause in patients presenting to the readers of Thorax is a matter of much debate. This confusion can be laid squarely at the door of the gastroenterologists who have taken one symptom of acidic reflux—heartburn—and made it the sine qua non for gastro-oesophageal reflux disease. This characterisation of gastro-oesophageal reflux disease as heartburn has led to the denial of the non-acid extra-oesophageal symptoms of reflux. In reality, however, reflux is almost universal in humans because our upright posture has disrupted the anatomy of the lower oesophageal sphincter. Measurement of electrical impedance within the gullet in fact shows that only a small number of reflux episodes are acidic (below pH 4)¹ and, while it takes a lot of acid to burn the hardy oesophagus, anyone who has performed a bronchoscopy will know that the delicate larynx and airways respond to the most gentle of stimulation. The difference between reflux causing respiratory symptoms and gastro-oesophageal reflux disease is neatly demonstrated by cough after meals. Patients with postprandial reflux cough do so approximately 10 min after food.² This is the time of peak transient opening of the lower oesophageal sphincter and combats aerophagy. However, heartburn does not occur until later because stomach acid has been neutralised by the meal.

What evidence is there that reflux is an important cause of chronic cough? In a recent survey of normal subjects, Ford et al³ found that 7% reported a chronic cough sufficient to interfere with activities of daily living. After correction for factors such as cigarette smoking, gastrointestinal symptoms including regurgitation and irritable bowel (but not heartburn) were highly correlated with cough. However, as with cigarette smoking and lung cancer, epidemiology can never prove a causal link—merely suggest associations. Can we be sure if there is no specific diagnostic test? The answer is not yet but, for the clinician, the precipitation of cough by factors known to cause transient opening of the lower oesophageal sphincter such as rising, phonation and postprandially gives the game away. Other extra-oesophageal symptoms such as dysphonia, rhinitis and a funny taste in the mouth are also present in subjects with cough with proven reflux disease.²

The approach taken in this latter study, to explore the symptoms of reflux cough by assessing patients with proven acid reflux, has been adopted in a study by Patterson and colleagues published in this issue of Thorax (see p 491).⁴ They investigated the profile of tachykinins present in induced sputum from patients with asthma (as defined by bronchial lability) and patients with cough without reversibility or methacholine hyper-responsiveness. They performed 24 h pH monitoring to define those in each group who had acid reflux and found that patients with acid reflux had higher tachykinin levels. Of the several possible explanations for this phenomenon, they favour the reflex neurogenic release of the peptides. What is very interesting about this study is that, for the first time, a difference has been detected in the profile of patients with different phenotypes of cough. When patients with chronic cough have previously been studied by histological examination,⁵ induced sputum inflammatory markers^6,7 or neurotrophin profiles,⁸ no difference has been detected, suggesting to some that chronic cough is a single syndrome. While this is still possible, and acid reflux may be merely stimulating an epiphenomenon of tachykinin release, this is the first study to define a unique phenotype which may have important consequences for treatment. Current treatment for reflux of respiratory importance is less than satisfactory. Because this sort of reflux is less acid-dependent, even twice daily proton pump inhibitors only produce a response in at most half of patients. Unsurprisingly, drugs that act on the motility of the gullet such as metaclopramide and domperidone can produce pleasing responses. Baclofen, which mimicks vagal inhibition of lower oesophageal sphincter opening, is our last specific treatment for reflux cough. The finding of higher levels of tachykinins in sputum from patients with acid reflux-related asthma and cough suggests an urgent need for neurokinin antagonists to be studied in these patients.